3 min read
Your best user research is already in your support inbox
Before you recruit a single participant, mine the thousands of support tickets you’ve already paid for.
AI
Sanket Daithankar
Product & Game Designer
I help teams ship products faster, with research backed design that keeps users coming back.
The research you already own
Most product teams ignore the cheapest research they have. Support tickets. Thousands of them, from real users with real friction, already transcribed. Nobody had to be recruited and no calls had to be scheduled.
Last month I exported six months of support tickets from our app: forty thousand rows of CSV. I pasted the whole thing into Claude with one prompt asking it to cluster by friction point, weighted by frequency, and link each cluster to the underlying flow. I also asked for a summary of top-performing products, total ratings, average ratings, and the percentage of negative feedback.
What twenty minutes gave me
Twenty minutes later I had a full picture of the product, without a single interview.
Positive themes showing what users genuinely valued
Recurring complaints ranked by how often they appeared
Feature requests, grouped and counted
A monthly trend showing exactly when negative reviews started to spike
Each cluster was tied back to the flow where the friction lived, so the findings were not abstract. They pointed to specific screens and moments in the journey.
The problems no one could see
Three of the top five problems were not on the roadmap. Not deprioritised. Not deferred. Just invisible.
That is the part that stuck with me. The signal had been there the whole time, sitting in a folder no one on the product team was reading. Support teams hear the pain first, but that knowledge rarely travels to the people deciding what to build next. The data was never missing. It was simply stored in the wrong place.
Before your next round of interviews
User interviews are valuable, and I am not suggesting you stop running them. But before you book the next round, check what your support data is already telling you. It will sharpen the questions you ask and show you where to dig deeper.
Start small. Export a few months of tickets, ask for clusters by friction point, and compare the top issues against your roadmap. If even one of them is missing, you have your answer.
You already paid for this research. Make use of it.
