Product Feed Management
Your product data decides whether Google shows your ads or someone else's. Most businesses have no idea how much this costs them.
The invisible layer that controls your Shopping performance
Every time someone searches for a product on Google, an auction happens. Google looks at all the advertisers selling that kind of product and decides which ones to show. The bids matter. The budget matters. But there's something that matters more than both of those combined: your product data.
Your product feed — the file that tells Google what you sell, what it looks like, what it costs, and what to call it — is the single most important factor in whether your Shopping ads and Performance Max campaigns succeed or fail. If your feed says “Blue Widget 500ml” and a customer searches “reusable water bottle for gym”, Google won't make the connection. Your product won't show. You won't even know the opportunity existed.
I'm Peter Empson. I've been optimising product feeds for UK e-commerce businesses for over 15 years, across catalogues from 50 products to 50,000+. Feed management is where I start with every Shopping and Performance Max client, because nothing else works properly until the data is right.
Same product, different results
What the platform generates
Lamp Black Table
What I'd write
Modern Black Table Lamp Adjustable Metal Desk Light
The second title matches searches for “black desk lamp”, “modern table lamp”, “adjustable desk light”, and “metal table lamp”. The first matches almost nothing.
Why your product feed matters more than your bids
Most businesses focus on bidding strategy, budget allocation, and campaign structure when they think about Google Ads performance. Those all matter. But they're secondary to a more fundamental question: is Google even showing your products for the right searches?
The feed is where that decision gets made. Google's algorithm reads your product titles, descriptions, categories, and attributes to understand what you sell. It uses that understanding to match your products against search queries. If the data is thin, generic, or poorly structured, your products enter fewer auctions, show for fewer searches, and lose to competitors whose data is better — regardless of how much you're willing to bid.
In my experience, feed restructuring alone — before touching a single bid or budget — typically produces a significant increase in how often products appear. Same products. Same prices. Same budget. The only difference is how well Google understands what you're selling.
This is also increasingly important for AI-powered shopping. Google's AI — whether in Performance Max, Shopping ads, or the new AI search experiences — makes decisions based on your product data. As Google leans harder into machine learning, the quality of your feed becomes even more decisive. Poor data doesn't just mean fewer impressions today; it means the AI learns the wrong things about your products and makes worse decisions over time.
What feed optimisation actually involves
It's not glamorous work, but it's high-impact work. Here's what I focus on, roughly in order of how much difference it makes.
Product titles — the biggest lever you have
Titles are what Google weighs most heavily when deciding which searches trigger your products. Most feeds use internal product names or manufacturer part numbers. I restructure titles around the language your customers actually use: brand, product type, key attributes (size, colour, material), and use case. This single change typically produces the largest improvement in impression volume.
The structure varies by product type. A clothing retailer needs colour and size front-loaded. A furniture retailer needs material and style. A technical product needs specifications. I tailor the formula to your catalogue and your customers.
Descriptions that convert, not just exist
Many feeds either have no descriptions at all, or use manufacturer boilerplate that reads like a spec sheet. Google uses descriptions to understand context and relevance. I write descriptions that are keyword-rich but natural — covering use cases, benefits, and the kind of language that appears in search queries. For large catalogues, I build templated approaches that scale without sacrificing quality.
Categories and product types
Google has its own product taxonomy — over 6,000 categories. Getting the right category means your products appear in the right context. Getting it wrong means competing in the wrong auctions or being excluded entirely. I map your catalogue to Google's taxonomy properly, and set up product type hierarchies that give you granular reporting and targeting control.
Missing and incorrect attributes
Colour. Size. Material. Gender. Age group. Pattern. These attributes feel like admin, but they're filters that shoppers actively use — and Google uses them to refine matching. A feed with missing attributes enters fewer auctions. A feed with incorrect attributes shows products to the wrong people.
I audit every required and recommended attribute for your product type, fix what's wrong, fill in what's missing, and set up processes so new products get added properly from the start.
Custom labels for bidding control
Custom labels are how you tell Google things about your products that aren't part of the standard feed — margin band, seasonality, bestseller status, clearance. These labels let me segment your campaigns by profitability, so high-margin products get the investment they deserve and low-margin products don't eat your budget. Without them, every product competes equally for the same budget, which is almost never what you want.
Merchant Centre: the bit nobody watches
Google Merchant Centre is where your feed lives. It's also where problems accumulate quietly until something breaks. Product disapprovals. Policy warnings. Feed processing errors. Outdated pricing. Missing images.
Most agencies treat Merchant Centre as “the client's problem.” Issues sit for days or weeks before anyone notices. By then, products have been suppressed, impressions have dropped, and revenue has been silently lost.
I monitor Merchant Centre as part of my ongoing management. Disapprovals get resolved within 24 hours. Feed errors get fixed before they compound. Policy changes get addressed proactively, not after Google suspends your account.
Common Merchant Centre problems I fix
- Products disapproved for missing identifiers (GTIN, MPN, brand)
- Price mismatches between feed and website
- Shipping configuration errors blocking entire regions
- Image quality rejections (too small, watermarks, promotional overlays)
- Policy violations from description content or product claims
- Feed processing failures that silently suppress products
Your feed is becoming your AI strategy
Google's direction is clear: more AI, more automation, more decisions made by algorithms rather than humans. Performance Max already uses machine learning to decide where and when to show your products. Google's AI-powered search experiences are starting to recommend products directly in conversational results.
In all of these systems, the quality of your product data is what the AI has to work with. Good data means the AI makes good decisions — showing your products to the right people at the right time. Poor data means the AI makes poor decisions — and you have no lever to fix it other than fixing the data.
Weak feed + AI
AI shows your products to broad, low-intent audiences. Budget spread thin across channels you can't see. Poor match between product and search query.
Strong feed + AI
AI understands your products properly. Matches to high-intent searches. Budget concentrated on channels that convert. Competitive in AI-powered shopping experiences.
This isn't theoretical. Businesses that invest in feed quality now are building a compounding advantage — one that gets stronger as Google's AI gets more sophisticated. The ones still running on auto-generated titles and minimal attributes are falling further behind with every algorithm update.
I work with whatever platform you're on
Your product feed starts with your e-commerce platform, but it shouldn't end there. Platform-generated feeds are a starting point — they get the basic data out, but they rarely produce the optimised titles, complete attributes, and custom labels that drive real performance.
I work with all major platforms and know the strengths and limitations of each:
For most clients, I use supplemental feeds to override and enhance the platform-generated data without changing anything on the website itself. This means I can optimise titles, add custom labels, fix categories, and fill in missing attributes — all without your development team needing to touch anything.
Frequently asked questions
How do I know if my feed needs work?
If you're running Shopping or Performance Max campaigns and you've never had your feed independently reviewed, it almost certainly does. Common signs: low impression share despite reasonable bids, products not showing for obvious search terms, Merchant Centre warnings you've been ignoring, or the same generic titles your platform generated automatically.
Can you optimise feeds for large catalogues?
Yes — that's where feed optimisation has the most impact. For large catalogues (1,000+ products), I use templated approaches and supplemental feeds to make systematic improvements at scale, then manually refine the top-performing products. I've worked with catalogues up to 50,000+ SKUs.
Do I need to change my website?
Usually not. I use supplemental feeds to override and enhance your product data without touching your website or product pages. The optimisations happen at the feed level, so your development team doesn't need to be involved.
How long before feed changes make a difference?
Google typically processes feed updates within a few hours. You'll see changes in impression volume within days, and meaningful performance impact within 2–4 weeks as Google re-evaluates which auctions your products qualify for.
Is feed management a one-off or ongoing?
Both. There's an initial audit and optimisation phase — typically 2–4 weeks depending on catalogue size — where I fix the structural issues and optimise titles, descriptions, and attributes. After that, ongoing management keeps the feed healthy as you add new products, prices change, stock fluctuates, and Google's requirements evolve. I include feed management as standard for all my Shopping and Performance Max clients.
What about feed management tools like Channable or DataFeedWatch?
Tools like these are useful for managing the technical side of feed delivery — rules, transformations, multi-channel distribution. But a tool doesn't know what a good product title looks like for your market, or which attributes are missing, or how to structure custom labels around your margins. The tool handles the plumbing. I handle the strategy and the product data quality that actually drives performance.
Want to know what your feed is costing you?
I offer a free feed review for any business running Google Shopping or Performance Max. I'll look at your product data, identify the biggest opportunities, and give you an honest assessment of what better feed quality would mean for your campaigns.