Why evidence and provenance matter

Marketing advice is easy to find; marketing advice you can trace is rare. Signal & Science labels what it observed, how confident it is, and what it could not verify, so you do not mistake a guess for a fact. This transparency is deliberate: a system that admits "Unknown" is one you can trust when it says "Observed."

Before acting on any recommendation, and before quoting any number from the product to a colleague or client.

You have read Diagnosis, plan, execution, and outcome are four different questions and How the Marketing Plan page works.

1. 1Open the Diagnostic page and find the confidence label (high, medium, low) on an observed area. Then check the evidence basis of the run as a whole: a banner above the report appears when the run was a direct scan only or found no observable evidence, and Diagnosis History shows each run's "Evidence map" as "Grounded read", "Direct scan only", or "Unavailable". Together they tell you how much weight to give a claim.
2. 2Find an Unknown. Confirm that it says the system could not verify enough, and notice that it does not score you zero, rank you, or call you weak.
3. 3Open the Marketing Plan page and read "What changed and why." Notice the careful wording: "what we think drove it" is a hedged hypothesis, not a proven cause.
4. 4Check for "Suggested changes to your plan." It appears only when there is a suggestion; when it is absent, the system has nothing better to propose rather than inventing something. That restraint is intentional.
5. 5Look at "Recorded task results": each entry shows when it was recorded, its verdict, and its source. A task still inside its window shows "Task completed. Measurement pending." on its card instead. Pending is a legitimate state, not a missing report.

Observed areas show their confidence and the sources behind them, and the run states its evidence basis. Where evidence is missing, you see Unknown rather than a guessed number or a placeholder that looks like customer data.