do it the way
that already worked.

run the 40% must-have survey on beta users

why now

shipping is cheap and noisy in 2026: a september 2026 count found show hn posts now make up over 12% of links submitted to hacker news, and marc lou’s 2025 review shows even a 100,000 follower audience did not save two launches. a launch spike says little; a must-have score from active users does, and superhuman found about 40 responses give a directional answer.

your first move

1. 1pick who gets surveyedonly users who used the core feature at least twice in the last 2 weeks. leave out friends, family and teammates.
2. 2send the four questionshow would you feel if you could no longer use {product}? (very, somewhat, not disappointed). who would benefit most? main benefit? how can we improve it? free google form or tally.
3. 3score it honestlyvery disappointed divided by all responses. low early scores are normal: superhuman started at 22% and reached 58% three quarters later.
4. 4profile the ‘very disappointed’ groupwhat role, company size and use case do they share? rewrite your ideal user description in their own words.
5. 5fix what holds back ‘somewhat’among ‘somewhat’ users whose main benefit matches the ‘very’ group, list what is missing. spend half the next sprint on that, half on the loved benefit.
6. 6survey again every 4 weekssame questions, new active users. track the score on one line in your weekly notes.