Signal collection
SignalForge starts with public sources: support threads, complaints, trend spikes, review pain, pricing complaints, workflow failures, and repeated operator questions.
Method
The goal is to make the process feel trustworthy, structured, and commercially serious. Not every interesting pattern deserves to become a business brief, so the system is designed to filter first and publish second.
Trust markers
Public inputs
Complaints, reviews, trends, and operator pain
Score dimensions
10 structured criteria per opportunity
Editorial pass
Human review before anything is published
Methodology
SignalForge starts with public sources: support threads, complaints, trend spikes, review pain, pricing complaints, workflow failures, and repeated operator questions.
We group repeated pain into opportunity clusters so one loud anecdote does not masquerade as a business.
Each cluster is scored across pain intensity, frequency, urgency, spend intent, workaround ugliness, trend strength, competition weakness, build simplicity, audience clarity, and execution speed.
A human-curated pass trims hype, removes weak evidence, tightens the wedge, and keeps the brief useful for an actual builder instead of a trend tourist.
Ideas only publish when the evidence is strong enough, the buyer is clear enough, and the MVP looks feasible for a solo builder or compact operator team.
Principle
The brief should show why the opportunity matters before it tries to sound clever.
Principle
Ideas need a believable first version and a sensible buyer, otherwise they stay out of the archive.
Principle
Every publishable brief should point toward validation, pricing, GTM, and a next step.
Score bands
The score is there to support judgement, not replace it. A strong score still needs clear evidence and a believable wedge.
| Band | Verdict | Meaning |
|---|---|---|
| 90–100 | Exceptional | Publish immediately if evidence is solid. |
| 80–89 | Strong | Excellent candidate for Idea of the Day. |
| 70–79 | Promising | Worth tracking, but needs more validation. |
| 60–69 | Borderline | Interesting, though not sharp enough yet. |
| Below 60 | Hold | Do not publish unless there is a strategic reason. |