How much PMax reach never gets a chance to convert

Performance Max expands into search categories automatically, and expansion is the point of the campaign type -- reaching adjacent demand a narrower campaign would miss. The failure mode is not expansion itself, it is expansion that keeps costing impressions in categories that never convert.

The pattern worth watching

Several search categories, each individually unremarkable, that together account for a meaningful share of a campaign's total impression volume while contributing zero conversions between them. No single category looks alarming; the sum does.

A worked example

Illustrative case: a search-category breakdown for a PMax campaign turns up nine categories, none individually accounting for more than 3 percent of impressions, together accounting for roughly a quarter of total impression volume across the quarter -- and zero conversions among all nine, quarter over quarter. No single line item would trigger a review on its own; summed, a quarter of the campaign's reach is going somewhere that has now proven, repeatedly, not to convert.

Why this is a volume question, not a spend question

Google does not report cost at the search-category level inside PMax, the same limitation behind the language-leakage pattern -- so this is a volume question, not a spend question: reach that never gets a chance to convert, sized in impressions, not euros. The distinction that actually matters: PMax testing new territory and finding nothing is different from PMax repeatedly returning to territory it has already proven does not convert. The second is what deserves attention.

What repeated testing costs, even without a euro figure

A category tested once and dropped is the system doing exactly what it is supposed to do. The same category showing up quarter after quarter, still at zero conversions, is not exploration anymore -- it is a standing allocation of impression volume that could otherwise reach a category that actually converts, and the only way to tell the two apart is to look at the same breakdown more than once.