Google Shopping and Performance Max don't read your website — they read your product feed. Weak titles, missing GTINs, poor categorisation and thin images cap your visibility and Quality regardless of bids, so feed quality is the ceiling your bidding strategy can never lift you past.
Book a free strategy callWith Search, you write the ad copy. With Shopping and Performance Max, Google builds the ad out of your product feed — the title, image, price, availability and attributes you send to Merchant Center. There’s no headline field to optimise, no description you craft by hand for the auction. The feed is the creative.
That means every weakness in your product data becomes a weakness in your ads. A vague title doesn’t just look untidy in a spreadsheet — it directly limits which searches your product can match. A missing GTIN doesn’t just fail a data check — it can restrict how prominently Google is willing to show that product at all. This is why I keep coming back to feed quality with ecommerce clients before touching bids: you can have a technically well-run campaign sat on top of feed data that’s quietly capping what it’s able to achieve.
It’s also why feed problems are so easy to miss. Nothing in the campaign interface flags ‘your titles are too vague’ or ‘your images are underperforming Google’s guidelines’ — you have to go and look at the feed itself, product by product, against how people actually search. Most accounts I look at have never had that review done.
If I had to pick one field to fix first, it’s the title. Google’s matching for Shopping and Performance Max weights the title heavily when deciding which searches a product is eligible to show against, which means a title written for your website (‘Handcrafted Oak Coffee Table — The Willow Collection’) is doing a different job to a title written for search matching (‘Oak Coffee Table 120cm Willow Collection Solid Wood’).
The practical fix is to lead with the attributes shoppers actually type — product type, brand, then the differentiating details (size, colour, material, model) — and drop marketing language that doesn’t map to a search term. This isn’t about stuffing keywords; it’s about structuring the title the way a shopper’s query is structured, so Google’s matching has something concrete to work with.
Worth checking your search terms report — Performance Max now has its own, alongside Standard Shopping and Search — against your current titles. If the phrases people actually search don’t appear anywhere in the title field, that’s a straightforward, fixable gap — not an algorithm problem, a data problem.
I see accounts where someone has spent weeks tightening tROAS targets on a Performance Max campaign that’s still capped by feed data — thin titles, no GTINs, a handful of products miscategorised into the wrong Google product category. No amount of bid strategy tuning recovers volume the feed itself is blocking. If Shopping or PMax performance has plateaued, the feed is usually the first place I look, before I touch a single bid.
It’s an easy trap because bidding changes feel like progress — you adjust a target, watch the graph, get a result within days. Feed fixes are slower and less visible in the interface, which is exactly why they get skipped, and exactly why they’re often the actual bottleneck.
Feed optimisation isn’t a one-off clean-up — it’s an ongoing discipline that has to sit alongside how you structure and sculpt Performance Max campaigns. This is one of the areas my agency, Growth PPC, specialises in for ecommerce and D2C accounts: feed optimisation and PMax sculpting go hand in hand, because custom labels built at the feed level are what make it possible to separate high-margin or bestselling products from the rest inside a campaign, rather than letting the algorithm treat every product the same.
That separation matters more than it sounds. Without it, PMax optimises toward whatever converts most easily — which is often your cheapest, thinnest-margin products — because that’s the only signal it has. With custom labels feeding margin or priority data into the feed, you can build campaign structures and bidding around what’s actually profitable, not just what’s easiest to sell.
Get the feed right first. It’s less exciting than a bidding tweak, but it’s the part of the account that determines what the rest of your Shopping and PMax strategy is actually allowed to achieve.
Prices, stock and availability need to sync as close to real-time as your platform allows — daily at an absolute minimum, ideally via an automated feed rather than a manual upload, since stale availability data leads to disapprovals and wasted clicks on out-of-stock products. Titles, images and categorisation don’t need daily changes, but should be reviewed whenever you add new products or notice a category underperforming.
Most products with a genuine manufacturer barcode should include it — it’s one of the clearest ways to help Google match your listing to its own product data. If a product genuinely has no GTIN (handmade or custom items, for example), Google allows you to mark it as exempt rather than leaving the field blank, which avoids the disapprovals that come from simply omitting it.
Standard Shopping campaigns only use your product feed and show in the Shopping tab, Google Search, Google Images and Search Partner sites. Performance Max uses the same feed but can also serve across Display, YouTube, Gmail, Discover and Maps from one campaign — which means feed quality matters even more in PMax, because it’s feeding a wider range of ad formats and placements, not just product listings.
Yes. Missing required attributes, mismatched prices between your feed and landing page, and policy violations in images or titles are common causes of item-level disapprovals. These reduce your active product count without necessarily showing up as an obvious problem until you check Merchant Center diagnostics directly.
Custom labels (custom_label_0 to custom_label_4) are five free-text fields you populate in your feed to segment products by whatever matters to your business — margin band, bestseller status, season, clearance stock. They’re not required, but without them you’re limited to letting Performance Max treat every product with roughly equal priority, which rarely matches your actual commercial priorities.
Start in Merchant Center diagnostics for disapprovals and warnings, then check your active vs. total product count — a large gap points to a feed issue before a bidding one. If products are active and approved but still underperforming, look at title structure, category accuracy and image quality against competitors’ listings for the same search terms before assuming the bid strategy needs adjusting.
Feed optimisation and PMax sculpting for ecommerce is one of the areas I focus on, through my agency Growth PPC as well as direct client work. Book a free strategy call and I'll give you an honest read on where your product data is capping performance.
Call 07410 907 104 Book a free strategy call