Sure. We noticed in our telemetry that a subset of users had lower monthly active usage than the rest of the cohort, so I ran a segmentation analysis in SQL to understand who they were. I found that users who had not completed onboarding had significantly lower retention at ninety days. I put together a deck, shared it with the PM, and they ended up adding an onboarding reminder to the product roadmap. It was a good example of using data to influence what got prioritised for the next quarter.
We were seeing flat Teams meeting adoption among large enterprise accounts in the financial services segment — something we flagged as a customer risk, not just a usage metric. I pulled seat utilisation data segmented by enterprise size and industry, and found that sixty-three percent of financial services accounts had IT admin policies blocking external meeting participants — a constraint consumer engagement frameworks would have completely missed. I brought that finding to the PM and the enterprise customer success team together. The product decision was to build IT admin-configurable meeting policies that enterprise admins could unlock without requiring a support ticket. Within two quarters, external-participant meeting usage in that segment increased thirty-eight percent. The insight had to come from understanding how enterprise IT governance actually works — not from optimising a metric.