ECC gets it built. This diagnoses why it is not selling once it is.

[maxxthemate-png]

Disclosure first: this is mine, and I am posting it here because the framing is complementary rather than competing.

ECC helps you build. What I built is for after that: the product works, it is deployed, and nobody is buying it. That is a different failure with a different diagnosis, and better engineering does not touch it.

why-isnt-it-selling is MIT and installs as a plugin. It asks five questions your code cannot answer, classifies one failure mode (traffic, activation, billing, or positioning), and returns ranked moves with a source URL on every claim. Billing is always item zero, because a product nobody can pay for has no conversion rate to diagnose.

Two things that make it worth a look even if you never install it:

It ships the negative results. Twenty candidate growth patterns went through an adversarial pass designed to refute them. Eleven died. All eleven are in the free tier with what killed them, and a few are advice you have definitely read approvingly. Publishing the graveyard is the whole point; every growth list you have seen contains only survivors, because nothing in the process tried to kill them.

It is honest about the ceiling. The most corroborated finding in the whole corpus is that distribution, not build capacity, is the bottleneck. The clearest case is a multi-agent build sprint that shipped 9 products and 280+ posts in 30 days and earned $0. Its own post-mortem: multi-agent systems can build anything, they cannot sell anything. That is an uncomfortable finding to package into a tool for people who like building, so I put it on the first screen instead of burying it.

No telemetry, no network calls: it declares allowed-tools: Read, Glob, Grep, AskUserQuestion, so it structurally cannot phone home.

Genuinely interested in people trying to break the method rather than just installing it.