GOOGLE SHOPPING FEED: TITLE AND ATTRIBUTES TO INTERCEPT THE RIGHT INTENT

Most explanations of the google shopping feed treat it as a data export chore: dump your catalog, map the columns, upload, done. That framing is why so many accounts I inherit have technically valid feeds that still barely show. In my experience the feed is not a passive inventory list; it is the targeting layer. In Search you tell Google which queries to compete for with keywords, but in Shopping you never do that directly. Instead Google reads your product data and decides, on its own, which searches your item is relevant to. So the words and values you put in the feed are the whole game. This article is about the two fields that carry most of that weight, the title and the attributes, and how I use them to intercept the buyer intent I actually want rather than whatever Google guesses from a thin, generic record.

What is a google shopping feed, really

What is a google shopping feed? Mechanically, it is the structured file of product data you submit to Merchant Center, where each item is one row or record carrying a defined set of attributes: id, title, description, link, image_link, price, availability, brand, and product identifiers like GTIN, among others. Google uses that data to match your products to the right queries and to assemble the ad or free listing automatically. I lean on the account-structure side of this in the shopping feed setup process, but the concept that matters here is simpler: the feed is the only thing Google has to decide what your product is and who should see it.

That is the mental shift I try to get every client to make. A feed is not paperwork you clear so the campaign can run. It is the input that determines eligibility, matching, and how you appear on the results page. If a record is thin, Google fills the gaps with inference, and that inference is frequently wrong, which is how you end up paying for clicks on searches that were never your product. A rich, accurate feed removes that guesswork. Everything downstream, from bidding to how you appear against a query, sits on top of the quality of this data, and it all connects back to how Google Shopping works as a system.

The title: your single highest-leverage field for intent

If I could only fix one thing in a feed, it would be the titles. Because Shopping has no keyword list, Google matches your product to a search largely by reading the title against the query. That means the title is your targeting, and a title written to look nice on a shelf almost always loses to one written to mirror how people actually search. The pattern I default to is front-loading the terms a buyer types: brand, then product type, then the distinguishing attribute, color, size, material, or model. A shopper hunting a specific item searches with those specifics, and a title that leads with them is what intercepts that high-intent query.

The front of the title matters most, both because Google appears to weight the opening words and because the display truncates. So I refuse to waste the first characters on filler like "Buy now" or a vague collection name. Google's own guidance is blunt about this: accurately describe the product, match the title to your landing page, and avoid promotional text, all-caps, or gimmicky characters. As Google's official product data specification sets out, the title has a 150-character limit and should carry the distinguishing features of a variant. I use that room to add the query-relevant specifics, not marketing adjectives. The test I apply is simple: read the title cold and ask which search it would win. If the answer is "a broad, cheap one," I rewrite it to name the exact product a ready buyer is looking for.

How I structure a high-intent title

For a branded item I lead with brand plus product type plus the key attribute, then secondary attributes if the character budget allows. For an unbranded or generic product I lead with the product type and the specifics that qualify intent, since there is no brand equity to trade on. Where the same platform generates hundreds of titles from a template, I make sure the template still front-loads the meaningful terms, which is where a supplemental layer or a set of feed optimisation rules earns its keep. The discipline is identical whether you have ten products or ten thousand: put the words a buyer would type where Google reads them first.

The attributes that decide which queries you match

The title does the most, but the other attributes decide the shape and precision of your matching. The ones I watch first are product_type and google_product_category, because they anchor Google on what the product fundamentally is; a wrong or missing category is a reliable way to get matched against the wrong intent. Next come the identifiers: GTIN and brand connect you to identifier-driven and branded searches, and products submitted without a GTIN that has one tend to get limited visibility. Then the variant attributes, color, size, gender, age_group, which qualify you for the specific long-tail queries where purchase intent is highest.

These fields are not optional polish; they are how you either intercept a precise "red running shoes size 44" search or drift into a generic "shoes" pool where you burn budget against browsers. The description and product_highlight attributes add semantic context that helps on ambiguous queries, so I write the description to describe the product plainly rather than to sell, since Google reads it for matching, not persuasion. Getting the required attributes right also keeps products from being disapproved in the first place, which is where a lot of the diagnostic work in Merchant Center feed verification ends up focused. A feed that is complete and accurate at the attribute level is one where Google has to guess about nothing, and that is exactly the position you want to be in.

Why the feed replaces keyword targeting

The point that clicks for most people is this inversion. In Search, you decide which queries to compete for by choosing keywords and match types; you tell Google what you want. In Shopping, you describe the product honestly and Google infers which queries are relevant. Your leverage moves from bidding on keywords to writing feed data that mirrors how buyers phrase their intent. That is not a limitation, it is where the skill lives, and it is why I treat title and attribute work as the core optimization rather than an afterthought.

You still shape the traffic after the fact. Since you can't pick the queries directly, the main steering wheel is exclusion: negative keywords for Shopping let you cut the searches the feed made you eligible for but that you don't want. But exclusion only trims a pool the feed created. If the feed is thin, you are trimming the wrong pool entirely, and no amount of negatives fixes a title that made you eligible for the wrong intent in the first place. Get the feed right, then use negatives and bidding to refine. That order is what reliably moves Shopping performance in the accounts I run.

The workflow I use to build a feed for intent

My process starts from the buyer, not the catalog export. Before touching titles I look at how people actually search for each product type, then I rewrite the titles to front-load those terms. I confirm google_product_category and product_type are set correctly for every item, fill in GTIN and brand wherever they exist, and complete the variant attributes for anything that comes in more than one version. Only after that do I look at price competitiveness and bidding, because a well-built record is what makes those levers meaningful.

I also treat the feed as a living asset, not a one-time upload. Availability and price drift, new products come in with lazy default titles, and platform exports quietly break formatting. So I audit on a schedule: check disapprovals, spot-check titles against real queries, and confirm the required attributes still validate. When the source data is locked by a platform I can't fully control, I layer a supplemental feed to override titles or add custom_label values without re-exporting everything. The takeaway I'd leave you with is the one that reframes the whole task: the google shopping feed is not a data chore, it is your targeting written down, and the accounts that win Shopping are the ones that treat every title and attribute as a deliberate bid for a specific buyer's intent.