How to use this library
This page exists for one reason: most leaders looking at Microsoft Copilot have heard the marketing pitch and now need to translate it into their own week. The use cases here are written from that angle. Pick the role that owns the workflow, read the scenario in the words an actual manager would use, and look at the recommended Copilot for the shape of work involved. The point is to move you from "AI in general" to "we could try this specific thing in the next month".
Why role-first, not product-first
Copilot is a family of products, not a single thing. Microsoft 365 Copilot, Copilot for Sales, Copilot for Service, Copilot for Finance, GitHub Copilot, Copilot for Security and Copilot Studio all do different jobs at different price points. The best way to choose between them is to start from the work, not the SKU. A finance team trying to speed up month-end commentary needs something different from a sales team trying to write better proposals, even though both will see the word "Copilot" in their license agreement.
That is also why most use cases here recommend more than one Copilot. The honest answer is that several products can deliver the outcome - what changes is how much heavy lifting each one does, how much context it brings out of the box, and how much it costs per user per month. Where one is a clear leader for a use case, we say so. Where two are reasonable, we name both.
Reading the effort and time-to-value labels
"Low effort" means you can stand it up in a couple of weeks, mostly through licensing and a focused team brief. "Medium effort" usually involves either configuring a specialist Copilot against your CRM or finance system, or building a small custom agent in Copilot Studio. "High effort" means a real engineering or security project - worth doing, but not the place to start your first pilot.
Time-to-value is the time from "we have decided to do this" to "the team is genuinely using it on real work" - not the time to install the software. We are deliberately conservative on this. The time it takes to build the agent is rarely the bottleneck; the time it takes for the team to trust the output and change their habits is.
A note on data and security
None of these use cases are safe to roll out without first reviewing your SharePoint and Teams permissions. Copilot will surface anything users already had access to but never noticed - which is where most early problems come from. Run the data readiness section of our Copilot buying checklist before you scale any of these patterns past the pilot stage. It is a 90-minute exercise that prevents the most common avoidable mistake.
What this page deliberately is not
It is not a vendor directory. It is not a list of every clever thing Copilot can theoretically do in a demo. It is the subset of use cases we see UK SMBs actually pilot successfully in a 30 to 90-day window with realistic budgets and existing teams. If a use case for your role is missing, that is usually because it sits further out on the maturity curve - worth coming back to once your first pilot is live and trusted, but not worth starting with.