← ClaudeAtlas

demand-reconlisted

Decide whether a product direction is worth building. Six stages, escalating in cost: stages 0 through 3 are free and offline, a veto stops the run, and only surviving directions reach the multi-track investigation. Use when the user says "is this idea worth building", "validate this idea", "help me research this direction", "should I build X", "competitor research", "demand analysis", or the Chinese equivalents.
mifanr/demand-recon · ★ 0 · Code & Development · score 72
Install: claude install-skill mifanr/demand-recon
# Demand recon ## Two premises **Search engines index supply, not demand.** Every result is proof somebody already built the thing and wrote about it. So "one search finds it" and "the ground is taken" are the same sentence. This is not a rule of thumb, it is an identity. A need with no name returns nothing, which is also why you cannot search your way to one. **The value of a public signal equals the delay between it becoming public and being read.** A platform changelog is hours, a job posting is weeks, a competitor's two-star reviews are long, and an audience who knows you is not a record at all, so its delay is infinite. This sets the ranking of sources, decides which channels stage 4 should run, and explains why distribution comes first in the whole method. Ordering is the method. Stages 0 through 3 are free, offline, and take minutes, and they kill most directions. Stage 4 costs three orders of magnitude more. Run the cheap ones first, stop on a veto, do not proceed. ## Stage 0 · Ask about them before asking about the direction Three questions, the cheapest gate in the method. The first two have nothing to do with the direction, which is why they come before every question that does. **Question one: how many people can they reach directly right now, and who are they?** Get a number: how large the existing audience is, and how many can be reached directly within a week. No number means zero. If a direction is already on the table, check it once more: are these pe