Quant teams wait for research. Research waits for a round of recruiting. Meanwhile the leak keeps its favourite story: “users don’t get the value.” The story may be true. It is unearned until someone who stalled says something specific.
Module 4 of Reading the Drop asks for one conversation, not twelve. Sampling is ugly: a recent trial who did not activate, recruited through whatever channel you already have. The transcript sits beside the stage chart. If they describe a permissions wall and your chart shows a drop after invite, you have a paired claim. If they describe pricing and your drop is before the paywall, you have a mismatch worth writing down.
This is not ethnography. It will not satisfy a research guild, and that is fine. The limitation is explicit: one voice can overfit. The memo must say so. We still prefer an overfit human sentence to a persona slide with no timestamps.
If you cannot recruit in time, the mark can still pass on an honest window — see the reservation in our reviews. The qualitative overlay is a habit, not a hostage situation.