How do you detect forced keywords and keyword stuffing?

Suggested searches unrelated to the article or to the ad that brought the visitor. The same few high-value terms appearing on pages about very different topics. High revenue per click with poor advertiser conversion on those terms. Terms supplied by the publisher that the feed repeatedly declines to show.

How do you prevent forced keywords and keyword stuffing?

Tie every supplied term to the page topic and to the wording of the incoming ad. Let the feed's own suggestions run alongside yours and compare. Review keyword reports for terms with high clicks and no downstream quality. Keep within the feed's limits on blocks and terms per page (RSOC policies).

Forced keywords and keyword stuffing: how it works, warning signs and defences

Forced keywords and keyword stuffing

The publisher loads the page with high-paying search terms that have little to do with what the visitor came for.

The short answer, from The Arbitrage Desk fraud map

How it works

On a related-search page the visitor sees a handful of suggested searches. The feed will suggest its own, and it also lets publishers propose terms through the terms parameter. Forcing keywords is legitimate when the terms fit the article. It becomes keyword stuffing when the terms are chosen for payout alone: an article about gardening whose suggested searches are all "mesothelioma lawyer" and "structured settlement".

The visitor may still click out of curiosity, but the advertiser on that expensive keyword receives someone who was reading about roses. Google says it reviews and filters publisher-supplied terms for relevance, engagement and performance, and in 2025 it limited new or restricted publishers to one related-search block per page with at most five terms.

Who pays for it

Advertisers in expensive verticals such as insurance, legal and finance, who pay premium prices for unrelated visitors. Publishers pay when revenue per click is discounted or the feed is restricted.

Who does it, and why

Arbitrageurs optimising purely on RPC. One high-paying keyword can earn several times a relevant one, which makes the temptation constant.

Warning signs

Suggested searches unrelated to the article or to the ad that brought the visitor.
The same few high-value terms appearing on pages about very different topics.
High revenue per click with poor advertiser conversion on those terms.
Terms supplied by the publisher that the feed repeatedly declines to show.

Defences

Tie every supplied term to the page topic and to the wording of the incoming ad.
Let the feed's own suggestions run alongside yours and compare.
Review keyword reports for terms with high clicks and no downstream quality.
Keep within the feed's limits on blocks and terms per page (RSOC policies).

An example

Illustrative: an article on "how to repot a houseplant" is given the terms "car accident attorney", "personal loan rates" and "online MBA". Suppose a relevant term pays $0.15 a click and the forced ones $1.50. Ten times the revenue per click looks like success for a week. Then advertiser data shows those clicks converting at close to zero, the price the feed pays for the account's clicks falls across the board, and the relevant terms earn less too.