Asset Groups

Run one asset group for a brand selling five product categories and Performance Max has no clean signal to work with. The algorithm learns from a blended pool — buyers of your $30 accessories and buyers of your $200 hero product treated identically. Budget follows whatever converts fastest, which is usually not whatever has the best margin.

An asset group is a container: up to 15 headlines, 5 descriptions, 20 images, and 5 videos, plus the audience signals that give the algorithm a starting point. Google’s AI assembles elements from inside the group into ads and distributes them across Search, Shopping, YouTube, Display, Gmail, and Maps.

How it shows up in the wild

KEH Camera (used camera and gear retailer, US): Inflow restructured KEH’s Performance Max setup so each asset group carried camera-brand-specific audience signals — Sony shoppers in one signal set, Canon shoppers in another, with separate seed lists for one-time purchasers, loyalists, and high-value repeat buyers. Feed versus non-feed performance was tracked separately at the asset group level, giving real-time signal on which creative types drove returns for which buyer profile. In Q1 2023, ad revenue was up 76.3% and transactions increased 44.1% versus the same period running Standard Shopping, per Inflow’s Performance Max case study.

Rothy’s (sustainable footwear, D2C): Organizing video and imagery into Performance Max asset groups let a single campaign distribute Rothy’s sustainability message across Search, YouTube, Display, and Gmail simultaneously. Google’s AI selected the right format for each placement automatically, removing the need to manage separate campaigns per channel. Conversions grew 60% and revenue grew 59%, per Google’s Performance Max resource page.

Why it matters

Asset group structure is the primary lever D2C brands have for giving Google’s AI coherent learning signal. A single generic asset group for a multi-SKU catalog means the algorithm optimizes toward an average. An average buyer profile is often no buyer profile at all.

My hunch is that most underperforming PMax campaigns trace back to a single asset group carrying mixed signals. Fixing structure usually moves performance more than adjusting bids or budgets. Google’s guidance is 3–7 asset groups per campaign, with each group generating at least 20 conversions per month before you consider splitting further.

Related terms

• Performance Max
• Audience Signals (Performance Max)
• Smart Bidding
• Creative Testing
• Dynamic Product Ads

Frequently asked questions

How many asset groups should a D2C brand run per PMax campaign? Google recommends 3–7, segmented by product category or audience intent. If an asset group generates fewer than 5 conversions a month, the algorithm lacks enough signal to optimize — merge it with a related group rather than letting it drag the campaign’s learning phase.

What’s the difference between an asset group and an ad group? Ad groups live inside Search, Display, or Shopping campaigns and hold ads targeting specific keywords or placements. Asset groups live inside Performance Max campaigns and hold creative elements that Google’s AI distributes across all channels automatically. The concepts are structurally similar but operate in entirely different systems.

Can I run separate asset groups for prospecting versus retargeting? Audience signals are suggestions. The algorithm decides who actually sees what — there is no hard audience separation inside PMax the way there is between ad sets on Meta. Brands expecting clean prospecting/retargeting splits inside a single PMax campaign are typically measuring a behavior the platform doesn’t support.