Reports landing
The umbrella usage dashboard. Pick Microsoft 365 Copilot from the report tile grid to drill into Readiness, Usage, Chat, Agents, and Search.
Docs ↗🏛️ MICROSOFT NATIVE REPORTING
What Microsoft already ships in the box
Before you stand up a single Power BI template, your tenant already has Copilot and Agent telemetry waiting for you in the Microsoft 365 admin center, Viva Insights Copilot Dashboard, Microsoft Purview, and the Power Platform admin center. The Analytics Hub picks up where these stop.
SOURCE 1
Microsoft 365 admin center
admin.microsoft.com → Reports → Usage → Microsoft 365 Copilot. Role required: AI Administrator, Reports Reader, or Global Reader. Data latency typically ~72 hours.
Copilot Readiness
License eligibility · which users meet the Copilot prerequisites · which M365 apps they actively use (Outlook, Teams, Word, Excel, PowerPoint, OneNote) · technical blockers. Use this to pick your first cohort.
Docs ↗Copilot Usage
Adoption + retention + engagement. Enabled users vs active users, active-day trends, per-app integration usage, top agents. Anonymizes user-level data by default — switch to identified mode if your privacy policy allows.
Docs ↗Copilot Chat usage
Specifically for unlicensed users interacting with Copilot Chat in Teams, Outlook, m365.cloud.microsoft, Edge, Word, Excel, PowerPoint, OneNote. Total active users, average daily active, prompts submitted, per-user activity.
Docs ↗Copilot Agents usage
Declarative agents, SharePoint agents, Custom engine agents. Total active users (licensed + unlicensed), total active agents, daily time series, active usage per user / per agent / per agent-user pair. Includes agents built by your org, Microsoft, or Microsoft Partners.
Docs ↗Copilot Search usage PREVIEW
Active usage and search-activity trends at org level, plus per-user search behavior. Reflects user behaviors within 1 hour (much fresher than the 72-hour reports above). Still in public preview.
Docs ↗SOURCE 2
Viva Insights · Copilot analytics
Two surfaces: the Copilot Dashboard (curated, in Teams) and the Advanced Insights Analyst workbench (custom queries, Power BI templates). License required: Viva Insights. Roles: AI Administrator (to enable) + delegated dashboard access or Analyst role.
Adoption metrics
User adoption rates and growth trends. Tracks the population curve: licensed → enabled → active → habitual.
Usage patterns
When and how users interact with Copilot — by day of week, by app, by time of day. Surfaces the difference between exploration and ingrained habit.
Productivity impact
The behavior-change layer. Measures the effect of Copilot on focus time, meeting load, after-hours work, and collaboration patterns. The closest native equivalent to Super User Impact.
ROI indicators
Data points to help assess return on investment — not a dollarized calc, but the inputs you need to build one.
Actionable insights
Recommendations engine: which cohorts to nudge, which apps to push, where adoption is stalling. The Dashboard's "what should I do next" panel.
Analyst workbench
Custom Copilot queries · Power BI templates · cross-data analysis · detailed executive reporting. This is the surface most of the Analytics Hub's Viva-based reports (Super Usage, Super Impact) plug into.
SOURCE 3
Microsoft Purview · Audit logs
The raw event stream behind every Copilot interaction — chat prompts, agent invocations, file activity. Not a "report" in the visual sense, but the truth-source most Analytics Hub reports build on. Role required: Audit Reader or above.
Unified Audit Log
Searchable in the Purview portal or via PowerShell / Graph API. Every Copilot interaction, every agent invocation, every file the model touched. Compliance + security teams already have access.
Docs ↗Compliance + DLP signals
Copilot interactions show up in DLP policy matches, sensitivity-label events, and Insider Risk Management. You don't need a custom report to know if Copilot saw a secret — Purview already flags it.
SOURCE 4
Power Platform + Copilot Studio analytics
If your org builds custom agents in Copilot Studio, the Power Platform admin center has consumption + performance metrics per agent. Complements (doesn't duplicate) the admin-center Agents usage report.
Agent consumption
Message-pack burn rate per agent, per environment. Tells you which agents are economically viable and which are runaway costs.
Agent performance
Containment rate, escalation rate, satisfaction scores, abandon points. Built into the Copilot Studio analytics tab per agent.
🔁 NATIVE VS ANALYTICS HUB
Where the Hub picks up
Native reports answer "what happened." Analytics Hub answers "so what, now what." Both are designed to coexist — the Hub never duplicates what's already in the admin center.
| You need to… | Start native | Layer Analytics Hub |
|---|---|---|
| See who's eligible & provisioned | Admin center · Copilot Readiness | M365 Readiness — adds cohort filters, stranded-seat alerts, executive narrative |
| Track adoption + retention | Admin center · Copilot Usage | AI-in-One + Super Usage — cross-tabs by department, persona, geo, tenure |
| Understand chat usage (licensed & unlicensed) | Admin center · Copilot Chat usage | Chat & Agent Intelligence — prompt topic analysis, app fan-out, repeat-user cohorts |
| Audit agent invocations | Admin center · Agents usage | AI-in-One + Chat & Agent Intelligence — who's building vs who's using, agent-to-user fan-out, governance posture |
| Quantify productivity lift | Viva · Productivity impact | Super User Impact + What I Did — peer-comparable deltas, dollarized in ROI Calculator |
| Build an exec ROI story | Viva · ROI indicators | ROI Calculator — scenario sweeps, payback model, defensible assumptions |
| Pull tenant-scale audit data on a schedule | Purview · Audit search (manual) | PAX — no row limits, schedulable, pushes to Lake / Warehouse / Fabric / SIEM |
| Survey sentiment + qualitative themes | Viva · Survey insights | Adoption & Sentiment — NPS layered on usage cohorts |
| Identify & recognize champions | — | CustomizeCopilot — Champion-ID, branding, theming |
| Measure GitHub Copilot for developers | GitHub Enterprise · Copilot metrics API | GitHub Copilot Impact — chat vs agent, language, model, acceptance rates |
START NATIVE · LAYER THE HUB
The Hub doesn't replace Microsoft's reports — it amplifies them.
Most teams run native + Analytics Hub side-by-side. The admin center gives you the official numbers your CFO trusts. The Hub gives you the slicers, comparisons, and ROI story that get the next budget cycle approved.