Weekly ecommerce tips, deals & news.
A custom label is an optional field in a product feed that you define yourself. Shoppers never see it. Instead your ad platform uses it to group products for bidding and reporting. Margin tier, season and bestseller status are the usual choices.
Custom labels work by carrying a value you invent into the feed, where your ad platform can filter on it. First you choose what it means. Then you build product groups that target only the products carrying that value.
Google Shopping gives you five custom label fields per product, numbered custom_label_0 to custom_label_4. AdTribes describes each as taking a text value of up to 100 characters.
Also, one product carries one value per field. So a product can be tagged along five separate dimensions at once, but never twice on the same one.
In practice most stores use two or three fields and leave the rest empty. Filling all five without a plan just creates work nobody maintains.
The values are yours to choose, and a handful of patterns come up repeatedly.
AdTribes sets out the same five strategies in its guide to Google Shopping campaign structure. These are conventions rather than rules, because the platform never reads the meaning of your values. So consistency inside your own account is the only thing that matters.
A label is only useful if something keeps it true. Two approaches exist, and only one of them survives a growing catalog.
Rule-based values update every time the feed rebuilds. So a product moving between margin tiers relabels itself. A hand-typed value stays exactly as wrong as the day somebody stopped maintaining it.
Therefore decide the rule before you settle on the label names. The naming is the easy half of the job.
A label sitting in the feed does nothing on its own. Your feed plugin writes the value during feed mapping, alongside the required attributes. Then it reaches Google Merchant Center with everything else.
Instead the work happens in your ad account. There you subdivide product groups by custom label rather than by category or brand. So high_margin products can take a higher bid than low_margin ones in the same campaign.
Combining two labels is where the value usually shows up. A group matching both bestseller and high_margin is your proven, profitable set. Therefore that group is the one that justifies your top bid.
Also, labels make reporting readable. Spend by margin tier answers a question that spend by category cannot.
The limits are published, and they are more generous than most catalogs need. Google’s product data specification allows up to 1,000 unique values account-wide for each custom label attribute.
In practice that ceiling is rarely the constraint. A store using six margin and season values sits nowhere near it. A store approaching it has usually started labeling by product rather than by group.
Also, feed tooling itself is well established. AdTribes Product Feed PRO, our own feed plugin, reports 80,000+ active installations. So writing custom labels into a feed is routine work rather than specialist work.
Still, no published figure tells you whether your own labels are working. That answer lives in your ad account, comparing return by label group. In practice a label that never changes a bid is just decoration.
Here’s a hypothetical example. So picture a pet supplies store running one Shopping campaign across about 900 products.
Every product sits in one product group, sharing one bid. First the campaign spends wherever clicks are cheapest. Then the report shows a blended return that hides everything useful.
Dog beds at 45% margin and bulk kibble at 8% get identical treatment. Consequently the cheap-click products soak up the budget. Meanwhile nobody can prove that from the reports they have.
They write custom_label_0 with three values: high, medium and low. Their feed plugin fills it from a margin calculation rather than by hand.
Then they split the campaign’s product groups on that label. High-margin products get a higher bid, and low-margin ones get a floor.
Total spend does not change at all. Instead the same budget moves toward products that can actually pay for it. In short, the mix changed rather than the amount.
Also, the labels did not improve a single product page. A well-bid product with a thin description still converts poorly, and no bidding rule rescues that.
Their season label went stale by March, because nobody owned the refresh. Therefore they moved it onto the same schedule as the feed itself. Even so, one label stayed wrong for weeks before anyone noticed.
Nobody had written down what high and low actually meant either. So a new hire guessed, and two products landed in the wrong tier.
| What you’re comparing | Custom label | Product category |
|---|---|---|
| Who defines it | You do | Google’s own taxonomy |
| Who sees it | Nobody outside your account | It shapes where ads appear |
| What it is for | Bidding and reporting groups | Classifying what the product is |
| How many per product | Five fields, one value each | One category |
| Can you rename values | Yes, any time | No |
A category says what a product is, while a custom label says how you want to treat it. So the two never compete for the same job. Getting a category wrong can cost you visibility, whereas getting a label wrong only confuses your own reports.
Meanwhile, for the attributes shoppers do see, feed optimization is the relevant discipline. Custom labels sit entirely on the other side of that line.
Both matter, though they fail differently. A wrong category can stop a product showing at all, while a wrong label only sends budget somewhere odd. Consequently the category deserves the stricter review.
Five per product, numbered custom_label_0 through custom_label_4. Each field takes a single value for that product. Most stores get everything they need from two or three, so there is no prize for filling all five.
No. They are internal fields, so they play no part in approval or in what shoppers see. A disapproval comes from the attributes describing the product, not from how you chose to group it.
Margin tier, assuming you know your margins. It is the one label that changes a bidding decision immediately. Bestseller status makes a reasonable second, though it needs refreshing as your sales data moves.
Custom labels matter because a flat catalog forces one bid onto products with very different economics. Instead labeling lets your budget follow margin rather than click price. So the same spend ends up doing more work.
Meanwhile your reporting finally answers the question owners actually ask. Which products are worth advertising, and which ones are merely cheap to click?
Copyright © StoreOwnerTips.com. All Rights Reserved.
