Talking to Users
Discovery, research, and the art of hearing what people mean, not just what they say.
The risk is not that the LLM gets labels wrong. It is that the researcher stops reading.
Using AI to Code Interviews Without Losing the Analysis
LLMs can label interview transcripts in seconds - the question is whether the labels mean anything without a human who…
ReadAbandonment is silent, and that silence makes it invisible to every quality metric you track.
Why Closing Tickets Does Not Fix Your Product
Engineering resolves reported problems - qualitative research reveals the problems no one knows how to report. A team t…
ReadPresenting options is not giving the room more information; it is giving them your uncertainty to manage.
The Four Questions Every Research Readout Must Answer
A research readout that does not answer 'so what do we build' is a document, not a decision. This article lays out the…
ReadMost AI product research is validation theater — a confirmation ceremony dressed up as research.
Writing Interview Questions for Products That Do Not Always Work the Same Way
When your product's output is non-deterministic, leading questions do not just bias the answer - they hide whether the…
ReadFive clean, non-overlapping buckets means you over-organized data messier than your categories admit.
How to Turn Messy Interview Transcripts Into Product Decisions
Thematic analysis is the process of reading raw qualitative data repeatedly until patterns emerge - then naming those p…
ReadEnd-user frustration that does not threaten renewal is the most dangerous kind.
Research Is Different When the Person Paying Is Not the Person Using
In B2B, the person who renews the contract and the person who uses the product daily are rarely the same person - and r…
ReadProtecting engineers from user conversations to preserve focus produces worse products.
How to Run Research as a Team Without It Dying in a Handoff
Continuous discovery only works if the people who build the product hear users directly. A PM relay is not a substitute…
ReadThe interview where the user agrees with your framing is almost always the least useful.
The User Interview That Actually Teaches You Something
A user interview is a conversation with an agenda you do not reveal - the goal is to understand how someone thinks, not…
ReadThe most valuable cohort is the abandoned one you are not recruiting.
Why Your User Segments Are Lying to You
Most research cohorts are built around whoever responds to in-app prompts - paying customers who have already solved th…
ReadSilent majorities are silent because they have lower expectations, not because they are satisfied.
When One Angry User Is Actually Right
Most teams dismiss loud complaints as outliers without testing whether the outlier reveals a structural problem the sil…
ReadThe sprint did not eat your research. Your team ate it.
Why Your Sprint Cycle Is Eating Your Research
Continuous discovery does not fit inside a sprint - it runs alongside one. Teams that wait for sprint capacity to do re…
ReadWho you think will pay is the loudest; who actually pays has the hardest deadline.
Research Before Building: Stress-Testing Inherited Assumptions
Why arriving at a problem statement independently is verification, not redundancy, and where inherited assumptions quie…
Read