RankAssurance
Owner Tests · Foundations · Updated 18 September 2026

The Four Gates.

Found, Matched, Believed, Clean — the Order Every Test Uses

The Short Answer

The Four Gates are the order search visibility is judged in: Found, Matched, Believed, Clean — skip a gate and the next one does not save you.

Four gates a page passes, in order, before a ranking — or an AI answer — can happen: Found, Matched, Believed, Clean. Skip a gate and the next one does not save you. This page is the frame's single definition.

The Question This Page Answers

Every test in this library leans on one frame: four gates a page passes, in order, before a ranking — or an AI answer — can happen. This page is the frame’s single definition. When a test says “gate 3,” this is what it means.

Skip a gate and the next one does not save you.

The Four Gates, in Order

Gate · The question · What it means
1. Found
Can the crawler find and fetch the page?A URL that cannot be reached, or tells readers to stay away, never enters the race. Nothing downstream matters until this opens.
2. Matched
Does the page answer the query?Fetchable is not the same as relevant. The page has to be about the thing a person asked, in a form a person can use.
3. Believed
Do independent pages treat it as a source?Mentions you do not control are the vote of confidence. A page can be found and matched and still lose to a page more of the web believes.
4. Clean
No spam patterns, no manual actions?The named policy objects — link schemes, scaled content, the rest — suppress what the first three gates earned.
The four gates search visibility is judged in, in order: Found — the crawler can reach and fetch the page cleanly; Matched — the page answers the query typed; Believed — independent pages treat it as a source on their own; Clean — no spam patterns and no manual action. Skip a gate and the next one does not save you.
The four gates in order. Skip a gate and the next one does not save you.

Why the Order Matters

The gates are a sequence, not a menu. A blocked page does not have an authority problem — it has a fetch problem, and buying mentions for it buys nothing. A perfectly clean page that answers no real query has nothing for gate 3 to believe. Diagnosis in this library always means: find the first gate that fails, fix that, and only then look downstream. That is the Ranking-Problem Test’s whole method.

The Same Gates on the AI Surface

When the surface is an AI answer instead of a results page, the gates do not change — the prize does. Reachable and readable is still gate 1; being about the buyer’s actual question is still gate 2; and the independent record — who cites you, who names you — is what answer engines lean on for gate 3. The Citation-Surface Test reads the whole frame on that surface.

Where Each Gate Gets Tested

GateThe tests that own it
FoundFirst-Week · First-Index · Redirect-Map · Preview-Control
MatchedTopic-Fit · Content-Quota · E-E-A-T · Helpful-Content
BelievedAuthority Gap · Earned-Mention · Link-Quality · Two-Site
CleanSuppression Type · Sister-Site · Would-This-Exist · Retired-Tactic

Questions

Is the Four Gates Frame Google’s?

No. It is this library’s editorial frame for reading the official documents in order — fetchability, relevance, independent belief, and the spam policies are all Google’s published concepts; the gate sequence is ours.

Which Gate Do I Check First?

Always gate 1, always in order. The Ranking-Problem Test is the ordered walk; stop at the first fail.

Do the Gates Apply to AI Answers?

Yes — unchanged. The surface changes what winning looks like (being named or cited instead of ranked), not what has to be true first.

Sources