Column · revenue vs retention
That the promise was clear enough and the price low enough for someone to risk it. That is real, and it is also the weakest form of proof, because it requires no evidence of quality.
That the first one delivered something the buyer wanted again. Nobody pays twice to be disappointed.
Because the window is short. With ten customers and a two-week history, retention is statistically invisible. The temptation is to declare success from the acquisition curve and stop measuring.
Record every purchase with a date and an identity. Then ask a single question monthly: of the people who bought last month, how many bought again? That fraction, however small the sample, is more informative than any growth rate you can compute from new buyers.
See the data this is built on
Open the snapshot