Claims We Refuse to Make
The claims HubSEO deliberately never makes. Truth-first means knowing where evidence ends.
HubSEO reports measured evidence, documented provider facts, and clearly labelled interpretation. It does not turn a proxy, simulation, heuristic, or model response into a fact. This page lists the claims the product deliberately refuses to make.
Identity and permission
HubSEO does not claim that:
- GPTBot is ChatGPT Search, or that ClaudeBot is Claude user retrieval.
- Training permission controls search visibility.
- Google-Extended controls Google Search inclusion.
- A spoofed user-agent request proves a real provider crawler was blocked.
- Robots permission proves a crawler visited.
- A crawler log proves provider identity without supported IP or DNS validation.
- A crawler hit produced an answer citation, referral, ranking, traffic, conversion, or revenue outcome.
Files and formats
llms.txt,llms-full.txt,skill.md, agent cards, or similar experimental files automatically improve rankings or citations. They are readiness signals, not proven levers.
Interpretation and outcomes
- AI-generated interpretation is measured technical evidence. It is labelled interpretation.
- A merged pull request changed production or fixed a finding. Only the verification re-scan closes a finding.
- A strong aggregate score means no critical blocker exists. Read the findings.
- Lab measurements are field Core Web Vitals, traffic, rankings, or business outcomes. They are lab measurements.
- An estimated forecast is real customer traffic, conversions, citations, or revenue. Forecasts are labelled forecasts.
The vocabulary of honesty
When evidence is unavailable, the product uses unknown, unavailable,
not observed, not verified, or pending. Never a guessed number in
disguise.
Why this page exists
Most SEO tooling lost trust by claiming outcomes it could not measure. The refusals above are not marketing hedging; they are the product's operating constraints, enforced in code: zero-mock invariants in the test suite, missing-data rules in every score, and statuses that distinguish measured from absent. See How scores are computed and Security.