Almost every platform change we have covered this month traces back to one asset. Google raised its Merchant Center image requirements, began reporting on AI-created product titles, and keeps feeding Performance Max from catalogue data. ChatGPT pinned a shopping tab that reads merchant product information. Social commerce platforms match products to viewers using the same structured fields. Your product feed stopped being a technical chore and became the raw material for paid, organic and assistant-led discovery at once. This complete guide covers product feed optimisation for ecommerce: what the feed is, which fields matter most, how to structure titles and attributes, and the review routine that keeps it healthy.

Product feed optimisation is the ongoing work of improving the structured product data a brand sends to platforms: titles, descriptions, attributes, images, prices, availability and identifiers. It matters because these systems no longer match ads to keywords you chose, they match products to demand using the data you supply, which makes the feed the closest thing modern ecommerce has to a universal targeting control.

Why Is the Feed More Important Than Campaign Settings Now?

Because automation ate the settings. Targeting, bidding, placements and increasingly creative are decided by the platforms, as we covered in our Google AI Max guide, which leaves budget, measurement and data as the levers an advertiser still owns. Of those, the feed is the one that determines what is even possible: a product with a vague title and a poor image cannot be matched to the searches or conversations where it would sell, however elegant the campaign structure around it. Across the accounts we audit at Webtopia, feed quality explains more performance variation between similar brands than campaign architecture does.

Which Fields Actually Move Performance?

In rough order of impact. Titles first, because they are the primary matching signal in Shopping and a heavy influence in AI descriptions. Images second, the deciding factor in whether a listing earns the click, and now governed by rising minimum standards. Product type and Google product category third, since miscategorised products compete in the wrong auctions. Identifiers next, GTIN and brand, which make richer listing formats and comparison eligibility possible. Then price and availability accuracy, where errors cause disapprovals and wasted spend on out of stock items. Then variant attributes, size, colour, material, pattern, age group and gender, which matter disproportionately in fashion and beauty. Descriptions come last for matching but first for assistants, which read them when explaining your product to a shopper.

How Do You Write Titles That Work?

Write for how customers ask, not for how your ERP labels. Front-load the important information in a consistent pattern: brand, product type, then the distinguishing attributes that a buyer would actually specify. A title reading Organic Cotton T-Shirt, White, Mens, Medium tells every system what it needs; one reading TS-4471-WHT-M tells them nothing. Three rules keep titles honest. Consistency across the catalogue, so performance differences reflect products rather than formatting. No promotional language, since Sale and Free Shipping in a title look like spam to a classifier and date instantly. And no keyword stuffing, because repetition dilutes the signal and reads badly in the AI-generated summaries that increasingly quote your data.

Google has begun reporting on AI-created product titles, which is worth watching closely: platforms will happily rewrite weak titles for you, and the resulting text becomes part of how customers perceive your brand. That is a governance question as much as a performance one, the theme of our AI creative governance guide: better to supply titles worth keeping than to inherit generated ones.

How Should You Handle Images?

To the recommendation rather than the minimum. Google's new floor is 500 x 500 pixels from 31 January 2027, with 1500 x 1500 recommended, and the practical guidance plus an audit routine sits in our Merchant Center image requirements post. Beyond size: neutral backgrounds on the main image, the product filling the frame without cropping, no overlaid text or watermarks, consistent aspect ratios, and additional images that show scale and detail. Assistants and visual search reuse these files, so an image built for a small website thumbnail is now a liability in three channels rather than one.

How Do You Structure a Feed Across Channels?

One source of truth, adapted per destination. Maintain the canonical catalogue in your store or a feed management tool, then use feed rules and supplemental feeds to adjust per channel: different title patterns where a platform's matching behaves differently, mapped categories per destination taxonomy, and channel-specific image crops where required. What you should not do is maintain parallel manual copies, because they drift within weeks and the drift always surfaces as a disapproval during peak trading. Segmentation matters too: split the catalogue into product groups by margin, bestseller status and seasonality, so campaigns can be steered towards the products worth selling rather than treating a thousand SKUs as equals, the approach behind our Google Shopping guide.

What Does a Feed Health Routine Look Like?

Three cadences. Weekly: check diagnostics for disapprovals and warnings and clear anything affecting a revenue-generating product, because a silently disapproved bestseller is the most expensive fifteen minutes you will ever save. Monthly: review titles, images and attributes for your top twenty products by revenue, and check price and availability accuracy against the store. Quarterly: audit the full catalogue, prune products that no longer sell, refresh categorisation, and confirm your feed rules still match each platform's current requirements. Add an alert for feed fetch failures, since a broken sync is invisible in campaign reporting until performance collapses.

How Does Feed Quality Show Up in the Numbers?

Indirectly, which is why it gets neglected. Better titles and categorisation raise impression share on relevant queries; better images raise click-through rate; accurate availability cuts wasted spend; complete attributes improve matching quality, which shows up as better conversion rates rather than more traffic. Watch impression share, click-through rate and conversion rate by product group rather than at account level, and judge the work on new customer CAC and MER over a quarter, the measurement frame from our analytics guide. Feed work rarely produces a dramatic week; it produces a structurally cheaper account.

The Bottom Line on Feed Optimisation

Your feed is now the ad, the answer and the shop window. It decides what Google can show, what Performance Max can build, what an assistant can say about your product and what social commerce can match. Treat it as marketing infrastructure with a named owner and a review cadence, not as a technical file someone set up once, and it will quietly improve every channel you run.

Keep Up With Feed Requirement Changes

Image minimums, title reporting, category taxonomies and destination rules all changed this year, mostly by email. We track every platform change that touches DTC spend, deadlines flagged, and send the ones that matter each Tuesday in Beyond the Clicks. Sign up at webtopia.co/newsletter.

Want Your Feed Audited?

If nobody owns your feed, book a call. Our google ads agency team will audit titles, attributes, images and diagnostics and show you what it is costing you, part of how we run growth as an ecommerce marketing agency.

Frequently Asked Questions

What is product feed optimisation?

Product feed optimisation is the ongoing work of improving the structured product data a brand sends to platforms: titles, descriptions, attributes, images, prices, availability and identifiers. Shopping, Performance Max, social commerce and AI assistants all match products using this data.

Why does the product feed matter more than campaign settings?

Because automation has taken over most campaign settings. With targeting, bidding and placements automated, the feed is one of the few inputs an advertiser still fully controls.

How should you write product titles for a feed?

Write them the way customers search, front-loading brand, product type and distinguishing attributes such as size, colour or material. Avoid internal SKU language, promotional text and keyword stuffing.

Which feed attributes have the biggest impact?

Titles and images do the heaviest lifting, followed by product type and category, GTIN and brand identifiers, price and availability accuracy, and variant attributes such as size and colour.

How often should you review your product feed?

Check diagnostics weekly, review titles and attributes for top revenue products monthly, and run a full catalogue audit quarterly.

Does the same feed work for every channel?

The core data should be one source of truth, but destinations differ, so use feed rules or supplemental feeds to adapt titles, categories and images per channel rather than maintaining separate copies.