Bite 1 / 8
Lesson 8.1

Measuring usage and impact

You'll be able to Name the Microsoft tools that measure Copilot usage and adoption.

Dashboard and reports

Use the Microsoft Copilot Dashboard (Viva Insights) and the admin-center usage report to see active users, actions per user, app usage and feature adoption. The Copilot Control System brings measurement together.

Check what you know

Which tool shows Copilot adoption and usage?

Which tool shows Copilot adoption and usage?
Takeaway

Measure with the Copilot Dashboard and the admin-center usage report.

Lesson 8.2

From usage to business outcomes

You'll be able to Explain how to connect Copilot usage to business outcomes, and how to treat ROI claims.

Usage is a means, not the end

The Copilot business impact report relates usage to business outcome measures you upload (for example, monthly deals closed). The impact report also estimates 'assisted hours' and a value figure — useful, but Microsoft itself labels these as estimates.

Be careful with ROI numbers

Treat specific ROI multiples or percentage gains as unverified marketing unless they trace to a primary source. Report what your own measures show. [author: do not paste vendor stats]

Check what you know

How should you treat a '40% productivity gain' headline?

How should you treat a '40% productivity gain' headline?
Takeaway

Tie usage to your own outcome measures — and don't quote unsourced ROI numbers.

Lesson 8.3

Service Health Reviews

You'll be able to Explain the purpose of a periodic Service Health Review.

Regular roundtables

A Service Health Review is a periodic roundtable where implementation and business stakeholders review progress, look at the data, and prioritise improvements. It keeps the operation honest and moving.

Check what you know

What is a Service Health Review for?

What is a Service Health Review for?
Takeaway

Hold regular Service Health Reviews to review data and prioritise improvements.

Lesson 8.4

Ongoing governance and operation

You'll be able to Explain why governance must continue as collaboration grows.

Keep access aligned to intent

Collaboration keeps changing, so oversharing can creep back. Ongoing governance — automated policies, regular monitoring, continuous improvement — keeps AI access aligned to intent. This is the 'Operate' half of deploy-then-operate.

Check what you know

Why does governance continue after go-live?

Why does governance continue after go-live?
Takeaway

Governance is continuous — collaboration drifts, so monitoring never stops.

Lesson 8.5

Continuous data governance

You'll be able to Name the recurring data-governance activities that keep the foundation healthy.

Recurring hygiene

  1. DSPM assessments — rerun Purview data risk assessments regularly.
  2. SAM / Purview reports — schedule and review them.
  3. Site-owner education — keep owners accountable for their content.

Check what you know

Which is a recurring data-governance activity?

Which is a recurring data-governance activity?
Takeaway

Keep the foundation healthy: recurring DSPM, scheduled reports, owner accountability.

Lesson 8.6

Support channels

You'll be able to Explain why sustained support matters beyond launch.

Help that lasts

Launch support fades; questions don't. Provide sustained help — a community, a help channel, refreshed guidance — so people keep getting unstuck after the launch buzz ends.

Check what you know

What does 'support channels' mean here?

What does 'support channels' mean here?
Takeaway

Support is a standing service, not a launch-week desk.

Lesson 8.7

Reporting to the AI Council

You'll be able to Explain how Run closes the loop back to the Stage 1 strategy and measures.

Close the loop

Report usage, impact and governance back to the AI Council against the success measures set in Stage 1. This is what makes the curriculum a loop, not a line: the Council decides what to adjust, fund or stop.

Check what you know

Reporting to the AI Council closes the loop back to which stage?

Reporting to the AI Council closes the loop back to which stage?
Takeaway

Report against the Stage 1 measures — Run closes the loop back to strategy.

Wrap-up

Sources & notes

Everything here traces to Microsoft Learn, retrieved June 2026.

Accuracy flags

  • Treat any specific percentage productivity gain as unverified marketing unless traced to a primary source. The Copilot impact report's 'assisted hours' is Microsoft's own estimate, with a default value rate of USD 72/hour (a U.S. Bureau of Labor Statistics estimate) — present it as an estimate, not a fact.
Fonts (Anton, Archivo Black, Inter) stand in for the licensed LearningBites faces — swap before publishing. Microsoft's Copilot and agent surface changes monthly; re-verify before reuse.