Forum Discussion
AI-generated feedback summaries for managers - is this a thing yet?
hi CandyMullany yes, it is becoming a thing , but what you’re trying to do (pull goals + feedback + 1:1 notes together automatically) isn’t fully “out-of-the-box” with Copilot yet.
Right now, Copilot works really well when:
The data is already in M365 (Teams chats, meeting notes, emails, docs)
You give it a clear prompt to summarize or draft
Where it struggles is exactly your use case , bringing together structured HR data (goals, ratings, progress) + unstructured notes across systems. Most of that data typically lives in tools like Workday or SuccessFactors, not in M365, so Copilot simply can’t “see” everything in one place.
That said, people are starting to use it in a practical way:
Drafting performance feedback from meeting notes and emails
Summarizing 1:1 discussions into talking points
Rewriting feedback to improve tone and clarity
Generating review bullets based on pasted inputs
Microsoft is also pushing this direction ,using Copilot + agents to aggregate signals (feedback, metrics, surveys) and generate manager insights, but that typically requires Copilot Studio or some level of customization/integration rather than just chat prompts.
What’s working today (in practice):
Instead of expecting full automation, most teams are doing something like:
Here are this employee’s goals, recent feedback, and my 1:1 notes, summarize strengths, gaps, and suggested review comments.
So basically:
You feed the context manually (or via docs)
Copilot structures and drafts the output
What’s missing (today):
Native integration with HR systems
Automatic “360 view” of an employee
End-to-end review generation without human input
If you want to make it work better:
Store 1:1 notes and feedback in OneNote / Loop / SharePoint (so Copilot can access them)
Use consistent templates for goals/feedback
Or go one step further: build a Copilot Studio agent that connects to your HR system
Bottom line: Yes , AI-assisted feedback writing is definitely emerging, and Copilot can already help a lot with drafting and summarizing. But the “fully data-backed, all-in-one feedback generator” you’re describing usually needs custom setup + better data integration, not just standard Copilot chat.