The Senior Microsoft 365 Administrator Operating Model
KTBR + KTBE Master Task Catalog, Automation Backlog, Calendar, Dashboard and Career Roadmap
Scope: Whole-of-tenant Microsoft 365 operations — Entra ID, Exchange Online, SharePoint Online, OneDrive, Teams, Intune, Defender XDR, Purview, M365 Admin, M365 Apps, Copilot, Viva, Windows 365, Planner/To Do, Power Platform, Microsoft Graph, PowerShell, Security, Compliance, Governance, Monitoring, Licensing, Business Continuity, Automation, Documentation.
Audience: Experienced M365 administrator / architect operating an enterprise tenant.
Currency note: Microsoft renames portals, retires modules and reshapes licensing continuously. Every product name, SKU, retirement date and certification code in this document should be re-verified against the Microsoft 365 admin center Message center, the Microsoft 365 Roadmap, and learn.microsoft.com before being adopted into an SOP. Items most prone to change are flagged with ⚠️.
Executive Summary
1. What KTBR means for an M365 Administrator
Keep The Business Running is everything that preserves the availability, integrity, security and compliance of the tenant as it exists today. It is defined by an obligation, not a project plan: if it stops, people cannot work, or the organisation becomes exposed.
For M365 specifically, KTBR is unusual compared with classic infrastructure ops because Microsoft owns the uptime, and you own the configuration. You are not patching Exchange servers; you are:
- Governing identity — the true perimeter of the tenant.
- Governing configuration drift across ~10 admin portals with thousands of independently mutable settings.
- Governing delegation — who can change what, and how that change is proven after the fact.
- Governing data boundaries — sharing, retention, residency, discoverability.
- Consuming a change stream you do not control (Message center + Roadmap) and deciding what breaks you.
- Owning recoverability, because Microsoft's SLA covers service availability, not your data or your misconfiguration.
The single most common senior-level failure is treating KTBR as ticket response. It is not. Roughly 70% of mature KTBR is preventive and evidentiary: monitoring, drift detection, expiry tracking, access review, audit retention and documentation.
2. What KTBE means for an M365 Administrator
Keep The Business Expanding is everything that increases the value density of the licences already being paid for, and reduces the human cost of running the estate.
M365 is chronically under-consumed. A typical enterprise pays for E5 and uses roughly E3 plus Teams. KTBE is the discipline of closing that gap deliberately:
- Capability activation — turning on paid-for capability (Purview DLP, Insider Risk, Entra ID Governance, Intune Suite, Defender features, SharePoint Advanced Management).
- Automation — removing the administrator from repetitive workflows via Graph, PowerShell, Power Automate, Azure Automation, Logic Apps.
- Self-service — pushing safe operations to owners, managers and the service desk with guardrails.
- AI enablement — Copilot readiness, Copilot Studio agents, agent governance.
- Cost engineering — licence right-sizing, SKU rationalisation, storage/archive tiering, credit consumption control.
- Modernisation — classic → modern SharePoint, hybrid → cloud-only identity, GPO → Intune, legacy auth → modern auth, on-prem file shares → OneDrive/SPO.
- Measurement — dashboards and executive reporting that convert admin activity into business language.
KTBE is what turns an administrator into an architect, and an operational cost centre into a platform function.
3. Difference between KTBR and KTBE
| Dimension | KTBR | KTBE |
|---|
| Purpose | Protect current state | Improve future state |
| Trigger | Event, schedule, obligation | Opportunity, strategy, backlog |
| Failure mode | Outage, breach, audit finding, data loss | Stagnation, waste, shadow IT, low adoption |
| Time horizon | Minutes → weeks | Weeks → quarters |
| Success measure | Nothing bad happened; you can prove it | Something measurably improved |
| Visibility to business | Only when it fails | Only when it is marketed |
| Budget language | Risk, compliance, uptime | ROI, capacity, productivity, cost avoidance |
| Skill emphasis | Depth, precision, discipline | Breadth, design, influence |
| Sourcing | Cannot be deferred | Can always be deferred (which is the trap) |
| Career effect | Keeps you employed | Gets you promoted |
The critical relationship: KTBE work is what permanently reduces KTBR volume. Automating joiner-mover-leaver is a KTBE project that removes a recurring KTBR burden forever. An organisation that never funds KTBE will see KTBR consume 100% of capacity within about 18 months of tenant growth.
4. How a senior M365 administrator should balance both
Five practical rules:
- Protect KTBE time structurally, not by willpower. Block it in the calendar as immovable. Unprotected KTBE time is always consumed by KTBR.
- Automate the KTBR that recurs. The rule: any manual task performed more than twice a month, or by more than one person inconsistently, becomes an automation backlog item.
- Run KTBR on a schedule, not on interrupt. Daily/weekly/monthly checklists batch the work. Interrupt-driven ops destroy deep-work capacity for KTBE.
- Convert every incident into a backlog item. Post-incident, ask "what preventive control or automation would have caught this?" This is the healthiest feeder for the KTBE queue.
- Report both. Executives fund what they can see. Report KTBR as risk reduction and stability; report KTBE as cost avoidance and productivity.
5. Recommended percentage of time between KTBR and KTBE
| Tenant maturity | KTBR | KTBE | Notes |
|---|
| Immature / firefighting (no automation, unknown drift, poor documentation) | 80% | 20% | Do not attempt Copilot rollout here. Stabilise first. Target: exit this state in 90 days. |
| Stabilising (monitoring in place, core automation started) | 65% | 35% | The most common real-world state. |
| Target steady state for a senior admin | 50–55% | 45–50% | Achievable once JML, reporting, licensing and drift detection are automated. |
| Mature platform team (KTBR largely automated + tiered to service desk) | 35% | 65% | Admin becomes platform owner/architect; L1 handles routine. |
Recommended target: 55% KTBR / 45% KTBE, with a hard floor of 25% KTBE. Below 25% KTBE, the tenant is decaying — capability goes unused, technical debt accrues, and the KTBR load ratchets upward.
A useful weekly shape for a single senior admin (40h):
| Block | Hours | Content |
|---|
| Daily operational checks | 5 | 45–60 min/day: service health, security, identity, mail flow, automation health |
| Incident / escalation | 8 | Reactive capacity, buffered |
| Scheduled KTBR (weekly/monthly cycles) | 6 | Reviews, reporting, patch/config cycles |
| Change & release | 3 | Message center triage, CAB, pilot rings |
| KTBE build | 12 | Automation, projects, modernisation |
| Learning + documentation | 4 | Non-negotiable, or the model collapses |
| Stakeholder engagement | 2 | Business owners, security, service desk |
Reading Guide for the Catalogs
The catalogs in Sections A and B use the compact nine-column schema requested. Because a full eighteen-field expansion for ~350 tasks would be unusable as a working document, the full field set (description, why it matters, business impact, tools, admin center, automation, monitoring, documentation, evidence, failure scenario, preventive action, escalation criteria) is expanded in Section A.27 — Deep-Dive Task Cards for the tasks where that detail actually changes behaviour: the high-risk, outage-causing and exposure-creating tasks. Treat the catalog as the index and the cards as the SOP source.
Legend
- Frequency: RT = Real-time/continuous, D = Daily, W = Weekly, M = Monthly, Q = Quarterly, Y = Yearly, AH = Ad-hoc, EV = Event-driven
- Priority: P1 Critical · P2 Major · P3 Normal · P4 Low
- Skill: B = Beginner, I = Intermediate, A = Advanced, E = Expert
- Automation: ★★★ fully automatable · ★★ substantially automatable · ★ partially / assisted · ☆ judgement-bound, keep human
- Flags: 🔥 outage risk if neglected · 🔒 security/compliance exposure · 👻 frequently forgotten · ⚠️ product/licensing detail likely to change
A. KTBR Master Task Catalog
A.1 Identity & Microsoft Entra ID
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| IDN-01 | Joiner provisioning (account, licence, groups, mailbox, manager, attributes) | Entra ID | Lifecycle | EV/D | P2 | I | ★★★ Graph + Entra Lifecycle Workflows / HR-driven inbound provisioning | New starters productive day one; avoids manual error |
| IDN-02 | Mover / role-change re-permissioning and stale access removal 👻🔒 | Entra ID | Lifecycle | EV | P2 | A | ★★ Lifecycle Workflows + dynamic groups + access packages | Prevents access accumulation, the #1 audit finding |
| IDN-03 | Leaver deprovisioning (disable, revoke sessions, block sign-in, convert mailbox, reassign data, remove licences) 🔒 | Entra ID / EXO / OD | Lifecycle | EV | P1 | A | ★★★ Graph + PowerShell runbook | Terminates access within minutes; legal/HR obligation |
| IDN-04 | Break-glass / emergency access account validation 🔥🔒👻 | Entra ID | Resilience | Q | P1 | A | ★★ Sign-in log alert on use; scripted credential test | Prevents total tenant lockout during MFA/CA/federation failure |
| IDN-05 | Conditional Access policy review, drift check and change control 🔥🔒 | Entra ID | Access control | W/M | P1 | E | ★★ Graph export + Git diff + What-If API | A bad CA policy is the fastest way to lock out an entire company |
| IDN-06 | Conditional Access exclusion audit (who is exempt, and why) 👻🔒 | Entra ID | Access control | M | P1 | A | ★★★ Graph report on excluded users/groups | Exclusions silently become permanent backdoors |
| IDN-07 | MFA / phishing-resistant auth coverage reporting and gap closure 🔒 | Entra ID | Authentication | W | P1 | I | ★★★ Graph authentication methods report | Credential attacks are the dominant breach vector |
| IDN-08 | Authentication methods policy migration & legacy method retirement ⚠️🔒 | Entra ID | Authentication | Q | P2 | A | ★★ Graph policy read + registration campaign | Retires SMS/voice; moves to passkeys/Authenticator number matching |
| IDN-09 | Risky user and risky sign-in triage (Identity Protection) 🔒 | Entra ID P2 | Threat | D | P1 | A | ★★ Auto-remediation via CA risk policies + Sentinel/Defender XDR | Detects account takeover before lateral movement |
| IDN-10 | Privileged Identity Management: eligible role review, activation audit, approval hygiene 🔒 | Entra ID P2 | Privilege | W/M | P1 | A | ★★ Graph PIM APIs + alerting on standing assignment | Standing Global Admin is an unacceptable enterprise risk |
| IDN-11 | Global Administrator count and privileged role inventory 🔒👻 | Entra ID | Privilege | M | P1 | I | ★★★ Graph directoryRoles export | Target: ≤5 GAs, all PIM-eligible, all phishing-resistant MFA |
| IDN-12 | Access reviews for privileged roles, guests and high-value groups | Entra ID Governance ⚠️ | Governance | M/Q | P2 | A | ★★★ Entra Access Reviews scheduled campaigns | Auditable proof of least privilege |
| IDN-13 | Guest / B2B lifecycle: inventory, inactivity, sponsor validation, removal 🔒👻 | Entra ID | External | M | P2 | I | ★★★ Graph guest report + lifecycle workflow | Dormant guests retain access to live data indefinitely |
| IDN-14 | Cross-tenant access settings (inbound/outbound B2B, trust MFA/device claims) review 🔒 | Entra ID | External | Q | P2 | A | ★★ Graph crossTenantAccessPolicy export | Controls which partner tenants can reach your data |
| IDN-15 | Entra Connect / Cloud Sync health, sync errors, sync account credential rotation 🔥👻 | Entra ID | Hybrid | D/M | P1 | A | ★★ Health alerts + Graph sync errors + agent version check | Sync failure silently freezes all identity changes |
| IDN-16 | Duplicate / orphaned / soft-deleted object cleanup | Entra ID | Hygiene | M | P3 | I | ★★★ Graph deletedItems + reconciliation script | Prevents licence waste and mail routing confusion |
| IDN-17 | Named location, trusted IP and country block-list maintenance | Entra ID | Access control | Q | P2 | I | ★★ Graph namedLocations | Stale IP ranges create both lockouts and gaps |
| IDN-18 | Self-service password reset / writeback health and registration coverage | Entra ID | Authentication | M | P2 | I | ★★★ Registration report | Deflects the highest-volume service desk ticket type |
| IDN-19 | Group sprawl control: naming policy, expiration policy, ownerless group remediation 👻 | Entra ID / M365 Groups | Governance | M | P2 | I | ★★★ Ownerless group policy + Graph report | Ownerless Teams/Groups become ungoverned data stores |
| IDN-20 | Administrative Unit and delegated-admin scope validation | Entra ID | Delegation | Q | P3 | A | ★★ Graph AU membership export | Keeps regional/BU delegation from over-scoping |
| IDN-21 | Tenant-wide sign-in failure trend analysis (legacy auth, blocked clients) | Entra ID | Monitoring | W | P2 | I | ★★★ Log Analytics / Sentinel workbook | Surfaces broken apps before users report them |
| IDN-22 | On-premises AD → cloud authentication method review (PHS/PTA/Federation) 🔥 | Entra ID | Architecture | Q/Y | P1 | E | ★ Documented design review | Federation outages take down all M365 auth |
A.2 Exchange Online
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| EXO-01 | Mail flow monitoring: queues, deferrals, NDR spikes, connector health 🔥 | Exchange Online | Operations | D | P1 | I | ★★★ Mail flow reports via Graph reports / EXO PowerShell + alerting | Mail is the most visible outage in any organisation |
| EXO-02 | Inbound/outbound connector configuration review and certificate validity 🔥🔒👻 | Exchange Online | Configuration | Q | P1 | A | ★★ Get-OutboundConnector export + expiry alert | Expired connector cert = total mail flow failure |
| EXO-03 | SPF, DKIM, DMARC record validation and DMARC enforcement progression 🔒👻 | Exchange Online | Email security | M | P1 | A | ★★★ DNS check script + DMARC aggregate report parsing | Prevents domain spoofing and inbound rejection by partners |
| EXO-04 | DKIM key rotation and CNAME validity ⚠️👻 | Exchange Online | Email security | Q | P2 | I | ★★★ Rotate-DkimSigningConfig scheduled | Stale keys weaken authentication posture |
| EXO-05 | Accepted domain, domain verification and MX record review 🔥 | Exchange Online | Configuration | Q | P1 | I | ★★ Automated DNS diff | Domain expiry/misconfiguration takes down mail |
| EXO-06 | Mailbox size, archive quota and retention-driven growth monitoring | Exchange Online | Capacity | W/M | P2 | I | ★★★ Get-MailboxStatistics scheduled report | Prevents mailbox-full outages for heavy users |
| EXO-07 | Auto-expanding archive enablement and health | Exchange Online | Capacity | M | P3 | I | ★★★ PowerShell | Avoids archive quota hard stops |
| EXO-08 | Shared mailbox inventory, permission audit and licensing check 👻🔒 | Exchange Online | Governance | M | P2 | I | ★★★ Graph/EXO permission export | Shared mailboxes >50GB require a licence ⚠️; permissions drift badly |
| EXO-09 | Full Access / Send As / Send on Behalf delegation audit 🔒 | Exchange Online | Governance | Q | P2 | I | ★★★ Get-MailboxPermission recursion | Undocumented delegation is a common data-exfiltration path |
| EXO-10 | Distribution list / M365 Group ownership, membership and dormancy review 👻 | Exchange Online | Governance | Q | P3 | I | ★★★ PowerShell inventory | Ownerless DLs cannot be maintained; dormant DLs leak |
| EXO-11 | Transport rule (mail flow rule) inventory, conflict and precedence review 🔒👻 | Exchange Online | Configuration | Q | P2 | A | ★★ Export + review + Git versioning | Rules silently intercept, redirect or bypass filtering |
| EXO-12 | Mailbox forwarding audit — external forwarding and inbox rules 🔒 | Exchange Online | Threat | W | P1 | I | ★★★ PowerShell + Defender alert policy | Classic BEC persistence mechanism |
| EXO-13 | Anti-spam / anti-phishing / anti-malware policy review (Standard/Strict preset drift) 🔒 | Defender for Office 365 | Email security | M | P1 | A | ★★ Configuration analyzer + export | Policy drift reduces protection silently |
| EXO-14 | Quarantine review and end-user release policy validation | Defender for Office 365 | Operations | D/W | P2 | I | ★★ Quarantine notification config + report | Legitimate mail trapped = business impact |
| EXO-15 | Message trace investigations for delivery disputes | Exchange Online | Support | AH | P3 | I | ★★ Self-service message trace tool for service desk | Deflects escalations |
| EXO-16 | Litigation hold / In-Place hold state validation on relevant mailboxes 🔒 | Exchange Online / Purview | Compliance | M | P2 | A | ★★★ PowerShell hold report | Failure to preserve = legal sanction |
| EXO-17 | Inactive mailbox management (post-termination retention) 👻🔒 | Exchange Online | Compliance | M | P2 | A | ★★ PowerShell inventory | Deleting a held mailbox destroys evidence |
| EXO-18 | SMTP AUTH / high-volume relay usage review and modernisation ⚠️🔥🔒 | Exchange Online | Legacy | Q | P1 | A | ★★★ Report SMTP AUTH sign-ins from Entra logs | Basic auth for SMTP is retiring ⚠️; unmigrated apps break |
| EXO-19 | Booking / resource mailbox calendar processing configuration | Exchange Online | Operations | Q | P3 | I | ★★★ Set-CalendarProcessing baseline script | Room booking failures are high-friction, high-visibility |
| EXO-20 | Exchange Online RBAC role group membership audit 🔒 | Exchange Online | Privilege | Q | P2 | A | ★★★ Get-RoleGroupMember export | Org Management is effectively mail-wide admin |
| EXO-21 | Outbound spam policy and restricted-sender remediation | Defender for Office 365 | Operations | D | P2 | I | ★★ Alert on restricted senders | Compromised account blocked = user cannot send mail |
| EXO-22 | Journaling / archiving integration health (where third-party archive exists) 👻 | Exchange Online | Compliance | M | P2 | A | ★★ NDR monitoring on journal recipient | Journal failure creates a silent compliance gap |
| EXO-23 | Hybrid Exchange server patching and eventual decommission ⚠️🔥🔒 | Exchange Hybrid | Hybrid | M/Q | P1 | E | ★ Manual, tracked | Unpatched hybrid Exchange servers are a top breach vector |
A.3 Microsoft Teams
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| TMS-01 | Teams service health and call quality monitoring (CQD) 🔥 | Teams | Operations | D/W | P2 | A | ★★ CQD + Power BI template + alerting | Poor meeting quality is the loudest user complaint |
| TMS-02 | Teams creation governance: template, naming, classification, approval 👻 | Teams | Governance | AH/M | P2 | A | ★★★ Power Automate request flow + Graph provisioning | Uncontrolled team sprawl destroys findability and governance |
| TMS-03 | Teams/Group expiration and dormancy review | Teams | Governance | Q | P3 | I | ★★★ Group expiration policy + usage report | Reclaims storage and reduces attack surface |
| TMS-04 | External access (federation) and guest access policy review 🔒 | Teams | External | Q | P2 | A | ★★ Get-CsTenantFederationConfiguration export | Controls who outside the org can reach staff |
| TMS-05 | Shared channel (B2B direct connect) governance 🔒👻 | Teams | External | Q | P2 | A | ★★ Cross-tenant access policy + Graph report | Shared channels bypass classic guest controls |
| TMS-06 | Teams app permission policies, app catalogue and third-party app approval 🔒 | Teams | Governance | M | P2 | A | ★★ Graph app policy export | Unvetted apps get broad data access |
| TMS-07 | Meeting policy review (recording, transcription, lobby, anonymous join) 🔒 | Teams | Governance | Q | P2 | I | ★★ PowerShell policy export | Anonymous join + auto-admit is a meeting-hijack risk |
| TMS-08 | Recording and transcript retention alignment with Purview ⚠️🔒👻 | Teams / Purview | Compliance | Q | P2 | A | ★★ Retention policy validation | Recordings accumulate in OneDrive indefinitely by default |
| TMS-09 | Teams Phone: number inventory, emergency address validation, calling plan usage 🔥🔒 | Teams Phone | Voice | M/Q | P1 | A | ★★ PowerShell number/E911 report | Incorrect emergency address is a life-safety and legal issue |
| TMS-10 | Auto attendant / call queue configuration and holiday schedule maintenance 👻 | Teams Phone | Voice | Q | P2 | I | ★★ Scripted holiday set updates | Missed holiday updates route customers to nobody |
| TMS-11 | Teams Rooms / device health, firmware and sign-in account expiry 🔥👻 | Teams Rooms | Devices | D/W | P2 | I | ★★ Teams Rooms Pro portal + alerting ⚠️ | Room device account password expiry kills all room meetings |
| TMS-12 | Direct Routing / SBC certificate and trunk health 🔥👻 | Teams Phone | Voice | M | P1 | E | ★★ Certificate expiry monitoring | Expired SBC cert = total voice outage |
| TMS-13 | Teams client version and new-client rollout compliance ⚠️ | Teams | Client | Q | P3 | I | ★★ Intune reporting | Old clients lose features and support |
| TMS-14 | Teams usage and adoption reporting | Teams | Reporting | M | P3 | I | ★★★ Graph reports API → Power BI | Evidence for licence and adoption decisions |
| TMS-15 | Teams live event / town hall / webinar configuration and capacity ⚠️ | Teams Premium | Events | AH | P3 | I | ★ Manual with runbook | Failed all-hands is a highly visible incident |
A.4 SharePoint Online
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| SPO-01 | Tenant storage consumption and site quota monitoring 🔥 | SharePoint Online | Capacity | W/M | P1 | I | ★★★ PnP/Graph storage report + threshold alert | Hitting tenant quota blocks uploads org-wide |
| SPO-02 | External sharing policy review (tenant + site level) 🔒 | SharePoint Online | External | M/Q | P1 | A | ★★★ Get-SPOTenant + site-level export | The single largest oversharing risk in M365 |
| SPO-03 | Anonymous ("Anyone") link inventory, expiry and revocation 🔒👻 | SharePoint Online | External | M | P1 | A | ★★★ Sharing report / Data Access Governance reports ⚠️ | Anonymous links are indexable, forwardable and permanent by default |
| SPO-04 | Oversharing and site permission sprawl review (broken inheritance, EEEU) 🔒 | SharePoint Online | Permissions | M | P2 | A | ★★ Data Access Governance + SharePoint Advanced Management ⚠️ | "Everyone Except External Users" on a site is a Copilot-era data leak |
| SPO-05 | Site collection admin and ownership validation (ownerless sites) 👻 | SharePoint Online | Governance | Q | P2 | I | ★★★ PnP inventory + ownership attestation flow | No owner = no lifecycle, no access decisions |
| SPO-06 | Site lifecycle: inactive site policy, archival, deletion 👻 | SharePoint Online | Lifecycle | Q | P3 | A | ★★ Site lifecycle management / M365 Archive ⚠️ | Reduces cost, risk and Copilot noise |
| SPO-07 | Recycle bin, versioning settings and version history storage impact ⚠️👻 | SharePoint Online | Capacity | Q | P3 | I | ★★★ Version trim policy via PowerShell | Version bloat consumes large amounts of paid storage |
| SPO-08 | Custom script / DenyAddAndCustomizePages posture 🔒 | SharePoint Online | Security | Q | P2 | A | ★★★ Set-SPOSite scripted enforcement | Custom script re-enablement is a privilege-escalation path |
| SPO-09 | Sensitivity label application on sites/containers and label drift ⚠️🔒 | SPO / Purview | Compliance | M | P2 | A | ★★ Graph + Purview reporting | Container labels enforce sharing/device conditions |
| SPO-10 | Search schema, promoted results and search health | SharePoint Online | Search | Q | P3 | A | ★ Manual + query testing | Bad search directly degrades Copilot grounding quality |
| SPO-11 | Restricted SharePoint Search / Restricted Content Discovery review ⚠️🔒 | SPO Advanced Management | Copilot readiness | M/Q | P2 | A | ★★ PowerShell configuration report | Prevents Copilot surfacing sensitive content pre-remediation |
| SPO-12 | Hub site structure and information architecture review | SharePoint Online | Architecture | Q | P3 | A | ★ Design review | Poor IA is the root cause of "we can't find anything" |
| SPO-13 | Classic site / legacy feature remediation ⚠️👻 | SharePoint Online | Modernisation | Q | P3 | A | ★★ Modernisation scanner | Classic features are being retired; late migration causes breakage |
| SPO-14 | Term store / managed metadata governance | SharePoint Online | Governance | Q | P4 | I | ★★ PnP export | Metadata quality drives search and retention accuracy |
| SPO-15 | SharePoint app catalogue / SPFx solution inventory and approval 🔒 | SharePoint Online | Governance | Q | P3 | A | ★★ PnP inventory | Custom solutions run with user context and can exfiltrate |
| SPO-16 | Site provisioning standardisation via template/policy | SharePoint Online | Governance | AH | P3 | A | ★★★ PnP provisioning + Power Automate | Consistency at creation eliminates remediation later |
A.5 OneDrive for Business
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| ODB-01 | OneDrive storage quota monitoring and per-user overrides | OneDrive | Capacity | M | P2 | I | ★★★ Graph/SPO PowerShell report | Full OneDrive stops sync and desktop file save |
| ODB-02 | Leaver OneDrive retention, delegated access and disposal 🔒👻 | OneDrive | Lifecycle | EV/M | P2 | A | ★★★ Runbook: grant manager access, apply retention, schedule deletion | Departing-employee data is both a risk and an asset |
| ODB-03 | Known Folder Move (Desktop/Documents/Pictures) enrolment coverage 👻 | OneDrive | Resilience | M | P2 | I | ★★★ Intune policy + reporting | Unredirected local files are unrecoverable on device loss |
| ODB-04 | Sync client health, version compliance and sync error reporting | OneDrive | Client | W/M | P3 | I | ★★ OneDrive sync health reports ⚠️ + Intune | Silent sync failure = silent data loss |
| ODB-05 | OneDrive external sharing exposure review 🔒 | OneDrive | External | M | P2 | A | ★★★ Sharing report | Personal-drive oversharing is less visible than SPO |
| ODB-06 | Files On-Demand and storage-sense policy enforcement | OneDrive | Client | Q | P4 | I | ★★★ Intune ADMX policy | Prevents endpoint disk exhaustion |
| ODB-07 | Retention and hold coverage for OneDrive accounts 🔒 | OneDrive / Purview | Compliance | Q | P2 | A | ★★ Purview policy validation | Retention gaps break eDiscovery defensibility |
| ODB-08 | Ransomware / mass-deletion detection and Files Restore readiness 🔥🔒👻 | OneDrive | Recovery | Q | P1 | A | ★★ Alert policy on mass deletion + restore drill | Files Restore is the primary user-level recovery path |
A.6 Microsoft Intune / Endpoint
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| INT-01 | Device compliance monitoring and non-compliant device remediation 🔒 | Intune | Compliance | D/W | P1 | I | ★★★ Graph deviceManagement reports + Power Automate notification | Non-compliant devices bypass Zero Trust assumptions |
| INT-02 | Configuration profile deployment health and conflict resolution 👻 | Intune | Configuration | W | P2 | A | ★★ Graph policy status export | Conflicting profiles silently fail to apply |
| INT-03 | App deployment success/failure monitoring and repackaging | Intune | Applications | W | P2 | I | ★★ Graph app install status | Failed deployments produce ticket floods |
| INT-04 | Windows Update rings / Autopatch health and deferral management 🔥🔒 | Intune / Autopatch | Patching | W/M | P1 | A | ★★★ Update report + exception workflow | Unpatched endpoints are the primary ransomware entry |
| INT-05 | Autopilot device registration, profile assignment and ESP validation 🔥 | Intune | Provisioning | W | P2 | A | ★★ Graph Autopilot device inventory | Failed enrolment blocks new-hire onboarding |
| INT-06 | BitLocker/FileVault encryption coverage and key escrow verification 🔒👻 | Intune | Encryption | M | P1 | I | ★★★ Graph encryption report | Missing escrowed key = unrecoverable device |
| INT-07 | MDM/APNs certificate, VPP token and DEP token expiry tracking 🔥🔒👻 | Intune | Certificates | M | P1 | I | ★★★ Graph expiry query + calendar alert 60/30/7 days | APNs expiry disconnects every managed Apple device |
| INT-08 | App protection policy (MAM) coverage for BYOD 🔒 | Intune | Data protection | Q | P2 | A | ★★ Graph policy report | Protects corporate data on unmanaged devices |
| INT-09 | Stale / duplicate device object cleanup in Intune and Entra 👻 | Intune / Entra | Hygiene | M | P3 | I | ★★★ Cleanup rules + Graph script | Stale objects corrupt compliance reporting and licence counts |
| INT-10 | Security baseline assignment and drift review 🔒 | Intune | Hardening | Q | P2 | A | ★★ Baseline comparison export | Baselines regress after Microsoft version updates |
| INT-11 | Defender for Endpoint ↔ Intune connector and onboarding coverage 🔒 | Intune / MDE | Integration | M | P1 | A | ★★★ Onboarding gap report | Unonboarded devices are invisible to the SOC |
| INT-12 | Endpoint analytics: startup performance, application reliability | Intune | Experience | M | P3 | I | ★★ Endpoint analytics + Power BI | Directly measurable employee productivity signal |
| INT-13 | Enrolment restriction, device limit and ownership-type policy review | Intune | Governance | Q | P3 | I | ★★ Graph export | Prevents unmanaged personal devices entering the estate |
| INT-14 | macOS / iOS / Android platform-specific policy parity review 👻 | Intune | Coverage | Q | P3 | A | ★★ Cross-platform policy matrix | Non-Windows platforms are chronically under-governed |
| INT-15 | Scripts and remediation (proactive remediation) health ⚠️ | Intune | Automation | M | P3 | A | ★★★ Remediation script reporting | Detect-and-fix at scale without user impact |
| INT-16 | Windows feature update readiness and end-of-support tracking 🔥⚠️ | Intune | Lifecycle | Q | P1 | A | ★★ Update Compliance / Autopatch reports | Out-of-support OS loses security updates |
| INT-17 | Certificate infrastructure (SCEP/PKCS/Cloud PKI) health ⚠️🔥 | Intune | Certificates | M | P1 | E | ★★ Connector health monitoring | Certificate failure breaks Wi-Fi/VPN at scale |
A.7 Microsoft Defender (XDR)
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| DEF-01 | Defender XDR incident triage, classification and closure 🔒 | Defender XDR | Threat | RT/D | P1 | A | ★★ Automated attack disruption + Sentinel playbooks | Detection without triage is worthless |
| DEF-02 | High-severity alert response SLA tracking 🔒 | Defender XDR | Threat | D | P1 | A | ★★★ Graph security API → ticketing integration | Proves SOC effectiveness to auditors |
| DEF-03 | Phishing / user-reported message investigation and tenant-wide purge 🔒 | Defender for Office 365 | Threat | D | P1 | A | ★★ Automated Investigation & Response (AIR) + ZAP | Contains campaigns before click-through spreads |
| DEF-04 | Safe Links / Safe Attachments policy coverage and bypass audit 🔒👻 | Defender for Office 365 | Email security | M | P1 | A | ★★ Configuration analyzer | Bypass rules and allow-lists erode protection quietly |
| DEF-05 | Tenant Allow/Block List hygiene and expiry review 🔒👻 | Defender for Office 365 | Email security | M | P2 | I | ★★★ PowerShell export + expiry enforcement | Permanent allow entries are exploited by attackers |
| DEF-06 | Microsoft Secure Score review and improvement action backlog 🔒 | Defender XDR | Posture | M | P2 | A | ★★★ Graph secureScore API → dashboard | The most usable executive security metric in M365 |
| DEF-07 | Vulnerability management (Defender Vulnerability Management) exposure review ⚠️🔒 | MDVM | Posture | W | P1 | A | ★★ Export to ticketing + Intune remediation | Converts CVEs into actionable endpoint tasks |
| DEF-08 | Attack surface reduction (ASR) rule coverage and audit→block progression 🔒👻 | Defender for Endpoint | Hardening | Q | P2 | A | ★★ Intune policy + report | ASR rules left in audit mode provide zero protection |
| DEF-09 | Defender for Cloud Apps: OAuth app risk, shadow IT, session policies 🔒👻 | Defender for Cloud Apps | Threat | M | P2 | A | ★★ App governance policies | Malicious consented OAuth apps persist through password resets |
| DEF-10 | Defender for Identity sensor health and on-prem AD posture alerts 🔒 | Defender for Identity | Threat | W | P2 | A | ★★ Sensor health monitoring | Hybrid attack paths start on-prem |
| DEF-11 | Threat analytics review and applicability assessment 👻 | Defender XDR | Intelligence | W | P3 | A | ★ Human review of Microsoft threat reports | Proactive defence against active campaigns |
| DEF-12 | Advanced hunting queries: scheduled detections and custom rules | Defender XDR | Detection | M | P2 | E | ★★★ KQL custom detection rules | Fills gaps in out-of-box detection |
| DEF-13 | Automated investigation & response (AIR) automation level review | Defender XDR | Automation | Q | P2 | A | ★★ Configuration review | Full automation dramatically reduces MTTR |
| DEF-14 | Exclusion review for antivirus/EDR 🔒👻 | Defender for Endpoint | Hardening | Q | P2 | A | ★★ Export and justify each exclusion | Exclusions are actively hunted by attackers |
| DEF-15 | Attack simulation training campaign execution and reporting | Defender for Office 365 | Awareness | Q | P3 | I | ★★ Scheduled campaigns + Power BI | Measurable human-risk reduction |
| DEF-16 | Email authentication failure and impersonation-protection tuning 🔒 | Defender for Office 365 | Email security | M | P2 | A | ★★ Priority account protection config | Executive impersonation is the top BEC vector |
A.8 Microsoft Purview
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| PUR-01 | DLP policy alert triage, false-positive tuning and rule refinement 🔒 | Purview DLP | Data protection | D/W | P1 | A | ★★ Graph/Activity Explorer export + Power BI trend | Untuned DLP is ignored DLP |
| PUR-02 | DLP policy coverage across Exchange, SPO, OneDrive, Teams, Endpoint, and AI apps ⚠️🔒👻 | Purview DLP | Data protection | Q | P1 | A | ★★ Policy matrix export | Endpoint and Copilot/AI surfaces are the common gaps |
| PUR-03 | Sensitivity label taxonomy, publishing and scoping review 🔒 | Purview Information Protection | Data protection | Q | P2 | A | ★★ Label policy export | Too many labels = zero adoption |
| PUR-04 | Auto-labelling policy simulation and rollout | Purview | Data protection | Q | P2 | A | ★★★ Simulation mode → enforce | Scales classification beyond user discipline |
| PUR-05 | Retention policy and label coverage validation across workloads 🔒 | Purview Data Lifecycle Mgmt | Compliance | Q | P1 | A | ★★ Policy inventory export | Both over-retention and under-retention create legal risk |
| PUR-06 | Disposition review queue processing 👻 | Purview | Compliance | M | P3 | I | ★★ Reminder automation | Unprocessed disposition = indefinite retention |
| PUR-07 | eDiscovery (Premium) case management, hold tracking and export ⚠️🔒 | Purview eDiscovery | Legal | AH/M | P1 | A | ★★ Graph eDiscovery API | Legal obligation; spoliation risk if mishandled |
| PUR-08 | Unified audit log ingestion, retention tier and search validation ⚠️🔒👻 | Purview Audit | Compliance | M | P1 | A | ★★★ Office 365 Management Activity API → SIEM | Default retention may be insufficient for your regulator |
| PUR-09 | Audit log search for privileged and sensitive operations 🔒 | Purview Audit | Compliance | W | P2 | A | ★★★ Scheduled KQL/Graph queries | Evidence for investigations and audits |
| PUR-10 | Insider Risk Management policy tuning and case review ⚠️🔒 | Purview IRM | Risk | W/M | P2 | E | ★★ Policy + HR connector integration | Detects data theft around resignations |
| PUR-11 | Communication Compliance policy review (regulated industries) ⚠️🔒 | Purview | Compliance | M | P2 | A | ★★ Policy + reviewer workflow | Regulatory requirement in finance/healthcare |
| PUR-12 | Compliance Manager assessment progress and control evidence 👻 | Purview Compliance Manager | Compliance | Q | P2 | A | ★★ Export improvement actions | Maps M365 config to ISO/NIST/GDPR/DPDP obligations |
| PUR-13 | Data Map / data classification scanning of connected sources ⚠️ | Purview | Discovery | Q | P3 | E | ★★ Scan scheduling | Extends classification beyond M365 |
| PUR-14 | Trainable classifier accuracy validation 👻 | Purview | Data protection | Q | P3 | E | ★ Sample-based validation | Poor classifiers cause both leaks and false positives |
| PUR-15 | Information barriers configuration and validation ⚠️🔒 | Purview | Compliance | Q | P2 | E | ★★ PowerShell segment export | Required for ethical walls in regulated firms |
| PUR-16 | Data residency / multi-geo configuration validation ⚠️🔒👻 | Purview / M365 Multi-Geo | Compliance | Q | P2 | E | ★★ PowerShell PDL report | Wrong preferred data location = sovereignty breach |
| PUR-17 | Data Security Posture Management for AI / Copilot data risk review ⚠️🔒 | Purview | AI governance | M | P2 | A | ★★ DSPM for AI reports | Shows what Copilot can actually reach |
A.9 Microsoft 365 Admin (Tenant)
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| M365-01 | Service health dashboard review and user communication 🔥 | M365 Admin Center | Operations | D | P1 | B | ★★★ Service Health Graph API → Teams channel/status page | Proactive comms cut ticket volume during outages |
| M365-02 | Message center triage, impact assessment and action assignment 🔥👻 | M365 Admin Center | Change | D/W | P1 | A | ★★★ Graph serviceAnnouncement API → Planner/DevOps backlog | The #1 forgotten task; Microsoft changes break tenants |
| M365-03 | Microsoft 365 Roadmap review for upcoming capability planning | M365 Roadmap | Planning | M | P3 | I | ★★ RSS/API ingestion | Feeds the KTBE backlog |
| M365-04 | Tenant-wide configuration baseline export and drift detection 🔥🔒👻 | All | Configuration | W/M | P1 | E | ★★★ Multi-module export to Git + scheduled diff | Detects unauthorised or accidental change |
| M365-05 | Admin role assignment review across all workload portals 🔒 | M365 Admin Center | Privilege | M | P1 | A | ★★★ Graph + workload-specific RBAC exports | Workload admin roles (EXO, SPO, Teams) are often forgotten |
| M365-06 | Domain inventory, verification and DNS record ownership 🔥👻 | M365 Admin Center | Configuration | Q | P1 | I | ★★ DNS monitoring + registrar expiry alerts | Domain expiry is a catastrophic, entirely preventable outage |
| M365-07 | Tenant contact details, technical/security notification addresses 👻 | M365 Admin Center | Configuration | Q | P3 | B | ★ Manual | Microsoft's breach notification goes to these addresses |
| M365-08 | Partner / delegated admin (GDAP) relationship review 🔒👻 | M365 Admin Center | Delegation | Q | P1 | A | ★★ Graph delegated admin relationships | CSP partners have held Global Admin far too often |
| M365-09 | Self-service purchase and trial control ⚠️👻 | M365 Admin Center | Governance | Q | P3 | I | ★★★ MSCommerce PowerShell module | Users buying their own Power BI/Copilot licences bypasses governance |
| M365-10 | Tenant-level privacy, telemetry and "connected experiences" settings | M365 Admin Center | Governance | Y | P3 | A | ★★ Policy export | Regulatory and works-council relevance |
| M365-11 | Support ticket lifecycle with Microsoft (severity, escalation, follow-up) | M365 Admin Center | Support | AH | P2 | I | ★★ Ticket tracker automation | Slow escalation extends outages |
| M365-12 | Multi-tenant organisation / cross-tenant sync health ⚠️👻 | Entra / M365 | Architecture | M | P2 | E | ★★ Sync error reporting | M&A scenarios; sync failures split the GAL |
| M365-13 | Tenant restrictions v2 configuration (preventing external tenant access) 🔒👻 | Entra / Network | Security | Q | P2 | E | ★★ Config validation | Prevents data exfiltration to personal/rogue tenants |
| M365-14 | Network connectivity / bandwidth assessment for M365 endpoints ⚠️👻 | Networking | Performance | Q | P3 | A | ★★ M365 network insights + endpoint IP/URL feed automation | Endpoint list changes break proxy/firewall rules |
A.10 Microsoft 365 Apps (Office)
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| APP-01 | Update channel strategy and version currency reporting ⚠️🔥 | M365 Apps | Client | M | P2 | I | ★★★ M365 Apps admin center inventory + Intune | Out-of-support builds lose security fixes |
| APP-02 | Cloud Update / servicing profile ring management ⚠️ | M365 Apps admin center | Client | M | P2 | A | ★★★ Servicing profile | Controlled rollout prevents mass regression |
| APP-03 | Add-in inventory, approval and blocking 🔒👻 | M365 Apps | Governance | Q | P2 | A | ★★ Integrated apps / centralised deployment report | Add-ins read document content |
| APP-04 | Macro / VBA policy and trusted location governance 🔒 | M365 Apps | Security | Q | P2 | A | ★★★ Intune ADMX policy | Macros remain a live malware vector |
| APP-05 | Office policy (Cloud Policy service) baseline and drift ⚠️👻 | M365 Apps admin center | Configuration | Q | P3 | A | ★★ Policy export | Cloud Policy silently supersedes GPO |
| APP-06 | Shared computer activation / licensing activation failures | M365 Apps | Licensing | M | P3 | I | ★★ Activation error report | Users locked out of Office is a P1 for them |
| APP-07 | App compatibility / readiness assessment before channel moves 👻 | M365 Apps | Client | Q | P3 | A | ★★ Readiness toolkit | Prevents breaking finance/engineering macros |
A.11 Microsoft Copilot
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| CPL-01 | Copilot licence assignment, reclamation and utilisation tracking ⚠️ | M365 Copilot | Licensing | M | P2 | I | ★★★ Graph usage reports + auto-reclaim policy | Copilot is expensive; idle seats are pure waste |
| CPL-02 | Copilot readiness: oversharing remediation before enablement 🔒👻 | SPO / Purview | Data governance | Q | P1 | E | ★★ SAM + DSPM for AI + DAG reports ⚠️ | Copilot surfaces every permission mistake ever made |
| CPL-03 | Copilot interaction auditing and eDiscovery coverage ⚠️🔒 | Purview | Compliance | Q | P2 | A | ★★ Audit log validation for Copilot events | Prompts and responses are discoverable records |
| CPL-04 | Copilot Chat / agent usage and data-boundary policy review ⚠️🔒 | M365 Copilot | Governance | M | P2 | A | ★★ Admin center policy export | Controls web grounding and non-licensed usage |
| CPL-05 | Copilot Studio agent inventory, ownership and publishing control ⚠️🔒👻 | Copilot Studio | AI governance | M | P2 | A | ★★ Power Platform admin center + Graph | Ungoverned agents connect to production data |
| CPL-06 | Copilot Studio capacity / credit consumption monitoring ⚠️ | Copilot Studio | Cost | W/M | P2 | A | ★★★ Capacity reports + threshold alerting | Runaway agent consumption creates surprise cost |
| CPL-07 | Agent connection and connector permission review 🔒 | Copilot Studio / Power Platform | AI governance | Q | P2 | A | ★★ Connector inventory | Agents inherit the creator's data access |
| CPL-08 | Copilot adoption and value measurement reporting ⚠️ | M365 Copilot | Reporting | M | P3 | I | ★★★ Copilot usage/dashboard reports → Power BI | Required to defend the renewal |
| CPL-09 | Restricted Content Discovery / sensitivity label enforcement for Copilot ⚠️🔒 | SPO / Purview | Data governance | Q | P1 | E | ★★ Configuration report | Keeps labelled content out of Copilot responses |
| CPL-10 | Copilot prompt-injection and third-party plugin risk review ⚠️🔒👻 | M365 Copilot | AI security | Q | P2 | E | ★ Threat modelling review | Emerging attack class; largely unmonitored today |
A.12 Microsoft Viva
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| VIV-01 | Viva Connections dashboard, card and audience-targeting maintenance ⚠️ | Viva Connections | Experience | M | P3 | I | ★★ PnP provisioning | The employee homepage; stale content kills usage |
| VIV-02 | Viva Engage community governance, ownership and moderation ⚠️🔒 | Viva Engage | Governance | M | P3 | I | ★★ Graph community inventory | Ungoverned communities become compliance liabilities |
| VIV-03 | Viva Insights privacy configuration and de-identification settings 🔒👻 | Viva Insights | Privacy | Q | P2 | A | ★ Config validation | Works councils and GDPR require careful handling |
| VIV-04 | Viva Learning content source connection health ⚠️ | Viva Learning | Content | Q | P4 | I | ★★ Connection check | Broken LMS connections make Learning empty |
| VIV-05 | Viva module licensing and retirement tracking ⚠️👻 | Viva | Lifecycle | Q | P3 | I | ★ Message center tracking | Several Viva modules have been retired or changed ⚠️ — verify current state |
| VIV-06 | Viva Amplify / internal communications campaign hygiene ⚠️ | Viva Amplify | Experience | Q | P4 | I | ★ Manual | Publication sprawl across channels |
A.13 Windows 365 / Cloud PC & AVD
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| W365-01 | Cloud PC provisioning policy, image and network connection health 🔥 | Windows 365 | Operations | W | P1 | A | ★★★ Graph cloudPC APIs + health alerts | Failed Azure network connection blocks all provisioning |
| W365-02 | Cloud PC licence assignment, size right-sizing and reclamation ⚠️ | Windows 365 | Cost | M | P2 | I | ★★★ Utilisation report + auto-reclaim | Cloud PCs are billed per seat regardless of use |
| W365-03 | Custom image lifecycle, patching and re-publish 👻 | Windows 365 | Images | Q | P2 | A | ★★ Image pipeline automation | Stale images extend provisioning and patch gaps |
| W365-04 | Cloud PC connection quality and user experience monitoring | Windows 365 | Experience | W | P3 | A | ★★ Endpoint analytics for Cloud PC | Latency complaints are the main support driver |
| W365-05 | Cloud PC restore point / backup policy validation 🔥👻 | Windows 365 | Recovery | Q | P2 | A | ★★ Policy export | Restore points are the only recovery path |
| W365-06 | Frontline / shared Cloud PC concurrency monitoring ⚠️ | Windows 365 Frontline | Capacity | M | P3 | A | ★★ Usage reporting | Over-subscription locks shift workers out |
| W365-07 | AVD host pool health, session host patching and scaling plans (where used) 🔥 | Azure Virtual Desktop | Operations | W | P2 | E | ★★★ Azure Automation scaling + Azure Monitor | Session host exhaustion = users cannot log in |
A.14 Planner, To Do & Project
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| PLN-01 | Planner plan sprawl, orphaned plan and ownership review ⚠️👻 | Planner | Governance | Q | P4 | I | ★★ Graph planner inventory | Plans outlive their groups and lose owners |
| PLN-02 | Planner Premium / Project licensing and feature usage review ⚠️ | Planner Premium | Licensing | Q | P3 | I | ★★ Usage report | Premium features are often licensed but unused |
| PLN-03 | Retention and eDiscovery coverage for Planner/To Do data ⚠️🔒👻 | Purview | Compliance | Q | P3 | A | ★ Coverage validation | Task data is often outside retention scope |
| PLN-04 | Roadmap/Project for the web integration and permission review ⚠️ | Project | Governance | Q | P4 | I | ★★ Dataverse security role export | Project data lives in Dataverse with separate RBAC |
A.15 Power Platform (overlapping M365 administration)
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| PPF-01 | Environment inventory, strategy and creation control 🔒 | Power Platform | Governance | M | P2 | A | ★★★ CoE Starter Kit + PPAC PowerShell | Uncontrolled environments fragment data and DLP |
| PPF-02 | DLP (connector) policy definition, scoping and drift review 🔒👻 | Power Platform | Data protection | M/Q | P1 | A | ★★★ Get-DlpPolicy export + Git | The primary control preventing data leaving via connectors |
| PPF-03 | Orphaned flow / app ownership reassignment 👻🔥 | Power Automate / Power Apps | Lifecycle | M | P2 | A | ★★★ CoE Starter Kit orphan detection | Business-critical flows die when the owner leaves |
| PPF-04 | Failed flow monitoring and remediation 🔥 | Power Automate | Operations | D | P1 | I | ★★★ CoE / Graph + admin alerting | Silent flow failure breaks business processes invisibly |
| PPF-05 | Connection reference and connector credential expiry 🔥👻 | Power Platform | Operations | M | P1 | A | ★★★ Connection status report | Expired connections are the top cause of flow outage |
| PPF-06 | Managed Environments coverage and premium licence compliance ⚠️ | Power Platform | Governance | Q | P2 | A | ★★ PPAC report | Licensing enforcement and sharing limits |
| PPF-07 | Dataverse capacity, storage and log capacity monitoring 🔥⚠️ | Dataverse | Capacity | M | P1 | A | ★★★ Capacity report + alert | Capacity overage blocks all writes to the environment |
| PPF-08 | Service principal / application user permission review in Dataverse 🔒 | Dataverse | Privilege | Q | P2 | A | ★★ Security role export | App users frequently hold System Administrator |
| PPF-09 | Power Apps sharing and audience scope review 🔒 | Power Apps | Governance | Q | P3 | I | ★★★ CoE sharing report | Apps shared with "Everyone" bypass intended scope |
| PPF-10 | ALM/solution deployment pipeline health | Power Platform | ALM | M | P3 | A | ★★★ Pipelines / Azure DevOps | Unmanaged direct-to-prod editing is unsupportable |
| PPF-11 | Citizen developer catalogue and support-tier definition 👻 | Power Platform | Governance | Q | P3 | A | ★★ CoE inventory + attestation flow | Defines what IT will and will not support |
| PPF-12 | Power BI workspace, gateway and dataset refresh health 🔥 | Power BI / Fabric ⚠️ | Operations | D/W | P2 | A | ★★★ Refresh failure alerting | Broken executive reports damage IT credibility fast |
A.16 Microsoft Graph
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| GRF-01 | Graph application permission inventory and least-privilege reduction 🔒 | Microsoft Graph | Privilege | M/Q | P1 | E | ★★★ Graph servicePrincipal appRoleAssignment export | Mail.ReadWrite/Directory.ReadWrite.All app grants are tenant-wide keys |
| GRF-02 | Admin-consented delegated permission review 🔒👻 | Microsoft Graph | Privilege | Q | P1 | A | ★★★ oauth2PermissionGrants export | Tenant-wide consent affects every user silently |
| GRF-03 | Graph API throttling and error monitoring in automation 🔥 | Microsoft Graph | Operations | W | P2 | A | ★★★ Retry/backoff + telemetry logging | Throttling causes partial, silent automation failure |
| GRF-04 | Deprecated / beta endpoint usage tracking in scripts ⚠️👻 | Microsoft Graph | Change | Q | P2 | A | ★★ Code scan for /beta usage | Beta endpoints change without notice and break jobs |
| GRF-05 | Graph activity logs review (who called what) ⚠️🔒 | Microsoft Graph | Audit | M | P2 | E | ★★★ Log Analytics ingestion + KQL | The only way to see app-level API abuse |
| GRF-06 | Change notification (webhook) subscription lifecycle and renewal 🔥👻 | Microsoft Graph | Operations | W | P2 | A | ★★★ Automated renewal job | Expired subscriptions stop event-driven automation silently |
| GRF-07 | Graph SDK / module version currency across automation estate ⚠️ | Microsoft Graph | Maintenance | Q | P3 | A | ★★ Dependency inventory | Breaking SDK changes take down runbooks |
| GRF-08 | Workload identity federation adoption (removing secrets) 🔒 | Entra / Graph | Security | Q | P2 | E | ★★ Migration tracking | Eliminates the secret-expiry outage class entirely |
A.17 PowerShell & Automation Estate
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| PSH-01 | Retired module migration (MSOnline / AzureAD → Microsoft Graph PowerShell) ⚠️🔥 | PowerShell | Modernisation | Q | P1 | A | ★★ Script inventory + static analysis | Retired modules stop working ⚠️ — verify current retirement status |
| PSH-02 | Script repository, version control and peer review discipline 👻 | PowerShell | Engineering | W | P2 | A | ★★★ Git + PR workflow + PSScriptAnalyzer | Undocumented one-off scripts are unmaintainable risk |
| PSH-03 | Credential and secret handling review (no plaintext, use Key Vault / MI) 🔒 | PowerShell | Security | Q | P1 | A | ★★★ Managed identity migration | Hardcoded credentials in scripts are a classic breach source |
| PSH-04 | Scheduled job / runbook execution monitoring and failure alerting 🔥 | Azure Automation | Operations | D | P1 | A | ★★★ Azure Monitor alert rules | Failed jobs create silent compliance and provisioning gaps |
| PSH-05 | Runbook idempotency and dry-run (-WhatIf) validation 👻 | PowerShell | Engineering | AH | P2 | A | ★★ Test harness | Prevents mass-change accidents |
| PSH-06 | Module version pinning and update testing in automation accounts ⚠️ | Azure Automation | Maintenance | Q | P2 | A | ★★★ Pinned modules + staged update | Auto-updating modules break production runbooks |
| PSH-07 | Bulk-change guardrails (batch size, throttle, rollback plan) 🔥 | PowerShell | Safety | AH | P1 | E | ★★ Framework-level controls | A bad bulk script is the fastest self-inflicted outage |
| PSH-08 | Script execution logging and change attribution 🔒👻 | PowerShell | Audit | M | P2 | A | ★★★ Transcript logging → central store | "Who changed this?" must be answerable |
A.18 Security Operations (cross-workload)
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| SEC-01 | Security incident response runbook execution (account compromise) 🔥🔒 | Cross | IR | EV | P1 | E | ★★★ Playbook: revoke sessions, reset, disable, hunt, purge | Speed of containment determines blast radius |
| SEC-02 | Compromised account forensics (inbox rules, OAuth grants, MFA method changes) 🔒👻 | Entra / EXO | IR | EV | P1 | E | ★★ Automated forensic collection script | Attackers persist through added MFA methods and OAuth apps |
| SEC-03 | App registration secret and certificate expiry tracking 🔥🔒👻 | Entra ID | Certificates | W/M | P1 | I | ★★★ Graph query + 90/60/30/7-day alerts | Expired secrets take down integrations without warning |
| SEC-04 | Enterprise application and service principal inventory review 🔒 | Entra ID | Privilege | M | P1 | A | ★★★ Graph export + risk scoring | Dormant/over-privileged SPNs are prime persistence targets |
| SEC-05 | User consent policy and admin consent workflow enforcement 🔒👻 | Entra ID | Governance | Q | P1 | A | ★★ Policy validation + request queue | Blocks illicit consent grant attacks |
| SEC-06 | Legacy authentication blocking verification 🔒 | Entra ID | Hardening | M | P1 | A | ★★★ Sign-in log query for legacy protocols | Legacy auth bypasses Conditional Access and MFA |
| SEC-07 | Zero Trust maturity assessment against Microsoft model | Cross | Posture | Q/Y | P2 | E | ★ Assessment workshop | Structures multi-year security investment |
| SEC-08 | Security baseline / CIS-M365 benchmark conformance review 👻 | Cross | Posture | Q | P2 | E | ★★★ Automated benchmark script (e.g. config assessment tooling) | Objective, external yardstick for tenant hardening |
| SEC-09 | Threat hunt execution against current threat intelligence 🔒 | Defender XDR | Detection | M | P2 | E | ★★ Saved KQL hunting queries | Finds what detection missed |
| SEC-10 | Tabletop exercise: tenant compromise / Global Admin loss 🔥👻 | Cross | Resilience | Q/Y | P1 | E | ☆ Human exercise | Reveals that the IR plan assumes access you may not have |
| SEC-11 | SIEM/Sentinel connector health and ingestion validation 🔒 | Sentinel | Monitoring | W | P1 | A | ★★★ Ingestion volume anomaly alerts | Silent connector failure blinds the SOC |
| SEC-12 | Privileged access workstation / admin isolation policy review 🔒👻 | Entra / Intune | Hardening | Q | P2 | E | ★★ Compliance reporting | Admin work from a general-purpose laptop is the weak link |
| SEC-13 | Password protection, banned password list and lockout policy 🔒 | Entra ID | Authentication | Q | P3 | I | ★★ Policy export | Cheap, high-yield credential hardening |
| SEC-14 | Token protection / continuous access evaluation configuration ⚠️🔒👻 | Entra ID | Authentication | Q | P2 | E | ★★ CA policy validation | Mitigates token-theft attacks that defeat MFA |
A.19 Compliance & Legal
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| CMP-01 | Regulatory obligation mapping to M365 controls (GDPR, HIPAA, SOX, DPDP, ISO) | Purview | Compliance | Y/Q | P2 | E | ★★ Compliance Manager assessments | Turns legal text into configuration |
| CMP-02 | Data subject request (DSAR) fulfilment process and SLA 🔒 | Purview / Priva ⚠️ | Privacy | AH | P1 | A | ★★ Priva subject rights requests ⚠️ | Statutory deadlines with financial penalties |
| CMP-03 | Records management file plan and event-based retention 👻 | Purview | Records | Q | P3 | E | ★★ File plan export | Required for regulated record classes |
| CMP-04 | Legal hold issuance, tracking and release 🔒 | Purview eDiscovery | Legal | AH | P1 | A | ★★ Hold report automation | Missed holds = spoliation |
| CMP-05 | Audit evidence pack production for internal/external audit 👻 | Cross | Audit | Q/Y | P2 | A | ★★★ Scheduled evidence exports to immutable store | Auditors need proof, not assertions |
| CMP-06 | Third-party/vendor access attestation into the tenant 🔒 | Entra ID | Compliance | Q | P2 | A | ★★ Access review campaigns | Vendors are a recurring audit finding |
| CMP-07 | Data residency and cross-border transfer documentation ⚠️🔒 | M365 | Compliance | Y | P2 | E | ★ Documentation | Sovereignty regulation is tightening globally |
| CMP-08 | Copilot/AI usage policy alignment with acceptable-use and privacy policy ⚠️ | Copilot | AI governance | Q | P2 | A | ★ Policy review | AI governance is now an audit topic |
| CMP-09 | Retention conflict analysis (which policy wins) 👻 | Purview | Records | Q | P3 | E | ★ Analysis with retention principles | Misunderstood precedence causes premature deletion |
A.20 Governance & Change Management
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| GOV-01 | Change advisory / approval process for tenant-level changes 🔥 | Cross | Change | W | P1 | A | ★★ Change request flow + approval logging | Prevents unreviewed tenant-wide changes |
| GOV-02 | Standard vs emergency change classification and post-implementation review | Cross | Change | W | P2 | A | ★★ ITSM integration | Balances speed and safety |
| GOV-03 | Pilot ring / staged rollout definition for Microsoft feature changes 👻 | Cross | Change | M | P2 | A | ★★ Ring membership groups automated | Contains blast radius of Microsoft-driven change |
| GOV-04 | Naming convention and tagging standards enforcement | Cross | Standards | Q | P3 | A | ★★★ Provisioning-time enforcement | Retro-fixing names is nearly impossible |
| GOV-05 | Tenant configuration standards document ("golden config") maintenance 👻 | Cross | Standards | Q | P2 | E | ★★ Config-as-code repo | The reference against which drift is measured |
| GOV-06 | Service ownership matrix / RACI review | Cross | Organisation | Y | P3 | A | ★ Documentation | Prevents "I thought security owned that" |
| GOV-07 | Exception register: approved deviations with expiry dates 🔒👻 | Cross | Risk | Q | P1 | A | ★★ Register with automated expiry reminders | Un-expiring exceptions become permanent risk |
| GOV-08 | Licence-to-capability governance (who can request what) | Licensing | Governance | Q | P3 | I | ★★ Request workflow | Prevents ad-hoc premium licence assignment |
| GOV-09 | Data classification scheme ownership and business sign-off | Purview | Governance | Y | P2 | A | ☆ Business workshop | IT cannot own classification alone |
| GOV-10 | AI/agent governance board and approval pathway ⚠️ | Copilot / Power Platform | AI governance | M/Q | P2 | E | ★★ Intake + review workflow | Agent sprawl is the next shadow IT wave |
A.21 Monitoring & Observability
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| MON-01 | Synthetic transaction monitoring (send mail, upload file, Teams login) 👻 | Cross | Monitoring | RT | P2 | E | ★★★ Scheduled runbooks + status dashboard | Detects tenant-specific issues Microsoft's dashboard misses |
| MON-02 | Log ingestion architecture: Entra, audit, Defender → Sentinel/Log Analytics 🔒 | Sentinel | Monitoring | M | P1 | E | ★★★ Diagnostic settings as code | No logs = no investigation capability |
| MON-03 | Alert rule tuning and noise reduction 👻 | Cross | Monitoring | M | P2 | A | ★★ Alert volume review | Alert fatigue is the real cause of missed incidents |
| MON-04 | Alert routing, on-call rotation and escalation path validation 🔥 | Cross | Monitoring | Q | P1 | A | ★★ Integration with paging tool | An alert with nowhere to go is not monitoring |
| MON-05 | Capacity trend forecasting (mailbox, SPO, Dataverse, Exchange archive) | Cross | Capacity | M | P2 | A | ★★★ Trend model in Power BI | Buy capacity before the wall, not after |
| MON-06 | Expiry watchlist: certificates, secrets, tokens, domains, licences 🔥🔒👻 | Cross | Monitoring | W | P1 | A | ★★★ Unified expiry dashboard | The single highest-ROI monitoring artefact in M365 |
| MON-07 | Service health API integration with internal status page 👻 | M365 | Communication | RT | P2 | A | ★★★ Graph serviceHealth → status page | Cuts inbound ticket volume during incidents |
| MON-08 | User experience monitoring (Endpoint Analytics, CQD, network insights) | Cross | Experience | M | P3 | A | ★★ Consolidated dashboard | Measures what users actually feel |
A.22 Licensing & Cost
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| LIC-01 | Licence assignment reconciliation (assigned vs purchased vs active) | M365 Admin | Licensing | M | P2 | I | ★★★ Graph subscribedSkus + usage reports | Direct, provable cost saving |
| LIC-02 | Unused / dormant licence reclamation (no sign-in in N days) | M365 Admin | Cost | M | P2 | A | ★★★ Automated reclaim workflow with manager approval | Typically 5–15% of spend in large tenants |
| LIC-03 | Group-based licensing health and assignment error remediation 🔥👻 | Entra ID | Licensing | W | P2 | A | ★★★ Graph licence error report | Assignment errors silently leave users unlicensed |
| LIC-04 | Service plan disablement to enforce capability governance | M365 Admin | Governance | Q | P3 | A | ★★★ Group-based licensing with disabled plans | Controls unwanted service exposure |
| LIC-05 | SKU right-sizing analysis (E5 vs E3 + add-ons) ⚠️ | M365 Admin | Cost | Q | P2 | E | ★★ Usage-based modelling | Six-figure decisions in mid-size enterprises |
| LIC-06 | Renewal calendar and true-up preparation 👻 | Commerce | Cost | Q/Y | P1 | A | ★★ Renewal tracker | Missing a renewal negotiation window is costly |
| LIC-07 | Copilot / premium add-on ROI review ⚠️ | M365 Copilot | Cost | Q | P2 | A | ★★ Usage vs cost dashboard | Defends or kills the renewal on evidence |
| LIC-08 | Consumption-based cost monitoring (Copilot credits, Power Platform, M365 Backup, Archive, Azure) ⚠️ | Cross | Cost | M | P2 | A | ★★★ Cost alerting | Consumption models create unbudgeted surprises |
| LIC-09 | Guest / F-SKU / kiosk licensing appropriateness review 👻 | M365 Admin | Cost | Q | P3 | I | ★★ Report | Frontline workers are frequently over-licensed |
| LIC-10 | Licence-affected feature dependency mapping ⚠️👻 | Cross | Planning | Y | P2 | E | ★ Documentation | Downgrading a SKU silently disables controls you rely on |
A.23 Business Continuity & Recovery
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| BCP-01 | Backup strategy definition and coverage validation (native vs third-party vs M365 Backup) ⚠️🔥👻 | Cross | Recovery | Q | P1 | E | ★★ Coverage matrix + job monitoring | Microsoft's shared responsibility model does not cover your deletions |
| BCP-02 | Restore testing: mailbox, site, OneDrive, Teams, Dataverse 🔥👻 | Cross | Recovery | Q | P1 | A | ★★ Scheduled restore drills with evidence | An untested backup is a hope, not a control |
| BCP-03 | Retention vs backup gap analysis (soft delete windows, hard limits) ⚠️👻 | Cross | Recovery | Q | P1 | E | ★ Analysis | Deleted-site and deleted-mailbox windows are finite ⚠️ |
| BCP-04 | Tenant-level disaster scenario runbooks (GA loss, CA lockout, sync failure, ransomware) 🔥 | Cross | Resilience | Q | P1 | E | ★ Runbook maintenance | The scenarios that end careers |
| BCP-05 | Out-of-band communication plan when M365 is down 🔥👻 | Cross | Resilience | Y | P1 | A | ★ Documented alternate channel | You cannot coordinate an Exchange outage over Teams reliably |
| BCP-06 | Configuration backup/export for all workloads (config-as-code) 👻 | Cross | Recovery | W | P1 | E | ★★★ Scheduled export to Git | Rebuilding CA/DLP/transport rules from memory is not viable |
| BCP-07 | Dependency mapping: which business processes fail with which M365 service 👻 | Cross | Resilience | Y | P2 | A | ★ BIA workshop | Drives prioritisation during real outages |
| BCP-08 | Recovery time / recovery point objective definition per workload 👻 | Cross | Resilience | Y | P2 | E | ☆ Business agreement | Sets realistic expectations before the incident |
| BCP-09 | Third-party backup vendor health, quota and job failure monitoring 🔥 | Backup vendor | Recovery | D/W | P1 | I | ★★★ API/alert integration | Silent backup failure is the classic recovery-day surprise |
A.24 Automation Estate Operations
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| AUT-01 | Automation inventory: every flow, runbook, script, agent and its owner 👻 | Cross | Governance | Q | P1 | A | ★★★ Central registry auto-populated | You cannot operate what you cannot enumerate |
| AUT-02 | Automation failure alerting and triage queue 🔥 | Cross | Operations | D | P1 | A | ★★★ Centralised failure channel | Failures must be loud, not silent |
| AUT-03 | Automation run-history retention and audit trail 🔒 | Cross | Audit | M | P2 | A | ★★★ Log Analytics ingestion | Change attribution for automated actions |
| AUT-04 | Identity used by automation: service principal / managed identity review 🔒 | Cross | Privilege | Q | P1 | E | ★★ Permission scoping report | Automation identities are usually the most over-privileged |
| AUT-05 | Long-running / cost-anomalous job detection ⚠️ | Cross | Cost | M | P3 | A | ★★★ Duration and consumption alerting | Runaway loops cost real money |
| AUT-06 | Automation change control and environment promotion (dev→test→prod) | Cross | ALM | M | P2 | A | ★★★ Pipelines | Direct-to-production automation edits cause outages |
| AUT-07 | Deprecation watch for connectors, actions and APIs used ⚠️👻 | Cross | Maintenance | Q | P2 | A | ★★ Dependency scan + Message center mapping | Deprecated connectors break flows on Microsoft's schedule |
| AUT-08 | Automation documentation and handover readiness (bus factor) 👻 | Cross | Documentation | Q | P2 | A | ★★ Template-enforced README per automation | Undocumented automation becomes unmaintainable within months |
A.25 Documentation & Knowledge
| ID | Task | Service | Category | Freq | Pri | Skill | Automation Opportunity | Business Impact |
|---|
| DOC-01 | Tenant architecture document (identity, mail flow, data flow, network) 👻 | Cross | Documentation | Q/Y | P1 | E | ★ Maintained diagram set | Essential for incidents, audits, onboarding and M&A |
| DOC-02 | Runbook library for the top 30 recurring operational tasks | Cross | Documentation | Q | P2 | A | ★★ Template + review cycle | Enables delegation and holiday cover |
| DOC-03 | Decision log / architecture decision records (why is it like this?) 👻 | Cross | Documentation | AH | P2 | A | ★★ ADR repo in Git | Prevents relitigating settled decisions |
| DOC-04 | Service desk knowledge base and self-help article maintenance | Cross | Enablement | M | P3 | I | ★★ Article freshness reporting | Directly reduces ticket volume |
| DOC-05 | Configuration inventory ("what is configured, where, by whom") 👻 | Cross | Documentation | M | P1 | A | ★★★ Automated config export → readable docs | Doubles as drift baseline and audit evidence |
| DOC-06 | Onboarding pack for new administrators | Cross | Enablement | Y | P3 | I | ★ Maintained pack | Reduces time-to-competence |
| DOC-07 | Post-incident review records and remediation tracking 👻 | Cross | Learning | EV | P2 | A | ★★ PIR template + backlog linkage | The main mechanism that converts incidents into improvement |
| DOC-08 | Vendor/partner contact matrix and support entitlement record 👻 | Cross | Support | Y | P3 | B | ★ Documentation | You need this at 02:00, not during business hours |
A.26 Highlighted Task Groups
🔥 Tasks that can cause a business outage if neglected
| ID | Task | Typical failure |
|---|
| IDN-04 | Break-glass account validation | Total admin lockout; no way to fix a bad CA policy |
| IDN-05 | Conditional Access change control | One policy scoped to "All users / All apps / Block" locks out the company |
| IDN-15 | Entra Connect / Cloud Sync health | All identity changes freeze; new hires cannot be created |
| IDN-22 | Federation / auth method architecture | ADFS or PTA agent failure blocks all authentication |
| EXO-02 / EXO-05 | Connector cert / MX / accepted domain | Total inbound or outbound mail failure |
| EXO-18 | SMTP AUTH deprecation ⚠️ | Line-of-business apps stop sending mail overnight |
| M365-06 | Domain registration expiry | Catastrophic, entirely preventable, whole-tenant |
| SEC-03 | App secret / certificate expiry | Integrations, SSO and automation fail without warning |
| INT-07 | APNs / DEP / VPP token expiry | Every managed Apple device drops out of management |
| INT-17 | SCEP/PKCS/Cloud PKI health | Wi-Fi and VPN certificate failures at scale |
| TMS-12 | Direct Routing SBC certificate | Complete voice outage |
| TMS-11 | Teams Rooms account expiry | Every meeting room fails simultaneously |
| SPO-01 | Tenant storage quota | Uploads blocked org-wide |
| PPF-05 / PPF-07 | Connection expiry / Dataverse capacity | Business process automation halts |
| BCP-09 | Backup job failure | Discovered only on recovery day |
| PSH-07 | Bulk-change guardrails | Self-inflicted mass misconfiguration |
🔒 Tasks that create security or compliance exposure if neglected
IDN-02, IDN-03, IDN-06, IDN-09, IDN-10, IDN-11, IDN-13, EXO-09, EXO-12, EXO-16, SPO-02, SPO-03, SPO-04, ODB-02, INT-01, INT-06, DEF-04, DEF-05, DEF-08, DEF-09, DEF-14, PUR-01, PUR-02, PUR-05, PUR-08, GRF-01, GRF-02, SEC-04, SEC-05, SEC-06, SEC-12, SEC-14, CPL-02, CPL-09, PPF-02, AUT-04.
The five that most reliably appear in breach post-mortems: legacy authentication still permitted (SEC-06), over-permissioned Graph application grants (GRF-01), unreviewed Conditional Access exclusions (IDN-06), standing Global Admin without PIM (IDN-10), and external forwarding rules created by an attacker (EXO-12).
👻 Tasks most frequently forgotten by M365 administrators
| Rank | Task | Why it gets missed |
|---|
| 1 | M365-02 Message center triage | It looks like marketing email; it is actually a change stream |
| 2 | SEC-03 App secret/certificate expiry | No native dashboard; discovered as an outage |
| 3 | IDN-04 Break-glass validation | Feels unnecessary until the one day it isn't |
| 4 | IDN-06 CA exclusion audit | Exclusions are added under pressure and never removed |
| 5 | BCP-02 Restore testing | Backups are configured once and never proven |
| 6 | SPO-03 Anonymous link inventory | Invisible unless deliberately reported on |
| 7 | INT-07 APNs/DEP/VPP token expiry | Annual cadence, easy to forget between admins |
| 8 | GRF-02 Admin-consented delegated permissions | Consent is granted once, reviewed never |
| 9 | PPF-03/PPF-05 Orphaned flows and expired connections | Owned by "the business", supported by nobody |
| 10 | M365-04 Configuration drift detection | Requires deliberate engineering; no built-in tenant-wide diff |
| 11 | EXO-22 Journal/archive integration health | Fails silently, discovered during litigation |
| 12 | GOV-07 Exception register expiry | Temporary exceptions outlive the people who approved them |
| 13 | DOC-03 Decision log | The knowledge leaves with the person |
| 14 | ODB-03 Known Folder Move coverage | Assumed complete, rarely verified |
| 15 | M365-08 GDAP partner relationships | Partner access persists long after the project ends |
Highest-risk tasks (require change control, peer review and a rollback plan)
| Task | Why it is high-risk | Required control |
|---|
| Conditional Access policy change | Immediate tenant-wide auth impact | What-If evaluation, report-only first, break-glass exclusion verified, staged ring, peer review |
| Bulk PowerShell/Graph mutation | Irreversible at scale | -WhatIf run, batch limit, export of prior state, approval |
| Tenant-wide sharing / external access change | Data exposure or collaboration breakage | Impact analysis, business sign-off, pilot |
| Retention or deletion policy change | Permanent data loss | Legal sign-off, simulation, no-delete-first principle |
| Transport rule change | Silent mail interception or loss | Test with pilot group, message trace validation |
| Licence removal / SKU downgrade | Data deletion after grace period ⚠️, feature loss | Dependency map, staged, backup first |
| DNS / domain record change | Mail and auth outage | Change window, TTL reduction beforehand, rollback record set |
| App permission grant | Tenant-wide data access | Least privilege review, justification, expiry |
| Directory sync rule change | Mass attribute overwrite | Staging mode, preview, small-scope pilot |
| Deleting an Entra group used for licensing/CA | Mass licence and access loss | Ownership check, dependency scan, soft-delete window awareness |
A.27 Deep-Dive Task Cards (full field expansion)
These cards expand the complete field set for the tasks where the detail materially changes execution.
Card 1 — IDN-03 Leaver Deprovisioning
| Field | Detail |
|---|
| Task name | Leaver / termination deprovisioning |
| Service | Entra ID, Exchange Online, OneDrive, Teams, Intune, licensing |
| Description | On termination, disable sign-in, revoke all refresh tokens and sessions, remove MFA methods used for persistence, convert mailbox to shared or apply retention, transfer OneDrive ownership, remove from groups/Teams, wipe or retire devices, remove licences on a defined delay, and archive the audit trail. |
| Why it matters | Terminated-user access is the highest-frequency insider risk and a standard audit test. Session revocation matters more than password reset: an active refresh token survives a disabled account in some paths until revoked. |
| Frequency | Event-driven (HR trigger), with a daily reconciliation sweep |
| Priority | P1 for involuntary termination; P2 for standard resignation |
| Business impact | Prevents data theft, satisfies audit, releases licence cost, preserves records for legal |
| Skill | Advanced |
| Recommended tools | Microsoft Graph PowerShell, Entra Lifecycle Workflows, Azure Automation, Power Automate approval, ITSM integration |
| Admin center | Entra admin center, Exchange admin center, SharePoint admin center, Intune |
| Automation | ★★★ HR system → Graph webhook/scheduled sync → runbook. Sequence must be ordered: block sign-in → revoke sessions → remove auth methods → hold/retention → data reassignment → group removal → device action → licence removal (delayed 30–90 days) |
| Monitoring | Alert on terminated-user sign-in attempts; daily reconciliation of HR terminations vs enabled accounts |
| Documentation | Deprovisioning runbook, data-retention decision matrix, delegation approval record |
| Evidence | Per-user deprovisioning log with timestamps for each step, retained per records policy |
| Common failure | Licence removed immediately → mailbox and OneDrive enter deletion countdown before data is preserved; or account disabled but sessions never revoked |
| Preventive action | Hard rule: never remove a licence before retention/hold is confirmed applied. Automate session revocation as step two. |
| Escalation | Involuntary termination, suspected data theft, or executive/privileged account → immediate escalation to Security and Legal; apply hold before any data action |
Card 2 — IDN-05 Conditional Access Change Control
| Field | Detail |
|---|
| Task name | Conditional Access policy review, change control and drift detection |
| Service | Microsoft Entra ID (P1/P2) |
| Description | Maintain the CA policy set as versioned configuration: exported regularly, reviewed for overlap and gaps, changed only through report-only → pilot → production progression, with documented exclusions. |
| Why it matters | CA is the enforcement point for Zero Trust in M365 and simultaneously the single most effective way to lock out the entire organisation. |
| Frequency | Weekly drift check; monthly full review; ad-hoc per change |
| Priority | P1 |
| Business impact | Protects every application; misconfiguration is an immediate company-wide outage |
| Skill | Expert |
| Recommended tools | Graph API (identity/conditionalAccess/policies), What-If tool, report-only mode, sign-in log workbook, Git for versioning |
| Admin center | Entra admin center → Protection → Conditional Access |
| Automation | ★★ Scheduled Graph export → JSON → Git commit → diff alert on unexpected change; automated report-only impact analysis from sign-in logs |
| Monitoring | Alert on any CA policy create/update/delete in the audit log; alert on report-only failure spikes |
| Documentation | Policy register: purpose, scope, exclusions with justification and expiry, owner, last review date |
| Evidence | Versioned policy JSON history plus approval records |
| Common failure | New policy scoped to All users without break-glass exclusion; or an exclusion group that has quietly grown to hundreds of members |
| Preventive action | Mandatory break-glass exclusion on every policy; every policy starts in report-only for at least one business week; exclusion groups are role-assignable, monitored and reviewed monthly |
| Escalation | Widespread sign-in failure after change → immediate rollback using break-glass account; declare P1 |
Card 3 — SEC-03 App Registration Secret & Certificate Expiry
| Field | Detail |
|---|
| Task name | Application credential expiry tracking and rotation |
| Service | Microsoft Entra ID (app registrations, service principals) |
| Description | Maintain a live inventory of every client secret and certificate on every app registration, with owner, purpose, expiry date and rotation procedure. |
| Why it matters | There is no native proactive alert. Expiry produces a hard, unannounced failure of integrations, SSO and automation. |
| Frequency | Weekly automated scan; monthly review |
| Priority | P1 |
| Business impact | Line-of-business integration outages, failed automation, broken SAML SSO |
| Skill | Intermediate |
| Recommended tools | Microsoft Graph PowerShell (Get-MgApplication), Azure Automation, Key Vault, Power Automate/Teams alerts |
| Admin center | Entra admin center → App registrations |
| Automation | ★★★ Scheduled Graph query listing passwordCredentials and keyCredentials with endDateTime; alert at 90/60/30/14/7 days to the registered owner and the platform team |
| Monitoring | Dashboard tile "credentials expiring in next 90 days"; zero-owner apps flagged separately |
| Documentation | App registry: app name, owner, business purpose, permissions, credential type, rotation method, dependency list |
| Evidence | Rotation history log |
| Common failure | The app has no owner recorded; expiry alert has nowhere to go; or rotation is done but the consuming system is never updated |
| Preventive action | Prefer certificates over secrets; prefer workload identity federation or managed identity over both, which removes the expiry class entirely. Enforce owner assignment at app creation. |
| Escalation | Expired credential on a production integration → P1; engage app owner and vendor immediately |
Card 4 — CPL-02 Copilot Readiness / Oversharing Remediation
| Field | Detail |
|---|
| Task name | Pre-Copilot data governance and oversharing remediation |
| Service | SharePoint Online, OneDrive, Purview, Microsoft 365 Copilot |
| Description | Before and during Copilot rollout, identify and remediate content that is broadly accessible but should not be — "Everyone Except External Users" grants, anonymous links, ownerless sites, unlabelled sensitive content — and constrain Copilot's grounding scope until remediation is complete. |
| Why it matters | Copilot does not break permissions; it makes existing permission mistakes instantly discoverable through natural language. Latent oversharing that was safe-by-obscurity becomes actively surfaced. |
| Frequency | Intensive pre-rollout project; quarterly thereafter |
| Priority | P1 before any broad Copilot enablement |
| Business impact | HR, finance, legal and M&A content surfacing to the wrong staff is a career-level incident |
| Skill | Expert |
| Recommended tools | SharePoint Advanced Management (Data Access Governance reports, Restricted Access Control, Restricted Content Discovery) ⚠️, Purview DSPM for AI, sensitivity labels, PnP PowerShell |
| Admin center | SharePoint admin center, Purview portal, M365 admin center |
| Automation | ★★ Scheduled DAG report extraction, EEEU detection scripts, automated remediation for known-bad patterns, label auto-application |
| Monitoring | Trend of sites with broad access; count of anonymous links; unlabelled sensitive-content volume |
| Documentation | Remediation plan per site tier; exception register for accepted broad access |
| Evidence | Before/after oversharing metrics; sign-off from data owners |
| Common failure | Copilot enabled tenant-wide on day one because licences were purchased, before any remediation |
| Preventive action | Phase the rollout: restrict search/discovery scope → remediate top-risk sites → pilot cohort → expand. Make site owners accountable via attestation. |
| Escalation | Confirmed exposure of regulated or restricted data via Copilot → security incident process, plus immediate scope restriction |
Card 5 — BCP-02 Restore Testing
| Field | Detail |
|---|
| Task name | Recovery validation across workloads |
| Service | Exchange Online, SharePoint Online, OneDrive, Teams, Dataverse, plus any third-party backup |
| Description | Quarterly, perform an actual restore — not a report review — of a mailbox, a site, a OneDrive account, a Teams channel's files and a Dataverse table, recording elapsed time and completeness. |
| Why it matters | Native soft-delete windows are finite ⚠️, retention is not backup, and third-party backup jobs fail silently. The only proof of recoverability is a completed restore. |
| Frequency | Quarterly, rotating workloads |
| Priority | P1 |
| Business impact | Determines whether a ransomware or mass-deletion event is a bad week or an existential event |
| Skill | Advanced |
| Recommended tools | Microsoft 365 Backup ⚠️, third-party backup console, PowerShell restore cmdlets, Files Restore, Recycle Bin |
| Admin center | Varies by workload |
| Automation | ★★ Automated test-restore into an isolated target with completeness comparison; automated evidence capture |
| Monitoring | Backup job success rate, protected-object coverage percentage, oldest recoverable point per workload |
| Documentation | Restore runbook per workload, RTO/RPO register, test results archive |
| Evidence | Dated restore test report with elapsed time, scope, success/failure and gaps identified |
| Common failure | Coverage gaps discovered during the test — newly created sites, Teams chat data, Planner, Dataverse and Power Platform assets are frequently unprotected |
| Preventive action | Auto-enrol new objects into protection scope; alert on any unprotected object older than 7 days |
| Escalation | Failed restore or coverage gap on business-critical data → risk register entry and remediation project with executive visibility |
B. KTBE Master Task Catalog
Same schema; here Automation Opportunity describes the build approach and Business Impact describes the value created.
B.1 Identity & Access Modernisation
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-IDN-01 | Passwordless / phishing-resistant authentication programme (passkeys, FIDO2, Windows Hello, certificate-based auth) | Entra ID | Zero Trust | Project | P1 | E | Registration campaigns, CA enforcement by ring, kiosk/frontline patterns | Removes the dominant attack vector and cuts password reset tickets sharply |
| E-IDN-02 | Entra ID Governance rollout: access packages, entitlement management, lifecycle workflows ⚠️ | Entra ID Governance | Governance | Project | P1 | E | Catalogue design, package per role, approval + expiry policies | Converts manual access administration into self-service with audit trail |
| E-IDN-03 | Full PIM adoption: zero standing privilege, approval + justification + time-bound roles | Entra ID P2 | Privilege | Project | P1 | E | Role-by-role migration, approver matrix, alerting | Reduces the impact of any single admin account compromise |
| E-IDN-04 | Dynamic group strategy replacing manual membership | Entra ID | Automation | Project | P2 | A | Attribute quality first, then rule design, then cutover | Eliminates a large recurring manual workload; improves accuracy |
| E-IDN-05 | Conditional Access policy consolidation and persona-based design | Entra ID | Zero Trust | Project | P1 | E | Persona model (internal, admin, guest, service account, frontline, BYOD), template-based rebuild | Fewer, clearer policies; fewer gaps; far easier to reason about |
| E-IDN-06 | Device-bound and compliant-device requirement expansion | Entra / Intune | Zero Trust | Project | P2 | A | Staged CA enforcement per app tier | Blocks credential-only access from unmanaged endpoints |
| E-IDN-07 | Guest self-service with sponsor approval and automatic expiry | Entra ID Governance | External | Project | P2 | A | Access packages + sponsor approval + time-bound assignment | Removes guest admin toil while tightening control |
| E-IDN-08 | Retire on-prem federation → cloud authentication (PHS/PTA + CA) ⚠️ | Entra ID | Modernisation | Project | P1 | E | Staged rollout with staged rollout feature, fallback plan | Removes a major single point of failure and an on-prem attack surface |
| E-IDN-09 | Workload identity federation to eliminate app secrets | Entra ID | Security | Project | P2 | E | Migrate CI/CD and automation identities off secrets | Removes an entire outage and breach class |
| E-IDN-10 | Cross-tenant collaboration model design (B2B direct connect, MTO) ⚠️ | Entra ID | Collaboration | Project | P3 | E | Partner tiering, trust settings, shared channel governance | Enables secure partner and post-M&A collaboration |
| E-IDN-11 | Entra Verified ID / verifiable credentials for onboarding or partner verification ⚠️ | Entra ID | Innovation | Project | P4 | E | Issuer setup, use-case pilot | Reduces identity-proofing friction; emerging capability |
| E-IDN-12 | Global Secure Access / Entra Internet & Private Access evaluation ⚠️ | Entra | Network security | Project | P3 | E | Pilot with a defined user group and app set | Consolidates VPN/SWG spend into the identity stack |
| E-IDN-13 | Service account discovery, ownership and modernisation to managed identities 👻 | Entra ID | Hygiene | Project | P2 | E | Inventory → classify → migrate → retire | Service accounts are the most neglected privileged population |
B.2 Security & Zero Trust Maturity
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-SEC-01 | Security automation with Sentinel/Defender playbooks (SOAR) | Sentinel / Logic Apps | Automation | Project | P1 | E | Playbooks for account compromise, phishing, risky sign-in, device isolation | Cuts mean time to respond from hours to minutes |
| E-SEC-02 | Secure Score improvement programme with quarterly targets | Defender XDR | Posture | Programme | P2 | A | Backlog of improvement actions with owners and dates | A single credible, trackable executive security metric |
| E-SEC-03 | Attack surface reduction rules from audit to block, at scale | Defender for Endpoint | Hardening | Project | P2 | A | Audit → analyse → exception → enforce, per rule | Blocks common malware execution techniques |
| E-SEC-04 | Automatic attack disruption and full AIR automation enablement ⚠️ | Defender XDR | Automation | Project | P2 | A | Raise automation level; validate containment behaviour | Machine-speed containment of active attacks |
| E-SEC-05 | Insider risk programme with HR connector integration ⚠️ | Purview IRM | Risk | Project | P2 | E | HR data connector, policy tuning, reviewer workflow, privacy controls | Detects pre-departure data exfiltration |
| E-SEC-06 | Privileged Access Workstation / secure admin workstation deployment | Intune / Entra | Hardening | Project | P2 | E | Dedicated device profile, CA restriction to PAW compliance | Protects the accounts that can destroy the tenant |
| E-SEC-07 | Token protection and continuous access evaluation enforcement ⚠️ | Entra ID | Zero Trust | Project | P2 | E | Staged CA rollout for supported clients | Defeats token replay attacks that bypass MFA |
| E-SEC-08 | Custom detection rules and hunting library development | Defender XDR | Detection | Programme | P2 | E | KQL library in version control, tested and documented | Detection tailored to your environment and threat model |
| E-SEC-09 | Security Copilot / AI-assisted SOC evaluation ⚠️ | Security Copilot | AI | Project | P3 | E | Use-case pilot: incident summarisation, hunting assistance | Force multiplier for small security teams |
| E-SEC-10 | Third-party risk: OAuth app governance and app-consent workflow | Defender for Cloud Apps / Entra | Governance | Project | P2 | A | App governance policies, consent request queue, review SLA | Prevents illicit consent attacks; visible shadow IT reduction |
| E-SEC-11 | Continuous compliance scanning against CIS/benchmark baselines | Cross | Posture | Programme | P2 | E | Scheduled assessment tooling, delta reporting, remediation backlog | External yardstick; audit-ready evidence |
| E-SEC-12 | Purple team / breach simulation against M365 attack paths | Cross | Validation | Q/Y | P3 | E | Scenario-based exercises with detection validation | Proves controls work, not just exist |
B.3 Automation & Engineering
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-AUT-01 | Joiner–Mover–Leaver end-to-end automation | Graph / Azure Automation / Power Automate | Lifecycle | Project | P1 | E | HR source of truth → event → orchestrated runbook → ITSM ticket closure | The single highest-ROI automation in M365 administration |
| E-AUT-02 | Configuration-as-code for tenant settings (CA, DLP, Intune, transport, SPO) | Graph / PowerShell / Git | Engineering | Project | P1 | E | Export → repo → PR review → deploy pipeline → drift alert | Turns tenant configuration into a reviewable, restorable asset |
| E-AUT-03 | Self-service portal for common requests (group, licence, mailbox, access) | Power Apps + Power Automate | Self-service | Project | P1 | A | Catalogue of request types, approvals, Graph execution, audit log | Deflects 30–50% of routine admin tickets |
| E-AUT-04 | Unified expiry monitoring service (secrets, certs, tokens, domains) | Graph + Azure Automation | Monitoring | Project | P1 | A | Single scheduled job feeding a dashboard and tiered alerts | Removes the most common preventable outage class |
| E-AUT-05 | Automated reporting pipeline: Graph reports → data store → Power BI | Graph / Power BI | Reporting | Project | P2 | A | Scheduled extraction, historical retention, refresh monitoring | Replaces manual monthly report assembly permanently |
| E-AUT-06 | Automated licence reclamation with manager approval | Graph + Power Automate | Cost | Project | P1 | A | Inactivity detection → approval → reclaim → audit | Direct, recurring, provable cost saving |
| E-AUT-07 | Event-driven automation using Graph change notifications | Graph webhooks + Azure Functions | Engineering | Project | P2 | E | Subscription lifecycle management, idempotent handlers | Near-real-time response instead of scheduled polling |
| E-AUT-08 | Automated onboarding of new sites/teams with governance baked in | PnP / Power Automate | Governance | Project | P2 | A | Request → approve → provision with template, labels, permissions, owners | Prevents the remediation work that follows uncontrolled creation |
| E-AUT-09 | Automated evidence collection for audit | Graph / PowerShell | Compliance | Project | P2 | A | Scheduled exports to immutable storage with retention | Turns audit season from weeks into hours |
| E-AUT-10 | ChatOps: Teams-based admin operations with approval and logging | Teams + Power Automate + Graph | Self-service | Project | P3 | A | Adaptive cards, role-checked actions, full audit trail | Fast, safe, visible delegated operations |
| E-AUT-11 | Automation framework standardisation (logging, retry, secrets, notification) | PowerShell / Azure | Engineering | Project | P2 | E | Shared module with common functions and conventions | Consistency, maintainability, lower bus factor |
| E-AUT-12 | Automated drift remediation (detect → alert → optionally self-heal) | Graph / Azure Automation | Engineering | Project | P3 | E | Desired-state definition per workload, controlled auto-correction | Configuration converges instead of decaying |
B.4 AI, Copilot & Agents
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-AI-01 | Copilot readiness programme (data hygiene, labels, permissions, search quality) | SPO / Purview | Foundation | Project | P1 | E | Oversharing remediation, label taxonomy, IA cleanup, restricted discovery | Determines whether Copilot delivers value or embarrassment |
| E-AI-02 | Copilot adoption programme: champions, scenario library, training, measurement | M365 Copilot | Adoption | Programme | P1 | A | Role-based scenario packs, champions network, usage dashboard | Copilot value is adoption-limited, not capability-limited |
| E-AI-03 | Copilot Studio agent development for high-volume internal queries (IT, HR, policy) | Copilot Studio | Automation | Project | P2 | A | Knowledge sources, topics, actions, testing, publishing to Teams | Deflects repetitive queries at low marginal cost |
| E-AI-04 | Agent governance framework: intake, review, ALM, ownership, decommission ⚠️ | Copilot Studio / Power Platform | Governance | Project | P1 | E | Environment strategy, DLP, approval board, agent registry | Prevents the next generation of shadow IT |
| E-AI-05 | Declarative agents / custom Copilot extensions for line-of-business data ⚠️ | M365 Copilot extensibility | Innovation | Project | P3 | E | Graph connectors or plugin design, security review | Brings non-M365 data into the Copilot experience |
| E-AI-06 | Microsoft Graph connectors to index external content for search and Copilot | Graph connectors | Knowledge | Project | P2 | A | Connector setup, schema mapping, ACL mapping, ranking tuning | Makes enterprise knowledge findable in one place |
| E-AI-07 | Copilot value measurement model (time saved, task deflection, cost avoided) | M365 Copilot | Measurement | Programme | P2 | A | Baseline survey + usage telemetry + business metric linkage | Required to justify a very expensive per-seat spend |
| E-AI-08 | AI acceptable-use policy, prompt hygiene guidance and data-handling training ⚠️ | Cross | Governance | Project | P2 | A | Policy + mandatory training + monitoring | Reduces the risk of sensitive data entering AI workflows |
| E-AI-09 | Copilot licence rotation / cohort model to maximise seat utilisation ⚠️ | M365 Copilot | Cost | Programme | P2 | A | Time-boxed cohorts, usage-based reassignment | Higher realised value per licence |
| E-AI-10 | Agentic process automation: multi-step agents with connectors and human approval ⚠️ | Copilot Studio / Power Platform | Innovation | Project | P3 | E | Process selection, connector security, approval checkpoints, observability | Next-generation automation beyond deterministic flows |
| E-AI-11 | AI-assisted operations for the admin team (script generation, log analysis, doc drafting) | Cross | Productivity | Ongoing | P3 | A | Tooling + guardrails + review discipline | Multiplies a small platform team's output |
B.5 Collaboration & Workplace Modernisation
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-COL-01 | SharePoint intranet modernisation with hub architecture and audience targeting | SharePoint Online | Modernisation | Project | P2 | A | IA design, hub structure, templates, content ownership model | Findability and communications effectiveness |
| E-COL-02 | File share → SharePoint/OneDrive migration with permission redesign | SPO / OneDrive | Migration | Project | P1 | E | Assess, classify, restructure permissions, migrate, decommission | Removes on-prem storage cost and enables Copilot grounding |
| E-COL-03 | Teams template and lifecycle standardisation by work pattern | Teams | Governance | Project | P2 | A | Templates per scenario, provisioning automation, archival policy | Consistency, less sprawl, faster team setup |
| E-COL-04 | Teams Phone / calling modernisation from legacy PBX ⚠️ | Teams Phone | Modernisation | Project | P2 | E | Number migration, routing design, E911, user readiness | Significant telephony cost reduction |
| E-COL-05 | Meeting room modernisation with Teams Rooms and managed services ⚠️ | Teams Rooms | Experience | Project | P3 | A | Device standardisation, management tooling, monitoring | Reliable hybrid meetings; measurable satisfaction gain |
| E-COL-06 | Viva Connections employee experience layer ⚠️ | Viva | Experience | Project | P3 | A | Dashboard design, audience targeting, mobile-first content | Single front door for the digital workplace |
| E-COL-07 | Knowledge management and search improvement programme | SPO / Search / Graph connectors | Knowledge | Programme | P2 | A | Metadata, ranking, promoted results, content freshness rules | Directly improves Copilot answer quality |
| E-COL-08 | External collaboration model design (guest vs shared channel vs B2B site) | Teams / SPO / Entra | Collaboration | Project | P2 | E | Decision tree, provisioning patterns, governance controls | Safe partner collaboration without shadow IT |
| E-COL-09 | Loop / co-authoring and modern content pattern enablement ⚠️ | M365 | Modernisation | Project | P4 | I | Pilot, governance and retention validation first | Reduces attachment-based collaboration |
| E-COL-10 | Frontline worker enablement (F SKUs, shared devices, Shifts, Walkie Talkie) ⚠️ | Teams / Entra / Intune | Inclusion | Project | P2 | A | Shared-device auth patterns, simplified UX, kiosk provisioning | Extends M365 value to the largest under-served population |
| E-COL-11 | Accessibility and inclusivity improvement across M365 experiences 👻 | Cross | Experience | Programme | P3 | I | Accessibility checker adoption, captioning, template compliance | Legal obligation in many jurisdictions; broad benefit |
B.6 Endpoint & Device Modernisation
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-END-01 | GPO → Intune settings catalogue migration | Intune | Modernisation | Project | P1 | E | Group Policy analytics, mapping, staged migration, AD DS retirement path | Removes domain-join dependency; enables true remote management |
| E-END-02 | Windows Autopilot / device provisioning modernisation | Intune | Provisioning | Project | P1 | A | Autopilot profiles, ESP, hardware vendor registration | Zero-touch deployment; faster onboarding |
| E-END-03 | Windows Autopatch adoption ⚠️ | Autopatch | Patching | Project | P1 | A | Ring design, exclusion policy, reporting | Removes recurring patch orchestration effort |
| E-END-04 | Hybrid join → Entra-only join transition | Entra / Intune | Modernisation | Project | P2 | E | Application dependency analysis, printing/file-share alternatives | Eliminates on-prem AD dependency for endpoints |
| E-END-05 | Intune Suite capability adoption (EPM, Remote Help, Advanced Analytics, Enterprise App Management, Cloud PKI) ⚠️ | Intune Suite | Capability | Project | P2 | A | Business case per module; pilot then expand | Replaces multiple third-party tools |
| E-END-06 | Local admin rights removal via Endpoint Privilege Management ⚠️ | Intune Suite | Security | Project | P1 | E | Discovery of elevation needs → rules → enforcement | One of the highest-impact endpoint security controls |
| E-END-07 | Third-party application patching via Enterprise App Management ⚠️ | Intune Suite | Patching | Project | P2 | A | Catalogue apps, automated update policies | Closes the largest remaining endpoint vulnerability gap |
| E-END-08 | Windows 365 / Cloud PC for contractors, BYOD, developers and BCP ⚠️ | Windows 365 | Modernisation | Project | P2 | A | Persona-based sizing, network design, image pipeline | Secure access without shipping hardware |
| E-END-09 | macOS / mobile management maturity to parity with Windows 👻 | Intune | Coverage | Project | P2 | A | Platform-specific baselines, compliance, app protection | Closes the most common device-governance gap |
| E-END-10 | Endpoint analytics-driven experience optimisation | Intune | Experience | Programme | P3 | A | Baseline, target setting, remediation scripts | Measurable productivity improvement (boot time, app hangs) |
B.7 Data, Compliance & Information Protection
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-DAT-01 | Sensitivity label taxonomy design, rollout and auto-labelling | Purview | Data protection | Project | P1 | E | Minimal taxonomy, pilot, auto-label simulation, then enforcement | Persistent protection that travels with the file |
| E-DAT-02 | Endpoint DLP and browser DLP expansion ⚠️ | Purview | Data protection | Project | P2 | A | Audit mode → tune → block on high-confidence policies | Covers the USB/upload/print exfiltration paths |
| E-DAT-03 | DLP for AI / Copilot interactions and generative AI apps ⚠️ | Purview | AI governance | Project | P2 | E | Policy scoping to AI surfaces; DSPM for AI monitoring | Prevents sensitive data flowing into AI tools |
| E-DAT-04 | Records management file plan and event-based retention implementation | Purview | Records | Project | P2 | E | Legal-led file plan, phased scope, disposition workflow | Defensible disposal reduces both cost and legal exposure |
| E-DAT-05 | Retention policy rationalisation and over-retention reduction 👻 | Purview | Cost/Risk | Project | P2 | E | Conflict analysis, business validation, staged tightening | Reduces storage cost and eDiscovery burden |
| E-DAT-06 | eDiscovery process automation and legal-hold self-service for Legal ⚠️ | Purview eDiscovery | Legal | Project | P3 | A | Delegated roles, templates, Graph API automation | Removes IT from the legal critical path |
| E-DAT-07 | Compliance Manager operationalisation with owner assignment | Purview | Compliance | Programme | P2 | A | Assessment selection, control ownership, evidence automation | Continuous rather than annual compliance |
| E-DAT-08 | Data minimisation / stale content disposal programme 👻 | SPO / Purview | Cost | Programme | P3 | A | Inactivity reports, owner attestation, archive tiering | Cuts storage cost and improves Copilot precision |
| E-DAT-09 | M365 Archive / tiered storage adoption for cold content ⚠️ | M365 Archive | Cost | Project | P3 | A | Identify cold sites, cost model, archive policy | Cheaper retention of inactive but required content |
| E-DAT-10 | Privacy programme automation: DSAR, consent, privacy risk ⚠️ | Purview / Priva | Privacy | Project | P2 | A | Request intake workflow, automated discovery, SLA tracking | Statutory compliance with less manual effort |
B.8 Cost Optimisation & Commercial
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-CST-01 | Licence baseline and continuous optimisation engine | Graph / Power BI | Cost | Project | P1 | A | Assigned vs active vs required model, monthly reclaim cycle | Recurring savings, typically the fastest ROI in the KTBE list |
| E-CST-02 | SKU rationalisation and add-on strategy ⚠️ | Commerce | Cost | Q/Y | P1 | E | Feature-usage mapping, scenario modelling for renewal | Major contract-level savings or capability gains |
| E-CST-03 | Third-party tool displacement by owned M365 capability 👻 | Cross | Cost | Programme | P1 | E | Inventory of tools whose function exists in E5/Intune Suite/Purview | Often the single largest documented saving |
| E-CST-04 | Storage cost optimisation (version trimming, archive, quota policy) | SPO / EXO | Cost | Project | P2 | A | Version policy, archive tiering, quota governance | Defers additional storage purchase |
| E-CST-05 | Consumption governance for credits, capacity and pay-as-you-go ⚠️ | Copilot Studio / Power Platform / Azure | Cost | Programme | P2 | A | Budgets, alerts, environment-level limits, chargeback | Prevents unbudgeted overruns |
| E-CST-06 | Chargeback / showback model for business units | Cross | Financial | Project | P3 | A | Cost allocation by department, reporting cadence | Changes consumption behaviour more than any policy |
| E-CST-07 | Renewal negotiation preparation pack with usage evidence | Commerce | Cost | Y | P1 | A | Usage data, growth model, competitive analysis | Materially improves negotiating position |
B.9 Operations Maturity & Enablement
| ID | Initiative | Service | Category | Freq | Pri | Skill | Build Approach | Business Value |
|---|
| E-OPS-01 | Tiered support model: what L1/L2/L3 owns, with delegated permissions | Cross | Operating model | Project | P1 | A | Least-privilege admin roles, runbooks, tooling for L1 | Frees senior capacity for KTBE work |
| E-OPS-02 | Service catalogue and SLA definition for M365 services | Cross | Operating model | Project | P2 | A | Catalogue, SLOs, request routing | Sets expectations; enables measurement |
| E-OPS-03 | Executive reporting pack (monthly) with risk, cost, adoption, incidents | Power BI | Reporting | Programme | P1 | A | Automated data pipeline, narrative layer | Converts admin work into business visibility and funding |
| E-OPS-04 | Adoption and change management practice for every rollout | Cross | Adoption | Programme | P2 | A | Champions, comms plan, training, measurement | Determines whether capability becomes value |
| E-OPS-05 | Message center → backlog automation with impact classification | Graph | Change | Project | P1 | A | API ingestion, classification, routing to Planner/DevOps | The change stream becomes managed instead of missed |
| E-OPS-06 | Innovation / pilot ring programme for preview features | Cross | Innovation | Programme | P3 | A | Volunteer cohort, feedback loop, go/no-go criteria | Early insight; controlled exposure to Microsoft change |
| E-OPS-07 | Platform team operating model: product ownership per workload | Cross | Organisation | Project | P2 | E | Named owners, roadmaps, backlogs per service | Ends the "everyone owns everything" failure mode |
| E-OPS-08 | Continuous learning programme for the admin team | Cross | Capability | Programme | P2 | I | Certification plan, lab tenant, internal brown-bags | Capability is the binding constraint on KTBE delivery |
| E-OPS-09 | Non-production / lab tenant for safe testing 👻 | Cross | Engineering | Project | P1 | A | Dev tenant with representative config | Enables testing without risking production |
| E-OPS-10 | Business relationship management with functional stakeholders | Cross | Engagement | Ongoing | P2 | A | Quarterly service reviews per business unit | Surfaces demand early; builds political capital |
B.10 Additional KTBE Opportunities Commonly Missed
| ID | Initiative | Why it is usually missed | Value |
|---|
| E-X-01 | Graph connectors to index ERP/CRM/ticketing content | Seen as a search feature, not an AI enabler | Transforms Copilot and search usefulness |
| E-X-02 | Bookings, Forms, Lists and Approvals as low-effort process tools | Considered "too small" to matter | Fast wins with near-zero build cost |
| E-X-03 | Power BI embedded operational dashboards for business owners | IT keeps data to itself | Owners self-serve; IT stops being a report factory |
| E-X-04 | Sensitivity-label-driven container protection (auto-configured sites) | Labels seen as a document-only control | Governance applied at creation, not retro-fitted |
| E-X-05 | Automated user offboarding knowledge transfer (OneDrive/Teams handover) | Treated as a manual HR ask | Retains institutional knowledge |
| E-X-06 | Language, translation and inclusivity features in Teams/Office | Never explicitly enabled or communicated | High value for global workforces |
| E-X-07 | Green/sustainability reporting from M365 usage data ⚠️ | Not on the IT radar | Increasingly a board-level reporting requirement |
| E-X-08 | Standardised email signature and branding management | Left to a third-party tool by default | Brand consistency and compliance disclaimers |
| E-X-09 | Meeting hygiene programme (Viva Insights-informed) ⚠️ | Considered a culture issue, not IT | Measurable reclaimed hours |
| E-X-10 | Tenant-to-tenant migration readiness capability | Only built during an M&A panic | Strategic capability for a growing organisation |
| E-X-11 | Developer enablement: Graph API access patterns for internal teams | IT hoards the API surface | Internal teams build their own value safely |
| E-X-12 | Shadow IT discovery → sanction or replace programme | Discovery happens, action rarely follows | Converts risk into either governance or savings |
C. Operational Calendar
C.1 Continuous / Real-time (automated, alert-driven — no human polling)
| Signal | Source | Routing |
|---|
| P1 Defender XDR incidents | Defender XDR | SOC on-call page |
| High-risk user / risky sign-in | Entra ID Protection | Security channel + auto-remediation |
| Service health advisory affecting your tenant | Graph serviceHealth API | Ops Teams channel + status page |
| Mail flow anomaly (queue depth, NDR spike) | EXO monitoring | Ops channel |
| Automation/runbook failure | Azure Monitor / Power Platform | Automation failure channel |
| Global Admin role activation | Entra audit log | Security channel |
| CA policy created/modified/deleted | Entra audit log | Change channel + Git diff |
| Mass file deletion / download anomaly | Defender for Cloud Apps / alert policy | Security channel |
| Tenant storage or Dataverse capacity threshold | Scheduled check | Ops channel |
| Backup job failure | Backup platform | Ops channel |
C.2 Daily (target: 45–60 minutes, batched)
| # | Check | Where | Time |
|---|
| 1 | Service health + active advisories; assess user impact; publish comms if needed | M365 admin center / status page | 5 min |
| 2 | Message center: triage new posts, classify (action / awareness / ignore), assign owners | M365 admin center | 10 min |
| 3 | Defender XDR incident queue: new, unassigned, ageing high-severity | Defender portal | 10 min |
| 4 | Identity Protection: risky users and risky sign-ins since yesterday | Entra admin center | 5 min |
| 5 | Privileged role activations in the last 24h — expected or not? | Entra PIM | 3 min |
| 6 | Mail flow: queues, NDR trend, restricted senders, connector errors | EAC / mail flow reports | 5 min |
| 7 | Quarantine: high-confidence phish releases requested; false positive check | Defender portal | 3 min |
| 8 | User-reported phishing submissions | Defender portal | 5 min |
| 9 | Automation health: failed runbooks, failed flows, failed scheduled jobs | Central failure channel | 5 min |
| 10 | Backup job status | Backup console | 3 min |
| 11 | Device compliance delta: new non-compliant devices | Intune | 3 min |
| 12 | P1/P2 ticket queue and escalations | ITSM | 5 min |
| 13 | Directory sync errors (hybrid tenants) | Entra Connect Health | 3 min |
| 14 | Dataverse / Power Platform capacity and failed flows | PPAC | 3 min |
Daily principle: anything on this list that can be turned into an alert should be. The daily check is for judgement, not detection.
C.3 Weekly (target: 3–4 hours)
| # | Review | Output |
|---|
| 1 | Conditional Access drift check (Git diff vs baseline) | Change record or rollback |
| 2 | Expiry watchlist: secrets, certificates, tokens, domains due in 90 days | Rotation tasks raised |
| 3 | MFA / phishing-resistant auth coverage delta; new users unregistered | Remediation list |
| 4 | Legacy authentication attempts in sign-in logs | App migration backlog |
| 5 | External forwarding rule report | Investigation of any new entries |
| 6 | Licence assignment errors and group-based licensing failures | Fixes |
| 7 | Intune: compliance %, update ring status, app deployment failures | Remediation tasks |
| 8 | Defender vulnerability exposure: new critical CVEs in the estate | Patch tasks to Intune |
| 9 | Sentinel / log ingestion volume anomaly check | Connector fixes |
| 10 | Storage trend: tenant SPO, top-growing sites, mailbox outliers | Capacity forecast update |
| 11 | Teams call quality (CQD) summary | Network/device actions |
| 12 | Automation estate: failure rate, long-running jobs, new automations registered | Backlog grooming |
| 13 | Change/CAB: review pending tenant changes, approve, schedule | Change calendar |
| 14 | Message center backlog burn-down | Assignments closed |
| 15 | Power Platform: new environments, new apps/flows, DLP violations | Governance actions |
| 16 | Configuration backup export verified (config-as-code commit succeeded) | Confirmation |
| 17 | Protected KTBE block — project delivery | Progress on current initiative |
C.4 Monthly (target: 1–1.5 days)
| # | Review | Output |
|---|
| 1 | Privileged role inventory: GAs, standing assignments, PIM compliance | Reduction actions |
| 2 | Access reviews cycle: privileged roles, guests, high-value groups | Attestation evidence |
| 3 | Guest account inventory: inactive, unsponsored, expired | Removals |
| 4 | Enterprise applications & service principals: new, dormant, over-permissioned | Consent revocations |
| 5 | Graph application permission review (delta since last month) | Least-privilege reductions |
| 6 | Licence optimisation: unused, dormant, duplicate, wrong-SKU | Reclaim batch + savings figure |
| 7 | Secure Score: movement, regressions, next improvement actions | Backlog with owners |
| 8 | DLP incident trend and false-positive rate | Policy tuning |
| 9 | Sharing exposure: anonymous links created, EEEU grants, external guest additions | Remediation |
| 10 | Shared mailbox and delegation audit delta | Corrections |
| 11 | Retention/hold coverage validation; inactive mailbox report | Compliance evidence |
| 12 | Audit log spot-check on privileged operations | Investigation notes |
| 13 | Copilot usage, licence utilisation, agent inventory | Reassignment + governance actions |
| 14 | Teams/SharePoint/Group sprawl: new, dormant, ownerless | Lifecycle actions |
| 15 | Endpoint estate: stale devices, encryption coverage, baseline drift | Cleanup + remediation |
| 16 | Automation inventory reconciliation and ownership confirmation | Registry updated |
| 17 | Cost review: consumption services, storage growth, forecast vs budget | Finance update |
| 18 | Executive report produced and circulated | Monthly ops pack |
| 19 | Documentation freshness check on top 20 runbooks | Updates |
| 20 | Post-incident review of the month's P1/P2 incidents | Remediation backlog items |
C.5 Quarterly (target: 3–5 days)
| # | Review | Output |
|---|
| 1 | Full Conditional Access policy review: purpose, overlap, exclusions, expiry | Consolidated policy set |
| 2 | Break-glass account test: sign in, validate exclusion, rotate credentials, reseal | Signed test record |
| 3 | Restore test across mailbox, site, OneDrive, Teams files, Dataverse | Dated restore evidence |
| 4 | Disaster scenario tabletop (rotate scenario each quarter) | Runbook improvements |
| 5 | Security posture assessment vs CIS/benchmark; Zero Trust maturity delta | Hardening backlog |
| 6 | Purview policy review: DLP, labels, retention, IRM, information barriers | Policy updates |
| 7 | eDiscovery and legal hold reconciliation with Legal | Hold register signed off |
| 8 | Compliance Manager improvement actions and evidence refresh | Compliance score movement |
| 9 | Intune baseline and platform parity review (Windows/macOS/iOS/Android) | Gap closure plan |
| 10 | Certificate/token annual-cycle items: APNs, DEP/VPP, SCEP, SBC, federation | Renewal scheduled |
| 11 | Domain and DNS review including registrar expiry and delegation | Confirmation |
| 12 | Exchange: connectors, transport rules, DKIM rotation, SMTP AUTH inventory | Cleanup |
| 13 | SharePoint: site lifecycle, inactive archival, IA review, storage strategy | Archive/delete batch |
| 14 | Power Platform: environment strategy, DLP, orphaned assets, capacity | Governance actions |
| 15 | Copilot governance review: agents, data exposure, adoption, ROI | Renewal evidence |
| 16 | Licence and SKU right-sizing analysis ahead of renewal cycle | Commercial recommendation |
| 17 | Vendor/partner access attestation (GDAP, third-party apps, contractors) | Revocations |
| 18 | Architecture review: identity, mail flow, data flow, network changes | Updated architecture docs |
| 19 | Automation portfolio review: value delivered, debt, deprecation risk | Roadmap update |
| 20 | KTBR/KTBE time-split analysis and rebalancing | Capacity plan |
| 21 | Skills and certification progress review | Learning plan update |
| 22 | Business stakeholder service reviews (one per major function) | Demand pipeline |
C.6 Yearly (target: 2–3 weeks spread across the year)
| # | Review | Output |
|---|
| 1 | Tenant architecture review: identity model, topology, hybrid dependencies, multi-geo | Target-state architecture |
| 2 | Business continuity and DR plan review with a full-scale exercise | Tested BCP |
| 3 | RTO/RPO renegotiation with business owners per workload | Signed service expectations |
| 4 | Licensing strategy and renewal negotiation | Contract outcome |
| 5 | Security strategy: threat model refresh, Zero Trust roadmap for next 12 months | Security roadmap |
| 6 | Governance framework review: policies, standards, exception register reset | Refreshed governance pack |
| 7 | Compliance obligation refresh (new regulation, new jurisdictions) | Updated control mapping |
| 8 | Data classification scheme and label taxonomy review with business owners | Confirmed taxonomy |
| 9 | M365 roadmap alignment: what Microsoft is shipping vs what you will adopt | Adoption roadmap |
| 10 | Third-party tool portfolio vs native M365 capability review | Displacement candidates |
| 11 | Full permission and delegation recertification across all workloads | Recertification evidence |
| 12 | Insurance / regulator / audit evidence pack assembly | Annual evidence archive |
| 13 | Team structure, on-call model and succession/bus-factor review | Org plan |
| 14 | Personal certification and skill roadmap for the next year | Development plan |
| 15 | Legacy retirement plan (hybrid Exchange, ADFS, classic SPO, legacy auth apps) | Decommission schedule |
D. Priority Matrix
D.1 Definitions
| Priority | Definition | Target response | Target resolution | Escalation |
|---|
| P1 | Complete or widespread loss of a critical service, or an active/confirmed security incident, or confirmed data loss/exposure | 15 min, 24×7 | 4 hours or until contained | Immediate: management, security, Microsoft Premier/Unified severity A |
| P2 | Major degradation, a significant user population affected, or a control failure creating material risk | 1 hour, business hours (24×7 for security) | 1 business day | Team lead; Microsoft severity B |
| P3 | Normal operational issue affecting individuals or a small group; standard requests with defined SLAs | 1 business day | 3–5 business days | Standard queue |
| P4 | Low-impact requests, cosmetic issues, improvements and enhancements | 3 business days | Backlog / next cycle | Backlog grooming |
D.2 Realistic M365 examples
P1 — Critical
| Scenario | Why P1 |
|---|
| Tenant-wide sign-in failure after a Conditional Access change | Everyone is locked out; revenue-affecting |
| All inbound mail rejecting (MX, connector cert, accepted domain issue) | Total loss of the most business-critical channel |
| Confirmed Global Admin compromise or unauthorised privileged role assignment | Full tenant control potentially in attacker hands |
| Active ransomware / mass encryption or mass deletion detected in OneDrive/SPO | Data destruction in progress |
| Directory sync completely stopped for >24h with pending terminations | Terminated staff retain access; provisioning frozen |
| Confirmed exfiltration of regulated data (PII/PHI/PCI) | Statutory notification clocks start |
| Break-glass accounts non-functional discovered during an incident | No recovery path |
| Teams Phone / emergency calling failure | Life-safety exposure |
| SharePoint tenant storage exhausted | Org-wide inability to save work |
| Microsoft major service outage affecting a critical workload | Business continuity invocation |
| eDiscovery hold failure on active litigation | Legal sanction risk |
| Expired app credential taking down a revenue-critical integration | Direct business process failure |
P2 — Major
| Scenario | Why P2 |
|---|
| A department cannot access a critical SharePoint site | Significant but contained population |
| Intune policy failure leaving a device fleet non-compliant | Control failure; not yet an outage |
| High-severity Defender incident on a single endpoint, contained | Real threat, limited blast radius |
| Mail delivery delays or intermittent NDRs to a major partner domain | Visible business friction |
| MFA registration campaign failing for a user cohort | Security posture gap accumulating |
| Backup jobs failing for one workload for several days | Recovery capability degraded |
| Key business Power Automate flow failing repeatedly | Business process degraded, manual workaround exists |
| Guest access broken for an active partner project | Delivery impact on a project |
| Copilot returning sensitive content to unauthorised users | Exposure requiring immediate scope restriction |
| DLP policy misconfiguration blocking legitimate business email | Productivity blocked for a function |
| Teams Rooms fleet offline in a major office | Meetings disrupted at scale |
P3 — Normal
| Scenario |
|---|
| Individual mailbox quota, delegation or calendar permission issue |
| New shared mailbox / distribution list request |
| Individual device enrolment failure |
| Single user cannot install an Office add-in |
| Standard licence assignment request |
| Site collection storage increase request |
| Message trace investigation for a disputed delivery |
| Room booking configuration change |
| Individual OneDrive sync error |
| Standard access request for an existing group |
P4 — Low
| Scenario |
|---|
| Cosmetic branding or theme changes |
| Nice-to-have reporting requests |
| Documentation updates |
| Non-urgent feature enablement request |
| Naming convention corrections on non-critical objects |
| Requests for training material |
| Exploratory "can M365 do X?" questions |
D.3 Priority modifiers
Escalate one level when any of these apply: an executive or VIP account is involved; the issue affects a regulated data set; there is evidence of an attacker; a statutory deadline is at stake; the issue is on a change freeze day or during period-end/payroll; or no workaround exists.
Downgrade one level when: a documented workaround exists and is in use; the affected population is a pilot group that accepted the risk; or the issue is scheduled to be resolved by an already-approved change within the SLA window.
E. Automation Opportunities
E.1 Master automation table
| Manual Task | Automation Method | Technology | Trigger | Output | Business Benefit |
|---|
| User onboarding | Orchestrated provisioning runbook | Graph + Azure Automation (+ Entra inbound provisioning) | HR event / scheduled sync | User, licence, groups, mailbox, device record, welcome mail, ITSM closure | Day-one productivity; error elimination |
| User offboarding | Ordered deprovisioning runbook | Graph + Azure Automation + Power Automate approval | HR termination event | Blocked account, revoked sessions, held data, reassigned OneDrive, released licence, audit record | Closes the largest insider-risk window |
| Role change re-permissioning | Dynamic groups + access packages | Entra ID Governance | Attribute change | Access added/removed automatically | Removes access accumulation |
| Licence assignment | Group-based licensing driven by attributes | Entra ID | Group membership change | Correct SKU + disabled service plans | Consistency; no manual assignment |
| Unused licence reclamation | Inactivity detection + approval workflow | Graph reports + Power Automate | Monthly schedule | Reclaim list, approvals, removals, savings report | Recurring hard-dollar savings |
| App secret/cert expiry tracking | Scheduled inventory + tiered alerts | Graph PowerShell + Azure Automation + Teams | Weekly schedule | Expiry dashboard, owner notifications | Removes a top outage cause |
| Guest lifecycle | Access reviews + inactivity cleanup | Entra ID Governance + Graph | Monthly / on expiry | Guests removed or re-attested | Reduced external attack surface |
| Access certification | Scheduled access review campaigns | Entra Access Reviews | Monthly/quarterly | Attestation records, auto-removal of denied | Audit evidence with no manual chasing |
| Conditional Access drift detection | Config export + version control diff | Graph + Git + pipeline | Daily/weekly | Diff report, alert on unapproved change | Detects unauthorised change |
| Tenant configuration backup | Multi-workload config export | PowerShell + Graph + Git | Daily | Versioned JSON of all key config | Restorability and audit trail |
| Message center triage | API ingestion + classification + routing | Graph serviceAnnouncement + Power Automate | Daily | Work items in Planner/DevOps with owner and due date | Nothing is missed from Microsoft's change stream |
| Service health communication | Health API → internal status page/Teams | Graph serviceHealth + Power Automate | Real-time | Automated user-facing notifications | Fewer duplicate tickets during outages |
| Mailbox/site/storage reporting | Scheduled extraction to warehouse | Graph reports API + Power BI | Daily | Trend dashboards with forecasts | Capacity bought before the wall |
| Mailbox permission audit | Scheduled recursive export | EXO PowerShell | Weekly | Delegation report with delta highlighting | Detects unauthorised delegation |
| External forwarding detection | Scheduled query + alert policy | EXO PowerShell + Defender | Daily | Alert on new forwarding rule | Early BEC detection |
| Anonymous sharing link review | Sharing report extraction | SPO PowerShell / Graph / SAM ⚠️ | Weekly | Link inventory with age and content sensitivity | Controls the largest oversharing vector |
| Oversharing remediation | Detection + owner attestation flow | PnP + Power Automate | Monthly | Site owner attestation, auto-remediation of known-bad patterns | Copilot-safe data estate |
| Device compliance chasing | Automated user notification workflow | Graph + Power Automate + Teams | Daily | User-targeted remediation instructions | Compliance without service desk effort |
| Stale device cleanup | Rule-based cleanup + Entra script | Intune cleanup rules + Graph | Monthly | Removed stale objects, reconciliation report | Accurate reporting and licence counts |
| Patch/update reporting | Update report extraction | Intune / Autopatch + Power BI | Weekly | Ring status, exception list | Evidence of patch compliance |
| Phishing response | SOAR playbook | Sentinel/Defender playbooks + Logic Apps | Alert | Purge, block sender, reset user, notify, hunt | MTTR from hours to minutes |
| Account compromise response | SOAR playbook | Logic Apps + Graph | Risk alert | Revoke sessions, disable, reset, remove rules/OAuth, ticket | Containment at machine speed |
| Group/Team provisioning | Request → approve → provision | Power Apps + Power Automate + PnP | User request | Governed Team/site with template, owners, labels | Sprawl prevented at source |
| Ownerless group remediation | Ownership attestation workflow | Entra ownerless policy + Power Automate | Scheduled | Assigned owners or archived groups | Lifecycle restored |
| Access request handling | Self-service catalogue | Entra access packages / Power Apps | User request | Approved, time-bound access | Ticket deflection + audit trail |
| Password reset | SSPR with registration enforcement | Entra ID | User action | Self-served reset | Largest single ticket category removed |
| Room/resource configuration | Baseline enforcement script | EXO PowerShell | Weekly | Consistent calendar processing | Fewer booking complaints |
| DKIM rotation | Scheduled rotation cmdlet | EXO PowerShell | Quarterly | Rotated keys, validation record | Maintained email authentication posture |
| Flow/app inventory & orphan detection | CoE Starter Kit + Graph | Power Platform | Weekly | Owner reassignment tasks | Business continuity for citizen-built assets |
| Failed flow alerting | Central failure notification | Power Automate + Teams | Real-time | Failure channel with context | Silent failures become visible |
| Backup verification | Automated restore test | Backup API + scripting | Quarterly | Restore evidence report | Proven recoverability |
| Audit evidence collection | Scheduled export to immutable store | Graph + PowerShell + storage | Monthly | Dated evidence packs | Audit prep collapses from weeks to hours |
| Executive reporting | Automated data pipeline + narrative | Graph + Power BI + Power Automate | Monthly | Published report pack | Consistent visibility, no manual assembly |
| Secure Score tracking | API extraction + trend dashboard | Graph secureScore | Weekly | Trend + regression alerts | Posture managed, not sampled |
| Copilot licence utilisation | Usage extraction + reassignment workflow | Graph reports + Power Automate | Monthly | Reclaim/reassign list | Maximises expensive seats |
| Certificate/PKI monitoring | Expiry scanning across sources | Azure Automation | Weekly | Unified expiry dashboard | Prevents the classic silent outage |
E.2 Technology selection guide
| Use | Choose | Why |
|---|
| Bulk directory/mailbox operations, complex logic, admin-heavy work | PowerShell (Graph SDK, EXO v3, PnP, Teams modules) | Richest coverage, best for admin-scale operations |
| Any tenant data read/write from code or scheduled jobs | Microsoft Graph | The strategic, unified API surface |
| Business-facing workflows with approvals, forms and notifications | Power Automate | Fast to build, business-owned, native approvals |
| Scheduled, unattended, credential-safe admin automation | Azure Automation (Runbooks + managed identity) | Managed identity, module control, run history |
| Event-driven, high-volume, integration-heavy orchestration | Logic Apps / Azure Functions | Enterprise connectors, scale, retry semantics |
| Security response orchestration | Sentinel playbooks / Defender automated response | Native to the security stack, incident-triggered |
| Reporting and analytics | Power BI (+ Fabric where present ⚠️) | Historical trend, distribution, drill-down |
| Endpoint-side remediation | Intune remediation scripts | Runs where the problem is; no infrastructure |
| Conversational/self-service front end | Copilot Studio agent in Teams | Natural-language intake to the same backend automations |
Architectural rule: put the logic in Graph/PowerShell, the orchestration in Azure Automation or Logic Apps, and the human interaction in Power Automate/Power Apps/Teams. Do not build complex conditional logic inside Power Automate — it becomes unmaintainable and untestable.
E.3 Ten example automations worth building first
- Expiry Sentinel — one runbook scanning app secrets, certificates, DEP/VPP/APNs tokens, SBC certs, domain registrations and licence renewal dates into a single dashboard with 90/60/30/7-day alerts to named owners.
- JML Orchestrator — HR-triggered joiner/mover/leaver with an ordered, idempotent, fully logged sequence and an ITSM callback.
- Config Vault — nightly export of CA policies, DLP policies, transport rules, Intune profiles, SPO tenant settings, Teams policies and Power Platform DLP to Git, with diff alerting.
- Message Center Router — daily ingestion, keyword/service classification, automatic work-item creation with a due date derived from Microsoft's stated timeline.
- Licence Recovery Loop — monthly inactivity report → manager approval card in Teams → automated reclamation → savings tracked cumulatively.
- Oversharing Watchdog — weekly detection of new anonymous links and EEEU grants on sensitive-labelled sites, routed to the site owner with a one-click remediation action.
- Compromise Containment Playbook — single-trigger response: revoke sessions, disable account, reset credentials, enumerate and remove inbox rules and OAuth grants, isolate device, open incident, notify.
- Automation Registry — self-populating inventory of every flow, runbook, script and agent with owner, last-run, failure rate and business criticality.
- Restore Prover — quarterly automated test restore into an isolated target with completeness comparison and evidence PDF generation.
- Ops Reporting Pipeline — nightly Graph extraction into a warehouse, Power BI model with 13-month history, and an auto-generated monthly executive pack.
F. Senior M365 Admin Dashboard
Design principle: one screen, three zones — Red now (act today), Amber trend (act this month), Green proof (evidence for stakeholders). Every tile must have a defined threshold and an owner; a tile with no action attached should be deleted.
F.1 Identity
| KPI | Target / threshold | Source |
|---|
| Active users (enabled, licensed, signed in ≤30d) | Trend | Graph reports |
| Guest users total / active / inactive >90d | Inactive <5% | Graph |
| MFA registration coverage | 100% | Graph authentication methods |
| Phishing-resistant auth coverage (admins) | 100% | Graph |
| Phishing-resistant auth coverage (all users) | Trending to 100% | Graph |
| Risky users (high/medium) open | 0 open >24h | Identity Protection |
| Risky sign-ins last 24h | Trend + spike alert | Identity Protection |
| Global Administrators | ≤5, all PIM-eligible | Graph directoryRoles |
| Standing (permanent) privileged assignments | 0 (excluding break-glass) | PIM |
| PIM activations last 7 days with justification | 100% justified | PIM |
| Legacy authentication attempts | 0 successful | Sign-in logs |
| Accounts without manager attribute 👻 | <1% | Graph |
| Break-glass last tested | ≤90 days | Manual/record |
| CA policies in report-only >30 days 👻 | 0 | Graph |
| CA exclusion group membership count | Trend, alert on growth | Graph |
| Stale accounts (no sign-in >90d, enabled) | 0 | Graph |
| Service accounts / SPNs with credentials expiring ≤90d | 0 unowned | Graph |
F.2 Exchange Online
| KPI | Target / threshold |
|---|
| Inbound / outbound message volume (24h) | Trend + anomaly |
| Delivery failure rate / NDR spike | <baseline +20% |
| Restricted (blocked) senders | 0 |
| Mailboxes >90% quota | 0 |
| Archive mailboxes >90% quota | 0 |
| Shared mailboxes >50GB unlicensed ⚠️ | 0 |
| New external forwarding rules (7d) | 0 unreviewed |
| Quarantine volume + release requests | Trend |
| Phish/malware messages blocked (7d) | Trend |
| DMARC policy state and alignment failure rate | p=reject target |
| Connector certificate expiry ≤90d | 0 |
| Transport rules changed (30d) | All change-controlled |
| Shared mailbox count growth | Trend |
| Mailboxes on hold vs expected | Match legal register |
| SMTP AUTH-enabled mailboxes ⚠️ | Declining to 0 |
F.3 Teams
| KPI | Target / threshold |
|---|
| Active users (daily/monthly) | Trend |
| Meetings held / minutes | Trend |
| Poor-quality call rate (CQD) | <3% |
| Teams created / archived (30d) | Net growth explainable |
| Ownerless teams | 0 |
| Teams with external guests | Inventory, reviewed |
| Shared channels with external tenants | Inventory, reviewed |
| Third-party apps in use / pending approval | Approved only |
| Teams Rooms online / healthy | >98% |
| Teams Phone: numbers assigned vs purchased | Utilisation >85% |
| Emergency addresses validated | 100% |
| Meeting recordings created (30d) + retention applied | 100% covered |
F.4 SharePoint & OneDrive
| KPI | Target / threshold |
|---|
| Tenant storage used vs entitled | <80% |
| Top 10 growing sites | Reviewed monthly |
| Sites >90% quota | 0 |
| Anonymous ("Anyone") links active | Trending down; 0 on sensitive sites |
| New external sharing events (7d) | Trend + review |
| Sites with "Everyone Except External Users" | 0 on sensitive sites |
| Ownerless sites | 0 |
| Inactive sites (>180d no activity) | Archival candidates |
| OneDrive accounts >90% quota | 0 |
| KFM (Known Folder Move) enrolment coverage | >95% |
| Sync errors reported (7d) | Trend |
| Sites with sensitivity label applied | Target % by tier |
| Version storage consumed 👻 | Trend, trimming policy active |
| Orphaned OneDrive accounts (deleted users) | Managed by policy |
F.5 Intune / Endpoint
| KPI | Target / threshold |
|---|
| Device compliance rate | >95% |
| Non-compliant devices by reason | Top reasons trending down |
| Devices not checked in >30d | 0 (cleaned) |
| Encryption coverage + escrowed keys | 100% |
| Update ring compliance / patch latency | >95% within N days |
| Devices missing critical CVE patches | 0 after SLA |
| Failed policy deployments | 0 unresolved |
| Failed app installs | <2% |
| Autopilot devices awaiting enrolment | Trend |
| Certificate/token expiry ≤90d (APNs, DEP, VPP) | 0 |
| Local admin rights exceptions ⚠️ | Trending to 0 |
| macOS/mobile compliance vs Windows parity | Gap closing |
| Endpoint analytics score (startup, reliability) | Above org baseline |
F.6 Security
| KPI | Target / threshold |
|---|
| Open incidents by severity | 0 high >24h |
| Mean time to acknowledge / resolve | Within SLA |
| Alerts suppressed/tuned last 30d | Justified |
| Microsoft Secure Score + delta | Improving, no regressions |
| Exposure score / critical vulnerabilities | Trending down |
| Devices with active EDR sensor | 100% |
| ASR rules in block mode vs audit | Target coverage |
| Attack simulation click rate / report rate | Click down, report up |
| New OAuth app consents (7d) | All reviewed |
| Over-privileged Graph app permissions | Trending down |
| Sentinel ingestion volume anomaly | Within band |
| Detection rule coverage vs MITRE ATT&CK | Gap map |
| Privileged operations outside change window 👻 | 0 |
F.7 Compliance
| KPI | Target / threshold |
|---|
| DLP incidents by policy, severity and workload | Trend |
| DLP false-positive rate | <10% |
| Policy overrides / justifications used | Reviewed |
| Sensitivity label coverage (documents, sites, emails) | Target % |
| Auto-labelling match volume and accuracy | Validated |
| Retention policy coverage by workload | 100% of in-scope |
| Items pending disposition review | 0 overdue |
| Active eDiscovery cases / holds | Matches legal register |
| Audit log retention tier and search availability ⚠️ | Meets regulatory minimum |
| Compliance Manager score + improvement actions | Improving |
| Insider risk cases open / aged ⚠️ | Within SLA |
| DSAR requests open vs statutory deadline | 0 breached |
| Data residency conformance ⚠️ | 100% |
F.8 Licensing & Cost
| KPI | Target / threshold |
|---|
| Licences purchased vs assigned vs active | Assigned ≈ active |
| Unassigned licences by SKU | <3% |
| Dormant licences (no sign-in 60d) | <2% |
| Licence assignment errors | 0 |
| Cost per active user by SKU | Trend |
| Copilot seats assigned vs weekly-active ⚠️ | >70% active |
| Consumption spend (credits, capacity, PAYG) vs budget ⚠️ | Within budget, alerted at 80% |
| Storage add-on spend vs organic growth | Trend |
| Realised savings YTD from reclamation | Cumulative figure |
| Days to renewal + true-up exposure | Tracked |
| Self-service purchases active ⚠️ | 0 unapproved |
F.9 Automation & Platform Health
| KPI | Target / threshold |
|---|
| Automations registered / with named owner | 100% owned |
| Successful runs (24h/7d) | >99% |
| Failed runs by automation | 0 unresolved >24h |
| Long-running / anomalous-duration jobs | Investigated |
| Flows with expired or expiring connections | 0 |
| Orphaned flows/apps (owner departed) | 0 |
| Automations using deprecated modules/APIs ⚠️ | Trending to 0 |
| Manual hours saved (cumulative) | Reported monthly |
| Config drift detections (7d) | 0 unapproved |
| Config backup last successful | ≤24h |
F.10 Additional KPIs senior admins should monitor (often missing)
| KPI | Why it matters |
|---|
| Message center items open past Microsoft's action deadline 👻 | Directly predicts self-inflicted outages |
| Days since last successful restore test 👻 | The truest measure of recoverability |
| Days since last break-glass validation 👻 | The truest measure of recoverability of access |
| Exception register items past expiry 👻 | Measures governance decay |
| Percentage of changes made through change control | Measures operational discipline |
| Percentage of admin actions performed by an eligible (PIM) role | Measures privilege maturity |
| Ticket volume by category, with automation-deflection percentage | Measures whether KTBE is reducing KTBR |
| KTBR/KTBE actual time split | Measures whether the operating model is real |
| Documentation freshness (% of runbooks reviewed in 6 months) | Predicts incident duration |
| Single-person dependencies (bus factor) per workload | Predicts organisational risk |
| Adoption depth per licensed capability (paid vs used) | Quantifies the value gap |
| Copilot response quality / user-reported satisfaction ⚠️ | Determines renewal defensibility |
G. 30 / 60 / 90-Day Improvement Plan
G.1 Days 1–30 — Stabilise
Guiding principle: you cannot improve what you cannot see, and you cannot defend what you cannot recover. This month buys visibility and removes catastrophic single points of failure. Resist every temptation to start a Copilot or automation project.
Objectives
- Establish complete visibility of tenant configuration, identity posture and change stream.
- Eliminate the top catastrophic-risk gaps (break-glass, privileged access, credential expiry, backup).
- Establish the daily/weekly operational rhythm and its evidence trail.
- Document the tenant as it actually is, not as it was designed.
Tasks
| Week | Task |
|---|
| 1 | Inventory: tenant config export across all workloads; commit to Git as the first baseline |
| 1 | Validate break-glass accounts: exist, excluded from all CA policies, credentials known and sealed, tested sign-in, alerting on use |
| 1 | Privileged role inventory: list every admin role holder in Entra, EXO, SPO, Teams, Intune, Purview, Power Platform |
| 1 | Confirm Message center and service health are being read daily by a named person |
| 1 | Establish the daily checklist and start recording completion |
| 2 | App registration credential inventory with expiry dates; alert anything <90 days |
| 2 | Identify and document all critical integrations and their identities |
| 2 | Verify backup coverage across all workloads; identify what is unprotected |
| 2 | MFA / phishing-resistant auth coverage report; identify unregistered and admin gaps |
| 2 | Legacy authentication usage report; list applications still using it |
| 3 | Conditional Access policy documentation: purpose, scope, exclusions, owner for every policy |
| 3 | External sharing posture report: anonymous links, guest count, EEEU grants |
| 3 | Licence position report: purchased vs assigned vs active |
| 3 | Set up centralised failure alerting for automations and scheduled jobs |
| 3 | Establish the change control process for tenant-level changes |
| 4 | Perform one restore test (mailbox + site) and document the result |
| 4 | Draft the tenant architecture document (identity, mail flow, data flow, dependencies) |
| 4 | Identify and log the top 10 risks with owners and target dates |
| 4 | Establish the monthly executive report format and produce the first one |
Deliverables
Tenant configuration baseline in version control · Privileged access inventory · Credential expiry register · Backup coverage matrix · Documented CA policy register · Risk register with owners · Daily/weekly operational checklists in use · Architecture document v1 · First monthly report.
KPIs
| KPI | Target by day 30 |
|---|
| Break-glass accounts validated | Yes, documented |
| Global Admin count | Known, reduced to ≤5 where feasible |
| Credentials expiring <90 days with no owner | 0 |
| Workloads with confirmed backup coverage | 100% known (not necessarily 100% protected) |
| Config baseline committed | Yes, automated nightly |
| Daily checklist completion | >95% |
| Documented CA policies | 100% |
Expected business outcome
The organisation moves from "we hope it is fine" to "we know where we stand." No new capability, but the probability of a catastrophic, preventable outage drops sharply, and every subsequent decision now has evidence behind it.
G.2 Days 31–60 — Optimise
Guiding principle: convert visibility into control, and convert repetitive work into automation. Every hour invested here returns hours every month thereafter.
Objectives
- Close the high-risk findings from month one.
- Automate the highest-volume recurring KTBR tasks.
- Establish governance that prevents new debt from accumulating.
- Deliver the first hard-dollar saving.
Tasks
| Week | Task |
|---|
| 5 | Implement PIM for all privileged roles; remove standing assignments |
| 5 | Block legacy authentication (staged, with exceptions registered and expiry-dated) |
| 5 | Build the Expiry Sentinel automation (secrets, certs, tokens, domains) |
| 5 | Implement CA drift detection with alerting |
| 6 | Build or complete JML automation — start with leaver (highest risk), then joiner |
| 6 | Implement group-based licensing with attribute-driven dynamic groups |
| 6 | Deploy automated licence reclamation with manager approval |
| 6 | Implement access reviews for privileged roles and guests |
| 7 | Message center → backlog automation |
| 7 | Consolidate and rationalise Conditional Access into a persona-based model (report-only first) |
| 7 | Establish Power Platform environment strategy and DLP policies |
| 7 | Remediate anonymous links and EEEU grants on sensitive sites |
| 8 | Build the automated reporting pipeline (Graph → store → Power BI) |
| 8 | Complete backup coverage gaps; automate backup failure alerting |
| 8 | Run quarterly-equivalent restore test across additional workloads |
| 8 | Implement self-service password reset coverage to 100% and measure ticket deflection |
| 8 | Tune DLP and Defender alert policies to reduce noise |
Deliverables
PIM-enforced privileged access · Legacy auth blocked · Expiry monitoring service · Automated JML · Group-based licensing · Licence reclamation workflow with quantified savings · Access review campaigns running · Persona-based CA model · Power Platform DLP · Automated reporting pipeline · Complete backup coverage.
KPIs
| KPI | Target by day 60 |
|---|
| Standing privileged assignments | 0 (excl. break-glass) |
| Successful legacy auth sign-ins | 0 |
| Manual JML effort | Reduced >70% |
| Licence savings identified and realised | Quantified figure reported |
| Credentials expiring without alert coverage | 0 |
| Access review completion rate | >95% |
| Automated report delivery | Monthly, unattended |
| Ticket volume in top 3 automated categories | Down >30% |
Expected business outcome
Recurring administrative toil falls measurably, the first documented cost saving lands on a finance report, and the security posture improves in ways that are provable to an auditor. Capacity is freed for phase three.
G.3 Days 61–90 — Expand
Guiding principle: now spend the capacity you created. Move from operating the platform to increasing what the platform delivers.
Objectives
- Deliver AI capability safely on top of a governed data estate.
- Push routine operations to self-service and lower support tiers.
- Modernise the areas creating the most future KTBR load.
- Establish the ongoing KTBE engine so improvement continues past day 90.
Tasks
| Week | Task |
|---|
| 9 | Copilot readiness assessment: oversharing remediation status, label coverage, search quality, restricted discovery configuration |
| 9 | Define AI/agent governance framework: intake, review, environments, DLP, ownership, decommission |
| 9 | Establish the pilot ring programme for Microsoft feature changes |
| 10 | Build the self-service request portal (groups, licences, access, mailboxes) with approvals |
| 10 | Define the tiered support model and delegate L1 operations with least-privilege roles |
| 10 | Launch the Copilot pilot cohort with scenario packs and measurement baseline |
| 11 | Build the first Copilot Studio agent for a high-volume internal query type |
| 11 | Begin endpoint modernisation: Autopatch or GPO→Intune migration, whichever is the larger source of toil |
| 11 | Implement sensitivity label auto-labelling in simulation mode |
| 12 | Deliver the third-party tool displacement analysis (what E5/Intune Suite/Purview already covers) |
| 12 | Publish the 12-month platform roadmap with KTBR/KTBE split and business alignment |
| 12 | Establish quarterly business service reviews with functional stakeholders |
| 12 | Conduct a tabletop exercise on tenant compromise; feed results into the backlog |
Deliverables
Copilot readiness report and remediation plan · AI/agent governance framework · Self-service portal in production · Tiered support model with delegated roles · First production agent · Endpoint modernisation underway · Auto-labelling simulation results · Tool displacement business case · 12-month roadmap · Stakeholder review cadence.
KPIs
| KPI | Target by day 90 |
|---|
| Sensitive sites with unresolved broad access | 0 |
| Self-service request deflection | >30% of eligible tickets |
| Copilot pilot weekly active usage | >70% of pilot seats |
| Agents in production with governance sign-off | ≥1, 100% governed |
| KTBE share of admin time | ≥35% |
| Documented savings (licence + tool displacement) | Quantified annualised figure |
| Restore tests completed | All critical workloads |
| P1 incidents caused by preventable config/expiry issues | 0 |
Expected business outcome
The tenant is now a governed platform rather than a collection of settings. AI capability is being delivered on a data estate that can safely support it, routine work is self-served, and the platform team has a visible roadmap and a seat at the business table. Critically, the improvement mechanism itself — backlog, roadmap, stakeholder cadence, measurement — is now permanent rather than a one-off project.
H. Top 50 KTBR Tasks (ranked by importance)
| # | Task | Why it is at this rank |
|---|
| 1 | Break-glass emergency access validation | Without it, every other recovery action may be impossible. It is the control that protects your ability to use all other controls. |
| 2 | Conditional Access change control and drift detection | The most powerful and most dangerous setting in the tenant; both a top security control and the top self-inflicted outage cause. |
| 3 | Privileged role management (PIM, GA minimisation, activation audit) | Compromise of one standing Global Admin is functionally equivalent to compromise of the entire organisation's data. |
| 4 | Leaver deprovisioning with session revocation | Terminated access is the most consistently exploited and most consistently audited gap. |
| 5 | Application credential and certificate expiry management | The most common source of preventable, unannounced, business-critical outages. |
| 6 | Backup coverage and proven restore capability | Retention is not backup; Microsoft's SLA does not cover your deletions. Recoverability must be proven, not assumed. |
| 7 | MFA / phishing-resistant authentication coverage | Credential attacks remain the dominant initial access vector; coverage gaps are where attackers enter. |
| 8 | Legacy authentication blocking | Legacy protocols bypass Conditional Access and MFA entirely, nullifying your other identity controls. |
| 9 | Defender XDR incident triage and response | Detection without disciplined triage produces alerts, not security. |
| 10 | Message center triage and change impact assessment | Microsoft changes your tenant on its schedule; missing this converts their change into your outage. |
| 11 | Service health monitoring and user communication | Determines whether an outage is a managed event or a crisis of confidence. |
| 12 | Mail flow monitoring (queues, NDRs, connectors) | Mail failure is the most visible and least tolerated outage in most organisations. |
| 13 | Graph application permission and consent review | Application permissions are tenant-wide keys that survive password resets and MFA. |
| 14 | External sharing and anonymous link control | The dominant data-exposure mechanism in M365, and the one Copilot makes visible. |
| 15 | Directory synchronisation health (hybrid) | Silent failure freezes all identity change, including terminations. |
| 16 | Device compliance monitoring and remediation | Compliance is the assumption underpinning every device-based Conditional Access rule. |
| 17 | Patch and update ring management | Unpatched endpoints remain the primary ransomware entry point. |
| 18 | Risky user and risky sign-in triage | The earliest reliable signal of account takeover in progress. |
| 19 | Configuration drift detection and baseline management | Without it, you cannot distinguish an authorised change from an attacker's change. |
| 20 | External forwarding and inbox rule auditing | The signature persistence technique of business email compromise. |
| 21 | Audit log ingestion, retention and search validation | No logs, no investigation, no evidence, no defensible position. |
| 22 | Conditional Access exclusion auditing | Exclusions are added under pressure and become permanent, undocumented backdoors. |
| 23 | Access reviews for privileged roles, guests and high-value groups | The mechanism that reverses inevitable access accumulation. |
| 24 | DLP policy tuning and incident triage | Untuned DLP is ignored, and ignored DLP protects nothing. |
| 25 | Retention and legal hold coverage validation | Both over-retention and under-retention create legal and financial exposure. |
| 26 | Domain, DNS and MX record management | Trivial to maintain, catastrophic to lose. |
| 27 | SPF/DKIM/DMARC posture | Determines whether your domain can be spoofed and whether your mail is accepted. |
| 28 | Enterprise application and service principal review | Dormant and over-permissioned service principals are prime persistence targets. |
| 29 | Storage and capacity monitoring across SPO, EXO, Dataverse | Capacity exhaustion is a hard stop with no graceful degradation. |
| 30 | Automation and scheduled job failure monitoring | Silent automation failure creates compliance and provisioning gaps nobody notices. |
| 31 | Guest and B2B lifecycle management | External identities accumulate indefinitely and retain access to live data. |
| 32 | Intune certificate and token expiry (APNs, DEP, VPP, SCEP) | Annual-cycle items that disconnect entire device fleets when missed. |
| 33 | Encryption coverage and key escrow verification | An unescrowed key means an unrecoverable device and unrecoverable data. |
| 34 | Shared mailbox and delegation auditing | Undocumented delegation is a persistent, low-visibility data access path. |
| 35 | Power Platform DLP and orphaned asset management | Business-critical citizen-built processes fail when their owner leaves. |
| 36 | Copilot data-readiness and oversharing remediation | Copilot converts historical permission mistakes into present-day incidents. |
| 37 | Change management and approval discipline | Most self-inflicted outages trace back to an unreviewed change. |
| 38 | Secure Score review and improvement backlog | The most usable structured posture metric available natively. |
| 39 | Anti-phishing, Safe Links and Safe Attachments policy integrity | Bypass rules and allow-list entries silently erode protection over time. |
| 40 | Tenant Allow/Block List hygiene | Permanent allow entries are actively targeted by attackers. |
| 41 | Teams external access, guest and shared channel governance | Modern collaboration paths that bypass classic guest controls. |
| 42 | Group and Teams lifecycle (ownerless, dormant, expiration) | Ungoverned containers become ungoverned data stores. |
| 43 | eDiscovery and legal hold execution | Direct legal exposure with judicial consequences for failure. |
| 44 | Licence assignment reconciliation and error remediation | Assignment errors leave users unlicensed and create silent service loss. |
| 45 | Alert tuning and noise reduction | Alert fatigue, not alert absence, is the usual reason incidents are missed. |
| 46 | Documentation and runbook maintenance | Directly determines incident duration and the viability of holiday cover. |
| 47 | Post-incident review and remediation tracking | The main mechanism converting failures into permanent improvement. |
| 48 | Partner/GDAP and vendor access attestation | Delegated partner access has repeatedly been the path into tenants. |
| 49 | Exception register with expiry enforcement | The mechanism that stops temporary risk acceptance becoming permanent risk. |
| 50 | Tenant architecture documentation currency | Required for incidents, audits, M&A, onboarding and every major decision. |
I. Top 50 KTBE Tasks (ranked by business value)
| # | Initiative | Business value rationale |
|---|
| 1 | Joiner–Mover–Leaver automation | Removes the largest recurring manual workload while simultaneously closing the largest security gap. Value on both axes at once. |
| 2 | Licence optimisation engine (baseline + continuous reclamation) | Fastest hard-dollar ROI available; recurring, provable, and immediately credible with finance. |
| 3 | Third-party tool displacement using owned E5/Intune Suite/Purview capability | Frequently the single largest documented saving in the entire portfolio, often six figures annually. |
| 4 | Configuration-as-code with drift detection | Converts the tenant into a reviewable, restorable, auditable asset; underpins every other improvement. |
| 5 | Copilot readiness (data governance and oversharing remediation) | Determines whether a very large AI investment produces value or a data incident. |
| 6 | Self-service portal for routine requests | Deflects a third or more of the ticket queue permanently; visible to every employee. |
| 7 | Entra ID Governance: access packages and entitlement management | Turns access administration into a self-service, auditable, time-bound process. |
| 8 | Security automation / SOAR playbooks | Cuts mean time to respond from hours to minutes on the incidents that matter most. |
| 9 | Passwordless / phishing-resistant authentication programme | Removes the dominant attack vector while eliminating the largest ticket category. |
| 10 | Copilot adoption programme (champions, scenarios, measurement) | Copilot value is adoption-limited; without this the licence spend is largely wasted. |
| 11 | Executive reporting pipeline | Converts invisible platform work into funded, understood, valued work. |
| 12 | File share to SharePoint/OneDrive migration with permission redesign | Removes on-prem cost and unlocks Copilot grounding on the organisation's real knowledge. |
| 13 | GPO to Intune modernisation | Removes domain dependency, enables true remote management, retires legacy infrastructure. |
| 14 | Local admin rights removal (Endpoint Privilege Management) ⚠️ | One of the highest-impact endpoint security controls available. |
| 15 | Zero standing privilege via full PIM adoption | Structurally limits the damage any single compromised admin account can cause. |
| 16 | Unified expiry monitoring service | Eliminates an entire class of preventable outages permanently. |
| 17 | Message center to backlog automation | Turns Microsoft's change stream from a risk into a managed pipeline. |
| 18 | Windows Autopatch adoption ⚠️ | Removes a recurring high-effort orchestration burden while improving patch compliance. |
| 19 | Sensitivity label taxonomy and auto-labelling | Protection that travels with data, applied at scale without relying on user discipline. |
| 20 | Tiered support model with delegated least-privilege roles | Frees senior capacity — the binding constraint on all other KTBE work. |
| 21 | Copilot Studio agents for high-volume internal queries | Deflects repetitive queries at low marginal cost; the clearest early agent use case. |
| 22 | Agent and AI governance framework ⚠️ | Prevents the next generation of shadow IT before it becomes unmanageable. |
| 23 | Automated reporting and analytics pipeline | Ends manual report assembly permanently and enables trend-based decisions. |
| 24 | SKU rationalisation ahead of renewal | Contract-level decisions worth far more than any operational efficiency. |
| 25 | Conditional Access persona-based consolidation | Fewer, clearer policies means fewer gaps, fewer lockouts and faster change. |
| 26 | Graph connectors for external content indexing | Transforms search and Copilot usefulness by including non-M365 knowledge. |
| 27 | Endpoint DLP and browser DLP expansion ⚠️ | Closes the exfiltration paths that email-only DLP leaves wide open. |
| 28 | Governed provisioning for Teams/sites/groups | Prevents at source the sprawl that otherwise requires perpetual remediation. |
| 29 | Intune Suite capability adoption ⚠️ | Replaces multiple third-party tools with already-licensable capability. |
| 30 | Automated evidence collection for audit | Turns audit season from weeks of manual work into an export. |
| 31 | SharePoint intranet and information architecture modernisation | Findability improvements compound across every knowledge-work task and improve Copilot. |
| 32 | Insider risk programme with HR integration ⚠️ | Detects pre-departure data exfiltration, a risk with no other reliable control. |
| 33 | Teams Phone migration from legacy PBX ⚠️ | Substantial, quantifiable telephony cost reduction. |
| 34 | Dynamic group strategy replacing manual membership | Removes a large, error-prone recurring workload and improves access accuracy. |
| 35 | Records management and defensible disposal programme | Reduces storage cost, eDiscovery burden and legal exposure simultaneously. |
| 36 | Windows 365 / Cloud PC for contractors, BYOD and BCP ⚠️ | Secure access without hardware logistics; also a genuine continuity capability. |
| 37 | Event-driven automation via Graph change notifications | Moves automation from scheduled polling to real-time response. |
| 38 | Automated licence reclamation with approval workflow | Recurring savings that require no ongoing human effort once built. |
| 39 | Non-production lab tenant | Enables safe testing; prevents production from being the test environment. |
| 40 | Workload identity federation replacing app secrets | Removes both an outage class and a credential-theft class permanently. |
| 41 | Data minimisation and stale content disposal | Reduces cost and improves Copilot precision by removing noise. |
| 42 | External collaboration model design | Enables partner work safely; prevents shadow IT collaboration tools. |
| 43 | Continuous compliance scanning against benchmarks | Provides an external, objective yardstick and audit-ready evidence. |
| 44 | Frontline worker enablement ⚠️ | Extends value to the largest under-served population in many organisations. |
| 45 | ChatOps for delegated admin operations | Fast, safe, fully logged operations without portal access. |
| 46 | Platform team product-ownership model | Ends the "everyone owns everything" failure mode that caps team throughput. |
| 47 | Chargeback / showback model | Changes consumption behaviour more effectively than any written policy. |
| 48 | macOS and mobile management parity ⚠️ | Closes the most common and most persistent device governance gap. |
| 49 | Meeting room and hybrid meeting modernisation ⚠️ | Highly visible experience improvement with measurable satisfaction impact. |
| 50 | Business relationship management cadence | Surfaces demand early, builds the political capital that funds everything else. |
J. What a Senior M365 Admin Should Automate
Level 1 — Automate Immediately (high volume, low complexity, high certainty)
These are deterministic, well-bounded and repeated constantly. Build them first; they pay back within weeks.
1.1 Credential and Token Expiry Monitoring
Architecture: Azure Automation runbook on a daily schedule, authenticating with a managed identity holding Application.Read.All and Directory.Read.All. It queries Graph for applications and servicePrincipals, extracting passwordCredentials and keyCredentials expiry dates, joins to an owner registry stored in a SharePoint list or Dataverse table, and writes results to a table consumed by Power BI. A companion Power Automate flow reads the table daily and posts adaptive cards to owners at 90/60/30/14/7 days, escalating to the platform team at 14 days. Extend the same job to Intune tokens (APNs, DEP, VPP) via deviceManagement endpoints, domain registrar expiry via a WHOIS API, and SBC certificates via a monitored endpoint check.
Why this shape: a single job, a single data table, a single dashboard. Splitting expiry monitoring per technology is the reason most organisations only cover half of it.
1.2 Leaver Deprovisioning
Architecture: HR system event (or a scheduled delta query against the HR source) triggers a Logic App, which invokes an Azure Automation runbook. The runbook executes a strictly ordered, idempotent sequence: block sign-in → revoke all refresh tokens → remove strong authentication methods → apply retention/litigation hold → convert mailbox and grant manager delegate access → transfer OneDrive ownership → remove from all groups and Teams → wipe or retire devices → schedule licence removal for T+N days. Every step writes a structured log entry with timestamp and result. A Power Automate flow raises an ITSM record and notifies the manager. Failure at any step raises an alert rather than silently continuing.
Why this shape: the ordering is the control. Runbooks give you managed identity, run history and module pinning; Power Automate gives you the human notification and approval layer. Do not attempt the whole sequence in Power Automate — the error handling and idempotency requirements exceed what it does well.
1.3 Joiner Provisioning
Architecture: Entra inbound provisioning from the HR source where the HR system is supported, otherwise a scheduled Graph runbook consuming an HR export. Create the account with standardised attributes (department, manager, location, employee ID, cost centre), let attribute-driven dynamic groups handle group membership, let group-based licensing handle licences, and let Lifecycle Workflows handle the timed pre-hire and day-one tasks. The runbook's only job is accurate attribute population — everything downstream should be attribute-driven.
Why this shape: hardcoding group and licence assignment into the joiner script is the most common design error. It makes every subsequent organisational change a code change. Get the attributes right and the rest becomes declarative.
1.4 Licence Reclamation
Architecture: Monthly Azure Automation job pulls getOffice365ActiveUserDetail and sign-in activity from Graph, cross-references subscribedSkus and per-user assignment, and produces a candidate list filtered by exclusion rules (leave of absence, service accounts, VIPs). Power Automate posts approval cards to each candidate's manager with a 5-day response window; approved reclamations execute via Graph, denied ones are logged with a re-review date, and non-responses default to reclamation with a 14-day licence-restore grace. A running total of realised savings feeds the monthly executive report.
1.5 Configuration Backup and Drift Detection
Architecture: Nightly Azure Automation job exports Conditional Access policies, authentication methods policies, named locations, Intune configuration and compliance policies, Purview DLP and retention policies, Exchange transport rules and connectors, SPO tenant settings, Teams policies and Power Platform DLP policies to normalised JSON. A pipeline commits to a Git repository with a service-principal identity. A post-commit check diffs against the previous commit and raises an alert into the change channel for any change not linked to an approved change record. The repository doubles as the restoration source.
1.6 Service Health and Message Center Distribution
Architecture: Power Automate or Logic App polls serviceAnnouncement/healthOverviews and serviceAnnouncement/messages every 15 minutes and daily respectively. Health issues affecting subscribed services post to the operations channel and update an internal status page. Message center posts are classified by service and keyword, and a work item is created in Planner or Azure DevOps with a due date derived from Microsoft's stated action deadline, assigned to the workload owner.
1.7 Other Level 1 candidates
| Task | Shape |
|---|
| Failed flow / runbook alerting | Central webhook → Teams channel with context and owner tag |
| Device compliance nudging | Graph query → Power Automate → adaptive card with self-remediation steps |
| Stale object cleanup | Scheduled Graph job with soft-delete-first policy |
| External forwarding detection | Daily EXO query → security channel |
| Mailbox/site quota alerting | Scheduled threshold check → owner notification |
| DKIM rotation | Quarterly scheduled cmdlet execution with validation |
| Room mailbox baseline enforcement | Weekly desired-state enforcement script |
| Graph webhook subscription renewal | Scheduled renewal job before expiry |
Level 2 — Automate Next (moderate complexity, requires design and governance)
2.1 Self-Service Request Portal
Architecture: Power Apps canvas or model-driven front end presenting a request catalogue (group creation, membership, licence request, shared mailbox, distribution list, site creation, access to a resource). Requests write to Dataverse. Power Automate handles approval routing based on request type and requester attributes. Execution occurs through a child flow or Azure Automation webhook using a dedicated service principal with narrowly scoped permissions — never the requester's credentials and never a Global Admin identity. Every request, approval and execution result is recorded in Dataverse for audit. Requests that fail validation return to the requester with a specific reason rather than a generic error.
Governance requirement: the executing identity must have the minimum permission needed per request type. Consider separate service principals per capability rather than one over-privileged automation identity. This is the difference between a self-service portal and a privilege-escalation vector.
2.2 Access Governance Automation
Architecture: Entra ID Governance access packages defined per business role, grouped into catalogues owned by business units. Policies define who can request, who approves, how long access lasts and what happens on expiry. Lifecycle workflows handle joiner/mover/leaver triggers. Scheduled access reviews cover privileged roles monthly, guests quarterly and high-value groups quarterly, with auto-removal on non-response for lower-risk populations and manual decision required for privileged ones. Review outcomes are exported monthly to the audit evidence store.
2.3 Security Response Playbooks
Architecture: Sentinel analytic rule or Defender XDR incident triggers a Logic App playbook. For account compromise: revoke sessions, disable the account, force credential reset, enumerate and export inbox rules and OAuth grants, remove attacker-created artefacts, isolate associated devices via Defender for Endpoint, purge related mail tenant-wide, create the incident record, and notify the SOC channel with a structured summary. Human approval gates should sit before destructive or business-impacting actions (account disable for an executive, device isolation for a production server) while allowing fully automatic execution of reversible containment steps.
Design principle: automate containment, gate eradication. The cost of an automated false-positive containment is minutes; the cost of an automated false-positive deletion is unbounded.
2.4 Governed Provisioning for Collaboration Workspaces
Architecture: Request through the self-service portal captures purpose, data sensitivity, owners (minimum two), external collaboration requirement and expected lifespan. Approval routes by sensitivity. Provisioning executes a PnP template applying site structure, sensitivity label, sharing configuration, retention, permissions and metadata. The workspace is registered in a governance table with a review date. Scheduled jobs enforce lifecycle: owner attestation at 6 months, inactivity detection at 12 months, archival recommendation thereafter.
2.5 Reporting and Analytics Pipeline
Architecture: Nightly Azure Automation or Azure Function extracts from Graph reports APIs, Defender APIs, Intune reporting, Purview, Power Platform admin APIs and commerce APIs. Data lands in Azure Storage or a SQL/Fabric target with historical retention beyond Microsoft's native reporting windows (which are short — typically 30 to 180 days ⚠️). Power BI models the data with 13+ months of history for trend and seasonality. Power Automate distributes a monthly PDF pack and posts weekly deltas to the operations channel.
Key design point: the value is in the history. Microsoft's native reports have short retention windows; your competitive advantage as an administrator is having thirteen months of trend when a question is asked.
2.6 Compliance Evidence Automation
Architecture: Monthly scheduled exports of access review outcomes, privileged role assignments, CA policy state, DLP incident summaries, retention policy coverage, backup test results, patch compliance and audit log search confirmations. Output written to immutable or WORM-configured storage with a defined retention period. Each export is timestamped, hashed and indexed against the specific control it evidences in your compliance framework.
2.7 Other Level 2 candidates
| Task | Shape |
|---|
| Oversharing remediation workflow | Detection → owner attestation card → one-click remediation → verification |
| Automated onboarding of new automations into the registry | Deployment pipeline writes metadata to the registry automatically |
| Capacity forecasting | Historical trend model with threshold projection and procurement lead-time alerting |
| Vulnerability-to-remediation pipeline | Defender exposure data → Intune remediation deployment → verification |
| Copilot seat rotation | Usage detection → reassignment workflow → cohort management |
| Change record correlation | Match every audit-log config change to an approved change record; alert on unmatched |
Level 3 — Strategic Automation (cross-platform, enterprise-scale, architecturally significant)
3.1 Full Identity Lifecycle Platform
Architecture: HR system as authoritative source, integrated via Entra inbound provisioning or a custom connector into a normalised identity data layer. That layer drives: Entra ID account state, attribute-driven dynamic group membership, group-based licensing, Entra ID Governance access packages, Intune device assignment, application entitlement provisioning (SCIM to SaaS applications) and downstream on-premises AD where hybrid remains. Exceptions route to a governed approval process rather than manual admin action. Reconciliation runs daily comparing HR truth to tenant state, alerting on any divergence.
What makes it Level 3: it crosses HR, identity, licensing, endpoint and third-party SaaS boundaries, and it requires organisational agreement about data ownership and attribute authority — which is harder than the technical build.
3.2 Tenant-as-Code Platform
Architecture: All tenant configuration expressed declaratively in a repository — Conditional Access, Intune policies, Purview policies, Exchange configuration, SPO settings, Teams policies, Power Platform DLP. Changes proposed by pull request with mandatory review. CI validates syntax and runs impact analysis (for example, CA What-If evaluation against a representative user set). CD deploys to a development tenant first, then to production on approval. Continuous reconciliation detects and optionally auto-corrects drift. The repository is simultaneously the change record, the audit evidence, the documentation and the disaster recovery source.
What makes it Level 3: it requires a non-production tenant, engineering discipline, pipeline infrastructure and a cultural shift from portal-clicking to code review. It is also the single highest-leverage investment a mature M365 team can make.
3.3 Unified Observability and Automated Response Platform
Architecture: All signal sources — Entra sign-in and audit logs, unified audit log via the Management Activity API, Defender XDR, Intune, Power Platform, Graph activity logs, custom synthetic transactions, automation telemetry — flow into Log Analytics/Sentinel. A curated set of analytic rules and workbooks provides operational and security views. Automated response playbooks handle defined scenarios. Synthetic transaction runbooks continuously validate real user journeys (send and receive mail, upload and download a file, join a Teams meeting, authenticate to a key application) from multiple network locations, feeding a tenant-specific health signal independent of Microsoft's own dashboard.
What makes it Level 3: it requires log architecture decisions with real cost implications, KQL engineering capability, and integration across the security and operations boundary.
3.4 Agentic Operations Layer
Architecture: A governed set of Copilot Studio agents and AI-assisted workflows sitting on top of the automation estate. A service desk agent handles tier-zero queries by querying documented knowledge and, where authorised, invoking narrowly scoped automation actions through connectors with per-action authorisation. An operations assistant summarises incident context, correlates signals and drafts communications. Every agent has a registered owner, a defined data scope, DLP policy coverage, logged interactions, an approval gate before any write action, and a scheduled review with a decommission date if unused.
What makes it Level 3: the governance is harder than the build. Agents inherit their creator's or connection's data access, they are non-deterministic, and they are the emerging shadow-IT frontier. Build the governance framework before the first production agent, not after the fifth.
3.5 Cross-Platform Business Process Automation
Architecture: Business processes spanning M365, ERP, CRM, ITSM and line-of-business systems, orchestrated through Logic Apps or Power Automate with Dataverse as the state store. M365 provides the human interaction surface (Teams approvals, Outlook actionable messages, Power Apps forms, SharePoint document management) while enterprise integration handles system-to-system reliability. Standard patterns for error handling, retry, dead-lettering, idempotency and observability apply across every process.
What makes it Level 3: the administrator is now delivering business capability, not IT capability. This is where the platform role converts fully into a business-value role — and where the M365 administrator's career trajectory changes.
3.6 Multi-Tenant / M&A Readiness Capability
Architecture: Reusable tooling and documented patterns for tenant assessment, cross-tenant identity synchronisation, B2B and multi-tenant organisation configuration, staged workload migration (mail, files, Teams, devices), coexistence during transition, and eventual consolidation. Configuration-as-code makes the target state reproducible; the automation estate makes the migration executable at scale.
What makes it Level 3: it is capability held in readiness rather than in production, and it is the difference between an M&A being an eighteen-month scramble and a managed programme.
K. Senior M365 Administrator Skill Roadmap
Classification: Must Know = you cannot do the job without it · Should Know = expected of a senior individual contributor · Advanced = differentiates you within the M365 community · Expert = you are the person others escalate to, and you shape architecture.
K.1 Microsoft Entra ID / Identity
| Skill | Level |
|---|
| Users, groups, licensing, directory roles, administrative units | Must Know |
| Authentication methods, MFA, SSPR, registration campaigns | Must Know |
| Conditional Access design, What-If, report-only, break-glass patterns | Must Know |
| Sign-in and audit log interpretation, KQL over identity logs | Must Know |
| Hybrid identity: Connect / Cloud Sync, PHS/PTA, staging mode, sync rules | Should Know |
| B2B, B2B direct connect, cross-tenant access, multi-tenant organisation | Should Know |
| PIM design: role settings, approval, alerting, zero standing privilege | Advanced |
| Identity Protection risk policies and remediation design | Advanced |
| Entra ID Governance: entitlement management, access packages, lifecycle workflows | Advanced |
| Application registration, SSO (SAML/OIDC), SCIM provisioning to SaaS | Advanced |
| Workload identities, managed identities, federated credentials | Advanced |
| Token protection, CAE, tenant restrictions, persona-based CA architecture | Expert |
| Identity architecture for M&A, multi-geo, sovereign and regulated scenarios | Expert |
K.2 Exchange Online
| Skill | Level |
|---|
| Mailbox types, permissions, delegation, quotas, archives | Must Know |
| Message trace, mail flow troubleshooting, NDR interpretation | Must Know |
| Exchange Online PowerShell (v3) and RBAC | Must Know |
| SPF, DKIM, DMARC design and DMARC enforcement progression | Must Know |
| Connectors, transport rules, accepted domains, hybrid mail flow | Should Know |
| Retention, holds, inactive mailboxes, journaling | Should Know |
| Anti-spam/anti-phishing policy architecture and preset security policies | Advanced |
| Migration and coexistence, hybrid decommissioning | Advanced |
| Mail flow architecture for regulated and complex routing scenarios | Expert |
K.3 Microsoft Teams
| Skill | Level |
|---|
| Teams, channels, membership, guest and external access | Must Know |
| Teams admin center policy model and PowerShell | Must Know |
| Meeting, messaging, app and calling policy design | Should Know |
| Teams and M365 Group lifecycle governance | Should Know |
| Call Quality Dashboard and network readiness analysis | Advanced |
| Teams Phone, Direct Routing, Operator Connect, E911 ⚠️ | Advanced |
| Teams Rooms fleet management and monitoring ⚠️ | Advanced |
| Voice architecture and enterprise migration from legacy PBX | Expert |
K.4 SharePoint Online & OneDrive
| Skill | Level |
|---|
| Sites, libraries, permissions, sharing, versioning | Must Know |
| SPO PowerShell and PnP PowerShell | Must Know |
| Tenant and site-level sharing controls | Must Know |
| Information architecture, hubs, navigation, search configuration | Should Know |
| Storage management, quotas, archival, version trimming | Should Know |
| Data Access Governance, Restricted Access Control, Restricted Content Discovery ⚠️ | Advanced |
| PnP provisioning templates and site lifecycle automation | Advanced |
| Migration at scale, permission redesign, Copilot-ready data architecture | Expert |
| SPFx and custom solution governance | Advanced |
K.5 Microsoft Intune / Endpoint
| Skill | Level |
|---|
| Enrolment, compliance policies, configuration profiles, app deployment | Must Know |
| Windows Autopilot and Enrollment Status Page | Must Know |
| Update rings, feature update policies, patch reporting | Must Know |
| Settings catalogue, ADMX, security baselines | Should Know |
| Conditional Access integration and device-based access | Should Know |
| Certificate infrastructure: SCEP/PKCS/Cloud PKI ⚠️ | Advanced |
| macOS, iOS and Android management parity | Advanced |
| Intune Suite: EPM, Enterprise App Management, Advanced Analytics ⚠️ | Advanced |
| Endpoint modernisation architecture (GPO retirement, Entra-only join) | Expert |
K.6 Microsoft Defender
| Skill | Level |
|---|
| Defender XDR incident model and triage workflow | Must Know |
| Defender for Office 365: policies, preset security policies, quarantine, TABL | Must Know |
| Secure Score interpretation and improvement planning | Must Know |
| Defender for Endpoint: onboarding, ASR, device isolation, live response | Should Know |
| Defender for Cloud Apps: app discovery, app governance, session policies | Advanced |
| Defender for Identity: sensors, on-prem attack path analysis | Advanced |
| Advanced hunting with KQL and custom detection rules | Advanced |
| Automated investigation, attack disruption and response design | Advanced |
| Threat modelling and detection engineering for M365 attack paths | Expert |
K.7 Microsoft Purview
| Skill | Level |
|---|
| Unified audit log search and retention model ⚠️ | Must Know |
| Retention policies and labels, and their precedence rules | Must Know |
| DLP policy structure, conditions, actions and tuning | Must Know |
| Sensitivity labels, encryption, container labels | Should Know |
| eDiscovery (Standard and Premium) and legal hold ⚠️ | Should Know |
| Auto-labelling, trainable classifiers, exact data match | Advanced |
| Insider Risk Management and Communication Compliance ⚠️ | Advanced |
| Information barriers, records management, disposition ⚠️ | Advanced |
| DSPM for AI and Copilot data governance ⚠️ | Advanced |
| Compliance architecture across multiple regulatory frameworks | Expert |
K.8 Microsoft Graph
| Skill | Level |
|---|
| Graph Explorer, permission model, delegated vs application permissions | Must Know |
| Microsoft Graph PowerShell SDK for daily administration | Must Know |
| Reports API for usage and activity data | Must Know |
Filtering, paging, batching, $select, advanced query parameters | Should Know |
| Throttling, retry and backoff patterns | Advanced |
| Change notifications (webhooks) and subscription lifecycle | Advanced |
| Graph connectors and external content indexing | Advanced |
| Graph-based platform design and least-privilege permission architecture | Expert |
K.9 PowerShell
| Skill | Level |
|---|
| Core language, pipeline, objects, error handling | Must Know |
| Workload modules: Graph SDK, ExchangeOnlineManagement, PnP, MicrosoftTeams, Intune via Graph ⚠️ | Must Know |
Bulk operations with safety controls (-WhatIf, batching, logging) | Must Know |
| Functions, modules, parameter validation, comment-based help | Should Know |
| Authentication patterns: certificate, managed identity, service principal | Should Know |
| Pester testing, PSScriptAnalyzer, CI validation | Advanced |
| Reusable framework design (logging, retry, secrets, notification) | Advanced |
| Building and maintaining a shared internal automation platform | Expert |
K.10 Power Platform & Power Automate
| Skill | Level |
|---|
| Power Automate cloud flows, triggers, actions, error handling | Must Know |
| Power Platform admin center: environments, DLP, capacity | Must Know |
| Connector security model and DLP policy design | Should Know |
| Power Apps canvas basics and delegation limits | Should Know |
| Dataverse: tables, relationships, security roles | Advanced |
| Solutions, ALM, pipelines, environment strategy | Advanced |
| Managed Environments, CoE Starter Kit, governance at scale ⚠️ | Advanced |
| Power BI / Fabric modelling for operational reporting ⚠️ | Advanced |
| Enterprise citizen-development governance model | Expert |
K.11 Copilot & Copilot Studio
| Skill | Level |
|---|
| M365 Copilot licensing, capabilities and data boundary ⚠️ | Must Know |
| Copilot readiness: permissions, labels, search quality prerequisites | Must Know |
| Copilot admin controls and usage reporting ⚠️ | Must Know |
| Copilot Studio: topics, knowledge sources, actions, publishing | Should Know |
| Agent governance: environments, DLP, ALM, ownership, review ⚠️ | Advanced |
| Graph connectors and declarative agents for enterprise data ⚠️ | Advanced |
| Copilot value measurement and adoption programme design | Advanced |
| AI risk: prompt injection, over-permissioned agents, data leakage ⚠️ | Expert |
| Enterprise AI architecture spanning Copilot, Studio and Azure AI ⚠️ | Expert |
K.12 Azure (as it supports M365)
| Skill | Level |
|---|
| Subscriptions, resource groups, RBAC, Azure Policy basics | Must Know |
| Azure Automation: runbooks, schedules, managed identity, modules | Must Know |
| Key Vault for secrets and certificates | Should Know |
| Log Analytics workspace design and KQL | Should Know |
| Logic Apps and Azure Functions | Advanced |
| Microsoft Sentinel: connectors, analytics rules, playbooks, workbooks | Advanced |
| Azure networking as it affects M365 (ExpressRoute, Private Link, DNS) | Advanced |
| Cost management and consumption governance | Advanced |
| Hybrid and multi-cloud identity/security architecture | Expert |
K.13 Security, Compliance & Governance (discipline, not product)
| Skill | Level |
|---|
| Zero Trust principles applied to M365 specifically | Must Know |
| Least privilege, separation of duties, privileged access management | Must Know |
| Incident response process and evidence handling | Must Know |
| Change management and configuration control discipline | Must Know |
| Regulatory frameworks relevant to your industry and geography | Should Know |
| Risk assessment, risk register, exception management | Should Know |
| Threat modelling for cloud collaboration platforms | Advanced |
| Security architecture and control design across the M365 stack | Expert |
| Governance framework design and business engagement | Expert |
K.14 Automation & AI Engineering (cross-cutting)
| Skill | Level |
|---|
| Identifying automation candidates and calculating ROI | Must Know |
| Idempotency, error handling, logging, alerting in automation | Must Know |
| Version control and code review discipline | Must Know |
| CI/CD pipelines for configuration and automation | Advanced |
| Infrastructure/configuration-as-code for M365 | Advanced |
| Agentic AI patterns, tool use, grounding and evaluation ⚠️ | Advanced |
| Responsible AI, evaluation and guardrail design ⚠️ | Expert |
| Platform engineering mindset applied to the M365 estate | Expert |
K.15 Non-technical skills that determine seniority
| Skill | Why it matters at senior level |
|---|
| Translating technical risk into business language | Determines whether your risks get funded |
| Writing clear documentation and decision records | Determines whether your work survives you |
| Stakeholder and vendor management | Determines what you can actually deliver |
| Estimation and prioritisation under constraint | Determines whether you burn out |
| Facilitating decisions you do not own (Legal, HR, Security, Finance) | Most M365 governance decisions are not IT's to make alone |
| Teaching and delegating | The only way to escape being the single point of failure |
L. Certification Mapping
⚠️ Verification requirement — read before using this section
Microsoft's certification portfolio changes frequently: exams are renamed, replaced, merged and retired, and new formats (Applied Skills, agent-focused credentials) are added regularly. This mapping was produced without live access to Microsoft's certification site and must be verified before you commit study time or budget.
Verify every item against:
https://learn.microsoft.com/credentials/ — the authoritative list of active certificationshttps://learn.microsoft.com/credentials/certifications/retired — retirement announcements- The specific exam page for current skills-measured documents and any dated change notices
Exam codes below are given as search keys, not as guarantees of current status.
L.1 Foundation credentials
| Exam / credential | Title (verify) | Maps to responsibilities | Recommendation |
|---|
| MS-900 | Microsoft 365 Fundamentals | Whole-of-M365 overview, licensing basics | Skip if already senior; useful for team members |
| SC-900 | Security, Compliance & Identity Fundamentals | Security/compliance vocabulary across the stack | Skip if already senior |
| AZ-900 | Azure Fundamentals | Azure services underpinning automation | Skip if already senior |
| PL-900 | Power Platform Fundamentals | Power Platform overview | Skip if already senior |
| AI-900 | Azure AI Fundamentals | AI concepts underpinning Copilot | Optional |
L.2 Core certifications for this role
| Exam / credential | Title (verify) | Catalog sections covered | Priority |
|---|
| MS-102 | Microsoft 365 Certified: Administrator Expert ⚠️ | A.1, A.2, A.7, A.8, A.9, A.22 — tenant, identity, threat protection, compliance | Highest. The definitive credential for this role. |
| SC-300 | Identity and Access Administrator Associate | A.1, A.16, A.18, B.1 — Entra ID, CA, PIM, governance, app integration | Highest. Identity is the perimeter. |
| MD-102 | Endpoint Administrator Associate | A.6, A.10, A.13, B.6 — Intune, endpoints, apps, Windows | High |
| SC-200 | Security Operations Analyst Associate | A.7, A.18, B.2 — Defender XDR, hunting, Sentinel | High |
| SC-401 | Information Security Administrator Associate ⚠️ | A.8, A.19, B.7 — Purview, information protection, DLP, insider risk. Believed to have replaced the retired SC-400 — verify. | High |
| MS-700 | Teams Administrator Associate | A.3, B.5 — Teams administration and governance | Medium–High |
| SC-100 | Cybersecurity Architect Expert | A.18, B.2 — security architecture across the estate. Requires a prerequisite associate certification. | High for architect track |
| AZ-104 | Azure Administrator Associate | A.17, A.24, B.3 — Azure Automation, Key Vault, Log Analytics | Medium–High for automation depth |
| AZ-500 | Azure Security Engineer Associate | A.18, B.2 — Azure-side security controls | Medium |
| AZ-305 | Azure Solutions Architect Expert | Architecture-track credential | Medium for architect track |
| PL-200 | Power Platform Functional Consultant Associate | A.15, B.3 — Power Platform and Dataverse | Medium |
| PL-600 | Power Platform Solution Architect Expert | A.15, B.4 — enterprise Power Platform architecture | Medium for architect track |
| PL-300 | Power BI Data Analyst Associate | F, E.5 — reporting and dashboards | Medium |
| AI-102 | Azure AI Engineer Associate | B.4 — AI services underpinning custom agent work | Medium |
| MS-721 | Collaboration Communications Systems Engineer Associate ⚠️ | A.3 — Teams Phone, meetings, devices | High only if you own voice |
| AB-620 | Designing and Building Integrated AI Agent Solutions in Copilot Studio ⚠️ | A.11, B.4 — Copilot Studio agents, connectors, ALM. Newer credential — verify code, title and availability. | High if agents are on your roadmap |
L.3 Applied Skills credentials ⚠️
Microsoft Applied Skills are shorter, scenario-based validations rather than full certifications. They are useful for demonstrating specific capability quickly and for team members who do not need a full certification. Titles change frequently — check the current catalogue — but the categories historically relevant to this role include:
| Area | Typical Applied Skills scope |
|---|
| Identity | Configuring secure access to workloads using Microsoft Entra ID |
| Defender | Defending against cyberthreats with Microsoft Defender XDR |
| Purview | Implementing information protection and data loss prevention |
| Sentinel | Configuring SIEM security operations using Microsoft Sentinel |
| Copilot Studio | Building agents with Microsoft Copilot Studio |
| Intune | Managing and securing endpoints |
| Power Platform | Creating and managing automated processes |
L.4 Certifications believed retired or superseded — verify before pursuing ⚠️
| Legacy code | Status (verify) | Successor |
|---|
| MS-500 (Security Administrator) | Believed retired | Split across SC-200, SC-300, SC-401 |
| MS-100 / MS-101 | Believed retired | MS-102 |
| MS-203 (Messaging Administrator) | Believed retired | No direct successor; Exchange content folded into MS-102 and role-based learning paths |
| SC-400 (Information Protection Administrator) | Believed retired | SC-401 |
| MS-600 (Teams Application Developer) | Believed retired | Developer-track alternatives |
| MS-220 (Exchange Online Troubleshooting) | Verify status | — |
| Legacy Microsoft 365 Enterprise Administrator Expert path | Superseded | MS-102 Administrator Expert |
L.5 Suggested sequencing for an experienced M365 administrator
| Phase | Credential | Rationale |
|---|
| 1 | MS-102 | Establishes the whole-of-tenant baseline and is the role's signature credential |
| 2 | SC-300 | Deepens the most consequential domain: identity |
| 3 | MD-102 or SC-200 | Choose by where your estate's risk and workload actually sit |
| 4 | SC-401 ⚠️ | Purview depth is the most common senior-level gap |
| 5 | AZ-104 | Unlocks credible automation and platform engineering |
| 6 | SC-100 or PL-600 / AZ-305 | Architect-track branching point: security architecture vs solution architecture |
| 7 | AB-620 ⚠️ / AI-102 | AI and agent capability — the current differentiator |
A caution worth stating plainly: certifications validate breadth of knowledge, not operational judgement. A senior administrator's real credential is a documented record of incidents handled, automations delivered with measurable savings, migrations completed, and audits passed. Pursue certifications to close known gaps and to satisfy partner or employer requirements — not as a substitute for building and documenting the operating model described in this document.
L.6 Continuous learning sources beyond certification
| Source | Use |
|---|
| Microsoft 365 Message center + Roadmap | The authoritative, tenant-specific change stream |
| Microsoft 365 admin and workload blogs (Tech Community) | Feature context and design intent |
| Microsoft Learn documentation | Authoritative reference; the only source to trust for current behaviour |
| Microsoft Graph changelog ⚠️ | API changes that break automation |
| Ignite and Build session recordings | Direction of travel, 12–18 months ahead |
| CIS Microsoft 365 Benchmark ⚠️ | External, structured hardening yardstick |
| A personal developer/lab tenant | The only safe place to actually test |
M. Final Senior Admin Operating Model
"If I were responsible for the entire Microsoft 365 tenant, this is exactly how I would run it."
The model rests on five convictions:
- Identity is the tenant. Everything else is configuration on top of it.
- If it is not monitored, it is not managed. If it is not documented, it does not exist. If it has not been restored, it is not backed up.
- Repetition is a design failure. Anything done manually more than twice a month is an automation defect, not a workload.
- The change stream is the job. Microsoft changes this platform continuously; managing that stream is the operational discipline.
- Capability the business does not use is money the business wasted. Activation and adoption are part of administration, not marketing.
1. Daily routine
| Time | Activity |
|---|
| First 45 min | The daily checklist (Section C.2) — service health, Message center, security incidents, identity risk, mail flow, automation failures, backup status, compliance delta. Judgement only; detection is automated. |
| Mid-morning | Escalations and P1/P2 work. Nothing else competes with this window. |
| Midday | Change execution within the approved change window; peer review of any change another admin is making. |
| Afternoon (protected 2h) | KTBE build block. No meetings, no tickets. This block is defended absolutely — it is the only reason the model works. |
| Late afternoon | Ticket queue, documentation of the day's changes, handover notes. |
| End of day | Confirm the daily checklist is recorded; confirm nothing critical is unowned overnight. |
Daily non-negotiables: every change made is recorded; every incident gets a written note the same day; every Message center item is triaged, not accumulated.
2. Weekly routine
Monday: Message center backlog burn-down; week's change calendar confirmed; expiry watchlist reviewed. Midweek: the Section C.3 review set — CA drift, MFA coverage, legacy auth, external forwarding, licence errors, Intune compliance, vulnerability exposure, storage trend, automation health. Friday: KTBE progress review, backlog grooming, the week's documentation updates, and a written weekly summary to the team and manager. One protected half-day for deep KTBE work.
3. Monthly routine
The Section C.4 review set, plus three things that matter more than the reviews themselves: the executive report (risk, cost, adoption, incidents, savings), the post-incident reviews converted into backlog items, and the KTBR/KTBE time-split measurement. If the split has drifted below 30% KTBE for two consecutive months, that is escalated as a capacity problem, not absorbed silently.
4. Quarterly routine
The Section C.5 set. Four items are treated as immovable regardless of workload: break-glass validation, restore testing, a disaster tabletop exercise, and the full Conditional Access review. These are the four that never feel urgent and are always the ones missing from the post-mortem. Quarterly also carries the business service reviews, the roadmap update, and a deliberate look at what Microsoft is shipping that changes the plan.
5. Yearly routine
Architecture review, BCP exercise at full scale, licensing strategy and renewal preparation, security and governance strategy refresh, full permission recertification, legacy retirement planning, and the honest annual question: what is still manual that should not be, and what did I not get to this year? Plus the personal development plan for the next twelve months.
6. KTBR priorities
In strict order when time is short: identity and privileged access → the ability to recover (break-glass, backup, restore) → the change stream (Message center, drift, expiry) → security detection and response → data exposure control → service availability monitoring → compliance evidence → everything else. Anything below the line gets deferred openly and recorded as accepted risk, never dropped silently.
7. KTBE priorities
In strict order: automate JML → automate licence optimisation (funds the rest) → configuration-as-code → self-service and tiered support (buys capacity) → Copilot readiness and adoption → modernisation of whatever generates the most KTBR load → agentic and AI capability. The sequencing is deliberate: the first four create the time and money that pay for the last three.
8. Automation strategy
Everything runs on managed identity or workload identity federation — no secrets in scripts, ever. Logic lives in Graph and PowerShell; orchestration in Azure Automation and Logic Apps; human interaction in Power Automate, Power Apps and Teams. Every automation has a named owner, a README, structured logging, failure alerting to a central channel, and an entry in the automation registry. Nothing reaches production without peer review and a test run in the lab tenant. Containment actions may be automatic; destructive actions require a human gate. The measure of success is not the number of automations built but the reduction in manual hours and in the ticket categories they were built to eliminate — reported monthly.
9. Security strategy
Zero Trust applied concretely: verify explicitly (Conditional Access on every application, phishing-resistant authentication for every administrator and progressively for every user), least privilege (zero standing privilege via PIM, least-privilege Graph application permissions, delegated workload roles instead of Global Admin), and assume breach (comprehensive logging into Sentinel, automated containment playbooks, quarterly tabletop exercises, tested recovery). Secure Score is tracked but never treated as the goal. The security backlog is prioritised by attack path, not by product feature. Every security control has an owner, a test, and evidence.
10. Governance strategy
A short, written tenant configuration standard ("golden config") that is the reference for drift detection. Change control on everything tenant-wide, with an emergency path that still requires retrospective documentation. Governance applied at creation, not retro-fitted — provisioning applies naming, labels, permissions, owners and retention automatically. An exception register where every exception has an expiry date, reviewed quarterly, with expiry enforced. Decisions recorded as ADRs so they are not relitigated. And an explicit acknowledgement that classification, retention and access decisions belong to the business — IT builds the mechanism and holds the business to using it.
11. AI / Copilot strategy
Readiness before rollout, without exception. Copilot is enabled only on a data estate where oversharing has been remediated, labels are applied, and discovery scope is constrained. Adoption is run as a programme — champions, role-based scenario libraries, training, measurement — because Copilot value is adoption-limited, not capability-limited. Agents go through an intake and review board, live in governed environments with DLP, have named owners, logged interactions, approval gates before write actions, and a scheduled review with automatic decommission if unused. Value is measured against a baseline and reported quarterly, because the renewal conversation will demand evidence. And the AI acceptable-use policy is written, trained and monitored before the first broad enablement, not after the first incident.
12. Reporting strategy
Three audiences, three products. Operational (daily/weekly, for the team): live dashboards with thresholds and owners. Managerial (monthly, for IT leadership): incidents, risks, capacity, cost, automation savings, adoption. Executive (quarterly, for the business): risk posture in business terms, cost avoided in currency, capability delivered, and the forward roadmap. All three are generated from the same automated pipeline with thirteen months of history — the narrative layer is the only manual part. The rule: never present a number you cannot drill into, and never present a metric with no owner or action attached.
13. Business stakeholder engagement
Quarterly service reviews with each major function — HR, Finance, Legal, Sales, Operations — covering what they use, what they need, what is coming, and what is costing them. A named business owner for each major capability. Legal and Compliance are consulted on retention, eDiscovery, labels and insider risk as decision-makers, not as reviewers. Security is a partner in the CA and privileged-access model, not an approver at the end. Finance sees licence optimisation savings in their own language. And the platform team markets its own work: capability nobody knows about is capability nobody uses, and unfunded platform work is usually invisible platform work.
14. Career and skill development
Certification closes known gaps (Section L sequencing), but the durable assets are different: a portfolio of automations with quantified savings, a record of incidents handled and what changed afterwards, architecture documents and decision records you authored, and people you taught to do parts of your job. Time is allocated explicitly — a few hours weekly, protected like the KTBE block — for learning, lab work and writing. The trajectory to aim at is not "better administrator" but platform owner and architect: someone who decides what the tenant should become, builds the mechanism to get it there, and can explain both to a board. The KTBR discipline in this document is what makes that trajectory possible; the KTBE work is what makes it happen.
How to use this document
| Turn it into | Using |
|---|
| M365 Admin SOP | Section A catalog + Section A.27 deep-dive cards + Section D priority matrix |
| Daily checklist | Section C.2, converted to a Planner/To Do recurring task or a Teams adaptive card |
| Monthly operations checklist | Section C.4, with evidence capture automated per Section E |
| Automation backlog | Section E.1 table + Section J levels, prioritised by Section I ranking |
| Career roadmap | Section K skill matrix + Section L sequencing, reviewed quarterly |
| Improvement programme | Section G, with Section F KPIs as the measurement layer |
| Risk register | Section A.26 highlighted groups, with owners and target dates added |
Before operationalising: re-verify every ⚠️-flagged item against current Microsoft documentation and your own tenant's licensing entitlements. Product capability, licensing boundaries, retirement dates and certification codes change on Microsoft's schedule, not yours — which is precisely why Message center triage (M365-02) sits at rank 10 in the KTBR priority list