Talking to one user a week
Not a strategy, a small habit, but the one that has reshaped the product more than anything else.
I talk to one user a week. Always one — not three, not the panel my advisor suggested. Thirty minutes, by video, scheduled the same way I schedule a barber appointment.
Why one#
Three users in a row turn into a montage in my head. One conversation is slow enough to remember in detail. I leave the call with maybe two real observations and one thing the user said that I cannot get out of my head all week. That third thing is where the product moves.
What I ask#
Three questions, in order. What did you try to do today before you opened this? What did you do after? What did you almost do but did not? The third question is the most valuable. The "almost" tells me where the product fails silently.
What I do not ask#
I do not ask for feature requests. People give them anyway, and I write them down, but I have never shipped one verbatim. The good ones are translations of the actual answer to question three.
The compounding#
Fifty conversations a year. The first ten taught me what the product was not. The next forty have been refining what it is. Both halves were worth the calendar slot.