PMax listing groups and feed signals
How Performance Max reads your product feed, listing groups, feed-driven targeting, and why feed quality matters even more when you can't see or steer the queries.
Performance Max runs your Shopping products off the same feed as everything else, but with almost no manual levers. Listing groups let you carve the catalogue into sets you can budget and target differently, and the feed's titles, categories and attributes do the actual query targeting. With no search-term report and no query controls, feed quality carries more weight in PMax than anywhere else.
PMax runs on your feed: with the levers removed
Performance Max serves your Shopping products from the same Merchant Center feed as Standard Shopping. The difference isn't the data source: it's how much control you keep. In Standard Shopping you have negatives, priorities and a full search-terms report. In PMax, almost all of that is gone: no positive keywords, no negative keywords at the campaign level to speak of, and only search category insights instead of the term-level report.
What remains? The feed, plus a small set of structural tools. That makes feed quality not just important but decisive: it's most of what you can actually influence.
Listing groups: how you carve the catalogue
Inside a PMax campaign, each asset group can target a listing group: a subdivision of your product feed. You build listing groups the way you'd build product groups in Standard Shopping, splitting on feed attributes:
| Split on | Use it to |
|---|---|
product_type / category |
Group by department for separate asset groups and creative |
| Brand | Isolate high-performing or brand-restricted lines |
| Item ID | Include or exclude specific products |
custom_label_0–4 |
Group by your own business logic: margin, lifecycle, seasonality |
Listing groups let you include or exclude product sets and route them to asset groups with matching creative. Crucially, PMax has no priority tiers: so separating products between PMax campaigns (or between PMax and Standard Shopping) is done at the product level via these listing groups and inventory filters, not by priority.
Custom labels are what make listing groups express your economics rather than just Google's categories: margin bands and lifecycle stages become the lines you carve along.
Feed signals: what PMax reads to target
With no keywords anywhere, PMax's product targeting leans almost entirely on feed data:
- Titles: still the primary query-matching signal. A weak title in PMax doesn't just lower click-through; it mis-routes the automated targeting itself, because Google has less to go on.
- Categories: a specific
google_product_categorygives PMax the context to match your product to the right intent. - Attributes: colour, size, material, gender, age group qualify the product and widen relevant matching.
- Identifiers: GTIN/MPN/brand drive catalog matching and richer placements.
- Images: PMax spans visual placements (Display, Discover, YouTube), so image quality does real targeting work here, not just Shopping thumbnails.
These are the same signals that drive feed quality and impressions in Standard Shopping: but in PMax there are no manual controls to compensate when the data is weak.
Deep dive Why feed quality matters MORE when you can't see the queries
In Standard Shopping, a weak feed is partly recoverable. Bad titles let junk queries through? Add negatives. Wrong product matching a query? Read the search-terms report, find it, sculpt it. The manual levers act as a safety net under imperfect data.
PMax removes the net. There's no term-level report to diagnose from, no negatives to cut waste query-by-query, no priority to reroute traffic. So every targeting decision is made by Google's models reading your feed: and if the feed is thin, the targeting itself degrades, silently, with no report that shows you why.
Concretely, in PMax a weak feed causes:
- Mis-routing: vague titles cause Google to match products to loosely relevant queries and placements, and you can't see or block them.
- Under-eligibility: attribute-light products are eligible for fewer relevant searches, capping reach with no error anywhere.
- Blended blame: aggregated reporting means you can't tell whether poor results come from the feed, the audience signals, or the creative. The feed is the most likely cause and the least visible one.
This is why accounts that clean their feed before scaling PMax consistently outperform accounts trying to fix PMax settings on top of weak data: there are barely any settings to fix. The lever is upstream. Enrichment strengthens titles, categories and attributes across the catalogue and writes them into a supplemental layer over your source feed, so PMax gets better targeting inputs without your original product data being touched. See the 20 enriched fields for what that covers.
Practical PMax feed checklist
- Get titles right first. They carry more weight in PMax than anywhere else, because nothing downstream can correct a bad match.
- Use specific categories. Top-level categories give PMax weak context; go deep.
- Complete the attributes the category demands: they widen relevant matching you can't otherwise influence.
- Build listing groups on custom labels, not just Google categories, so you can budget by margin and lifecycle.
- Separate products at the product level. No priority tiers exist in PMax: use listing groups and inventory filters to keep campaigns from overlapping. See Standard Shopping vs PMax for running both together.
In PMax, the feed isn't one input among many: it's the input you have. Invest there.
Frequently asked questions
What is a listing group in PMax?
A listing group is a subdivision of your product feed inside a PMax campaign's asset group, you split by category, brand, item ID, product type or custom label to create sets of products you can include, exclude, or (with the right structure) treat differently. It's PMax's equivalent of Standard Shopping's product groups.
Can I target keywords in PMax?
No. PMax has no positive keyword targeting for Shopping inventory. Its product targeting runs on your feed, titles, categories, attributes and images, plus Google's audience signals. The feed is the targeting.
If I can't see queries in PMax, how do I improve results?
You improve the inputs you can control, feed quality above all. Better titles, specific categories, complete attributes and clean identifiers directly improve how PMax matches and ranks your products. When you can't sculpt queries, you sculpt the data.