Google Performance Max (often abbreviated PMax) is a fully automated, goal-based campaign type within Google Ads that serves ads across every Google-owned and operated channel — Search, Display, YouTube, Gmail, Google Maps, and Discover — from a single unified campaign. Advertisers supply creative assets, audience signals, and a conversion goal, and Google's machine learning systems handle placement, bidding, audience selection, and creative assembly automatically.
Performance Max replaced Smart Shopping and Local campaigns as the default campaign type for e-commerce and local advertisers in 2022, signaling Google's strategic commitment to AI-driven, cross-channel campaign management. The campaign type represents a fundamental shift from channel-specific buying to outcome-optimized delivery across the entire Google inventory ecosystem.
How Performance Max works
Asset Groups are the building blocks of a PMax campaign. Each asset group contains a set of creative inputs — up to 15 images, 5 videos, 5 headlines, 5 descriptions, and a final URL — that Google's system assembles into ads formatted for each channel and placement. Multiple asset groups can exist within a single campaign, typically organized around different product lines, audiences, or creative themes.
Audience Signals allow advertisers to provide directional guidance to the algorithm without hard constraints. Signals can include customer match lists, remarketing audiences, custom intent audiences, and demographic indicators. Google's system uses these signals as a starting point but is free to expand delivery beyond the suggested audiences when it predicts strong conversion probability.
[Smart Bidding](/glossary/smart-bidding) integration means all PMax campaigns run on automated smart bidding strategies — Target CPA, Target ROAS, or Maximize Conversions. There is no option to use manual bidding. The system optimizes bids in real time at each auction based on predicted conversion likelihood, user context, and competition.
Search themes (a newer feature) allow advertisers to provide keyword-level signals to guide search inventory targeting within PMax, partially restoring the search intent control that was lost when Smart Shopping campaigns were deprecated.
Budget optimization is handled automatically across all channels within the campaign. Google allocates spend toward the channel and placement combinations predicted to deliver the best performance against the campaign goal at any given moment.
How AI improves Performance Max
Performance Max is Google's most complete expression of AI-powered advertising, consolidating every optimization variable — bidding, placement, creative, audience, and budget — into a single automated system. The underlying models are trained on Google's aggregate conversion data across billions of queries and interactions, giving them signal richness that no individual advertiser can replicate.
Creative asset diversity is the primary lever advertisers retain. The more varied and high-quality the asset group inputs, the more effectively Google's system can personalize ad assembly for different users and contexts. Soku AI addresses this constraint directly by generating diverse headline, description, and image variations at scale, ensuring PMax campaigns have the creative breadth needed to optimize effectively across all channels.
Dynamic creative optimization is built into PMax's core architecture — the system assembles responsive ads dynamically, selecting and combining assets based on predicted performance for each specific impression opportunity.
Performance Max vs. standard Google Ads campaigns
Unlike traditional Search or Display campaigns where advertisers control keywords, placements, bids, and creative independently, PMax delegates all of these decisions to Google's algorithms. This produces better average outcomes for advertisers with sufficient conversion volume, but removes the granular control that experienced buyers use to protect against waste and surface specific insights.
The tradeoff is efficiency for control. PMax typically outperforms manually managed campaigns in conversion rate optimization and ROAS at scale, but makes it harder to understand exactly what is working and why — a challenge for advertisers who need channel-level or audience-level performance breakdowns.
Challenges and considerations
Reporting opacity is the most persistent criticism of PMax. The campaign provides limited visibility into which channels, placements, audiences, and asset combinations are driving performance. Advertisers cannot see search terms at the keyword level without supplementary Search campaigns, and channel-level spend breakdowns are only partially available.
Brand search cannibalization occurs when PMax captures search traffic that would have converted through lower-cost branded search campaigns. Without careful exclusion management, PMax can inflate costs by buying traffic the advertiser would have acquired more efficiently through other campaign types.
Creative asset requirements are substantial. PMax performs best with video assets — campaigns without video are limited in their YouTube reach. Producing quality video at the volume needed to supply multiple asset groups is a significant production challenge for many advertisers.
Learning period sensitivity means that budget changes, asset edits, or goal modifications reset the algorithm's learning phase. During the learning period, performance is typically volatile and CPA may spike before stabilizing.
Audience signal limitations mean that PMax may not honor exclusions with the same strictness as standard campaigns. Certain audience exclusions — particularly for sensitive categories — require explicit configuration that is easy to overlook during campaign setup.
