Interviews and surveys, plus the material you’re already sitting on: support tickets, sales calls, churn responses. AI does the synthesis across all of it, so findings come from hundreds of sources rather than the six people there was time to call. I then bring that into the room where ideas get formed, so options are pressure-tested early.
Decisions are being made from the loudest voice in the room. You have opinions about customers but no evidence, or evidence so old it predates the current product. Sales, support and product each believe something different about who the customer is. Someone says “customers want X” and nobody can point to where that came from.
Interviews and surveys where they're warranted, but I start with what you already hold: support tickets, sales call recordings, churn responses, review sites, in-product feedback.
AI does the synthesis across all of it, which changes the economics. Findings come from hundreds of sources rather than the handful there was time to read, and the loop can run continuously rather than once a quarter.
Then I go and talk to people, with the desk work already done, so interviews test hypotheses rather than fish.
The part that matters most is what happens next. I bring findings into the room where ideas get formed, so options get pressure-tested early rather than research being presented after the decision is made.
Two to four weeks. Can also run as an ongoing arrangement where the synthesis refreshes continuously and I report on what's changed.
We've never done research. Where do we start?
With what you already have. Almost every company is sitting on years of unread customer evidence in support and sales tooling.
How do you stop research becoming a report nobody reads?
By tying every finding to a decision that's currently open. If a finding wouldn't change what you do next, I leave it out.
Do you talk to our customers directly?
Yes, with your introduction. I can also work only from existing material if customer access is difficult.