Why stopping early is the cheapest decision
Dropping a weak hypothesis before you build costs a few weeks of research. Dropping it after launch costs months of work. The goal is not to talk yourself out of ideas; it is to find out what an idea has earned.
Seven warning signs
- Nobody can describe the last time the problem hurt.
- People complain but spend nothing: no money, no time, no workaround.
- Your main feature is already bundled into products buyers use. In the sample reports, an AI Gmail reply tool was already duplicated by Google and Superhuman.
- The one differentiator is unproven. The collision repair sample could not show that supplement omissions can be predicted before teardown, and the vehicle audio sample could not show 95% component accuracy from 30 seconds of phone audio.
- The price does not fit the alternatives. The contract renewal sample found no sufficient reason to add another tool at $750 per month.
- You cannot name the buyer who signs.
- You keep changing the idea after each bad result.
Pass, pivot, or stop?
Killgate's sample reports separate the outcomes. RESEARCH_PASS means the idea is worth direct buyer validation, not that it is ready to build. RESEARCH_PIVOT means the problem is real but this specific product is too close to existing ones, so a new hypothesis is needed. RESEARCH_FAIL means this hypothesis has not earned buyer validation or build investment yet, which is not the same as KILL.