How SignalForge reasons from evidence
A deliberate path from a market question to a commercial hypothesis—and then to the cheapest useful test.
The evidence set is deliberate
SignalForge retrieves candidates through multiple research paths. Deterministic anchor checks classify each source as Direct, Adjacent, or Weak. You choose which sources become evidence; newly retrieved candidates are never silently selected.
- Direct: clearly about the target market, buyer, or workflow.
- Adjacent: useful broader or analogous context.
- Weak: generic overlap or insufficient anchor evidence.
Observation stays separate from inference
Observed claims should be directly supported by selected evidence. Synthesized patterns combine signals across sources. Hypotheses are commercial inferences that still require validation. User-defined scope—such as Germany or English—is context, not something sources independently proved.
Counter-evidence is part of the method
SignalForge surfaces missing evidence, contrary cases, existing alternatives, and reasons a hypothesis may fail. Vendor-authored claims remain attributed rather than becoming observed market facts.
Strength and payment evidence remain qualitative
Opportunities use Strong, Moderate, or Exploratory evidence wording and Direct, Indirect, or Missing payment evidence. These are not proprietary numeric scores. Specific prices, sample sizes, and thresholds in a Validation Plan may be AI-proposed starting points unless evidence explicitly supports them.
What the method does not claim
- YouTube engagement is not demand.
- Comments are not payment evidence by themselves.
- Popularity is not profitability.
- SignalForge produces hypotheses, not guarantees.
- The user's source selection and judgment remain material.