Google Ads — About Ad Rank

Google Ads Help, the platform’s own account of how it decides which ads show and in what order. Held as the first-party half of search-ad-auction; the theory half is gsp-auction-paper.

Tier T3, and the reason is the point. This is the seller of the auction describing the auction. It is the only authoritative statement of what the mechanism does today, and it is also promotional in places (“Higher quality ads can often lead to lower CPCs”). Usable for what is offered, per ../QUALITY.md; not usable as evidence that the mechanism works as claimed.

The six factors

Ad Rank, per this page, is set by:

1. Your bid — “the maximum amount you’re willing to pay for a click on your ad.”
2. Ad and landing page quality — “the usefulness and relevance of your ad and landing page.”
3. Ad Rank thresholds — “minimum thresholds that an ad must achieve to show.”
4. Auction competitiveness — how closely ranked ads compete.
5. Search context — “search terms, person’s location, device type, time of search.”
6. Ad assets — how formats and assets affect performance.

Quality is “summarized in your Quality Score,” visible in the account, and Google says its quality measures “help determine the Ad Rank thresholds for your ad, and whether your ad is qualified to appear at all.”

Read against the 2007 paper

Three things have changed since gsp-auction-paper described the mechanism.

Quality became the gate, not only a multiplier. In 2007 the quality score rescaled the bid; here a threshold can keep an ad out of the auction entirely, whatever the bid. That is a different lever: not “you pay more,” but “you do not appear.”

The auction is contextual. Location, device and time enter Ad Rank directly, so there is no single price for a keyword — the 2007 model’s one auction per keyword is now one auction per query-context.

The formula is no longer stated. The paper could write the mechanism as bid × γ. This page gives six named factors, one composite score and no weights. The authors noted the same trend in 2006 — Google’s pricing “was more transparent” early on, when quality score was just estimated click-through rate.

What this does and does not close

With gsp-auction-paper it closes the spoke’s coverage gap: the paid half now has the mechanism under b2b-ppc-metrics and groas instead of only the metrics on top of it.

What no source here supplies is any independent measurement of the auction — how much of a CPC difference quality score actually explains, or where the thresholds sit. That is the same shape as this spoke’s standing complaint about generative-engine advice: the party being optimized against is also the party publishing the rules.

Related

search-ad-auction · gsp-auction-paper · b2b-ppc-metrics · groas · google-search-essentials · google-guidance-seo-authority