Sources
Where Our Claims Come From
RankOps runs on one sourcing rule: a claim we can't verify doesn't ship. This page lists where our numbers come from — the published scorecard, owner-verified case studies, live proof pages anyone can click, and our own tracking data — so you can check us the way AI engines check everyone.
Our evidence, itemized
- The 17-signal scorecard — published in full, weights included; the same deterministic checker runs the free score, the client dashboard, and our own daily self-audit
- Case-study numbers — client-verified: Triple B's Mechanics figures come from the owner's own public Google review, quoted verbatim; the undeployed audit from documented health checks on the live site, published anonymously
- Live proof — 241 neighborhood pages across 25 NC cities, clickable rather than claimed
- Tracker evidence — every "cited / not cited" verdict stores the engine's actual answer text; failed checks are shown as failed, never smoothed into results
- Industry statistics — used only with source and date attached in the content where they appear; no orphaned "studies show" claims
What we deliberately don't claim
Guaranteed rankings, guaranteed AI citations, or specific outcomes for your business — nobody controls AI outputs, and results vary by market. What we publish instead is method, prices, and verifiable befores-and-afters. Questions about any claim on this site: ask Tyler directly.
Questions
How does RankOps source its claims?
One rule: a claim that cannot be verified does not ship. Numbers come from the published scorecard, owner-verified case studies, live proof pages anyone can click, and RankOps' own tracking data, which stores the engine's actual answer text.
What does RankOps deliberately not claim?
Guaranteed rankings, guaranteed AI citations, or specific outcomes for your business. Nobody controls AI outputs, and results vary by market and starting point. An agency promising otherwise is describing something it cannot deliver.
Are failed checks hidden?
No. Every cited or not-cited verdict stores the engine's actual answer text, and failed checks are shown as failed rather than smoothed into results. A log that only shows wins is a marketing asset, not evidence.