The controls that changed PMax
For two years PMax was a black box. Through 2025 Google opened it up, and we use every part of that. Campaign-level negative keywords expanded from 100 to 10,000 in March 2025, channel-level reporting now shows where spend lands, and full search-term reporting reveals what PMax is actually matching.
That matters because the most common PMax failure is brand cannibalisation: the campaign claims credit for searches people were always going to make. Brand exclusions and negatives stop it.
- Channel-level reporting to see Search vs Display vs YouTube spend
- Up to 10,000 campaign negative keywords (matched to Search)
- Brand exclusion lists to stop cannibalisation
- Asset groups split by audience and intent, not left as one blob
Feeding PMax clean signals
PMax is only as good as the conversion data it learns from. We wire first-party conversion tracking, exclude junk conversions, and where possible feed offline conversions so the algorithm optimises to real revenue rather than form fills. Without that, PMax happily buys cheap, worthless clicks.
On lead-gen accounts we point PMax at qualified leads, not raw enquiries, which keeps it away from the low-quality traffic that flatters a dashboard.
When PMax is the wrong answer
PMax is not always right. New accounts with little conversion history often do better on standard Search first, building data before handing it to the algorithm. We will tell you when a tightly run Search campaign beats PMax, rather than defaulting to whatever Google is promoting that quarter.
Common questions
Often yes, but only with the controls Google added through 2025: channel-level reporting, search-term data and up to 10,000 negative keywords per campaign. Run blind, PMax wastes budget on brand traffic and weak placements. Run with those levers and clean conversion data, it becomes a strong scale channel beside standard Search.