PRODUCT FEED STRATEGY: BESTSELLERS FIRST FOR BUDGET

Almost every underperforming Shopping account I inherit has the same hidden problem: the budget is spread evenly across a catalog where the products are anything but equal. A handful of items carry the revenue, and the rest quietly drink spend that should be compounding on the winners. A google ads product feed strategy for bestsellers exists to fix exactly that, and it's the single highest-leverage change you can make to a feed that's already technically clean.

The idea I keep coming back to is simple to say and easy to get wrong: your feed is not just a data file, it's a budget-allocation instrument. A google ads product feed strategy bestsellers approach uses that fact directly, because the group you can see in the feed is the group you can fund separately. The whole game is turning the feed into something that lets you point money at your proven sellers instead of spraying it across everything.

That's what the feeder strategy is for, and I'll walk through it the way I run it on live accounts: what the strategy actually is, how I find the real bestsellers from data, how I tag them in the feed, how I give them their own budget priority, and how the whole thing settles into a tiered structure that keeps feeding money to what works.

What the feeder strategy actually is

The feeder strategy google ads practitioners talk about is, at its core, a way of tiering the product feed by performance and letting budget flow down those tiers. The top tier is your bestsellers: proven, efficient, high-volume products that earn the priority. Below them sits a tier of steady mid-performers, and below that a tier of unproven or long-tail items that haven't shown they can convert yet. The name comes from the way products "feed" upward: a new item starts in the bottom tier and only graduates once its own data justifies more budget.

What makes it a feed strategy rather than a bidding tweak is that the segmentation lives in the data source, not in a spreadsheet on the side. The tier a product belongs to is written into the feed as a value Google can read, which means your campaigns can target it directly. Without that, you can't fund bestsellers separately no matter how good your bidding is, because Google has no way to tell them apart from the rest of the catalog.

This is where the feeder strategy connects to how you build the feed in the first place. The columns you keep decide what you can segment on later, which is why I treat the tiering decision as part of the google shopping feed template rather than an afterthought. If the label column isn't in the template from the start, adding it retroactively across a live feed is far more friction than reserving it up front.

How I identify true bestsellers, not favourites

The mistake I see most often is deciding bestsellers by instinct: the products the client is proud of, the newest launch, the highest-margin line. None of that is a bestseller unless the data agrees. I pull the products that actually drive the bulk of revenue and conversions over a rolling window, usually the last 30 to 90 days depending on how fast the catalog moves, and I start from there rather than from anyone's opinion.

But volume alone isn't enough, and this is the part inference matters. A product that spends heavily and sells a lot isn't automatically a bestseller for budget purposes if it does so at a poor return. So I filter twice: first by sales volume, then by efficiency. A true bestseller sells in volume at a return that comfortably clears the account's target. A high-spend, low-return item is not a bestseller, it's a leak, and putting it in the priority tier would defeat the whole point.

Because the label is a claim about current performance, it has to be refreshed on a schedule or it lies. A product that was a bestseller last quarter may have cooled off; a quiet SKU may be climbing. This ties into ongoing Shopping campaign optimization, because the bestseller list isn't a one-time export, it's a rolling decision you revisit as the data moves.

Tagging bestsellers in the feed with a custom label

Once I know which products qualify, I flag them in the feed with a custom label, and I reserve one label slot for selling rate so its meaning never drifts. Google's own documentation uses exactly this pattern: a custom label defined for selling rate, with values like best seller and low seller. That's not a coincidence, it's the intended use, and it's the mechanism the whole feeder strategy depends on. You can read the attribute-level rules in Google's custom labels documentation, which spells out the format, the one-value-per-product rule, and the limits.

Two operational details matter here. First, you submit only one value per custom label attribute per product, so a SKU is either flagged for the bestseller tier or it isn't, not both. Second, you have up to five custom labels available, numbered custom_label_0 through custom_label_4, so pick one and keep it dedicated to selling rate rather than mixing season, margin, and bestseller status into the same column. Consistency in what a label means is what makes the segmentation trustworthy months later.

The way I actually apply the label is with feed rules, not manual editing, so it stays current on its own. A feed rule can set the selling-rate label automatically based on data you already submit, which means the day a product's numbers cross my threshold, its label updates on the next refresh without anyone touching the file. The label is only as reliable as the rule that maintains it.

Giving bestsellers real budget priority

Tagging bestsellers does nothing on its own; the payoff comes from acting on the tag. Once the label exists, I split the flagged products into their own campaign or, at minimum, their own product group with a dedicated budget. This is the crux of the whole google ads product feed strategy for bestsellers: a shared budget means a few high-spend, low-return items can starve your proven winners of impressions at the exact moment they'd convert. A separate budget removes that competition entirely.

With bestsellers isolated, I can also set a target that fits their economics rather than the catalog average. Proven products usually justify more aggressive scaling, so I give them room to spend as far as demand allows instead of capping them at a budget sized for the whole account. The mid and long-tail tiers run on tighter targets and smaller budgets, which stops them cannibalizing the spend that should be compounding on the best items.

The bidding lever you pull on the bestseller tier depends on how much data those products generate. When a segment has enough conversion history, letting the system optimize toward value tends to outperform manual tinkering, which is why I lean on Shopping bidding strategies tuned per tier rather than one target across the account. The bestseller tier and the unproven tier are different problems, so they get different targets.

The feeder tiers in practice

Put together, the structure is three layers that money moves through. The top tier is bestsellers, funded first and scaled hardest. The middle tier is reliable mid-performers on a steady but capped budget. The bottom tier is new arrivals and long-tail items on a small budget whose only job is to generate enough data to prove themselves. Budget priority is explicit: bestsellers are funded to the point of diminishing returns before the lower tiers get aggressive spend.

The movement between tiers is the whole point of the word "feeder." A bottom-tier product that starts converting efficiently graduates into the mid tier, and a mid-tier product that becomes a consistent, efficient seller graduates into the bestseller tier. It works the other way too: a bestseller that loses its efficiency drops down. Because the tier is written into the feed as a label, this promotion and demotion happens by updating the label, and the campaigns follow automatically.

What I like about this structure is that it survives a growing catalog. New products don't disrupt your winners because they land in the bottom tier by default, on a contained budget, until they earn their way up. That containment is really a way to avoid cannibalization between unproven products and proven ones. The bestseller tier stays protected precisely because everything else has to prove itself before it can compete for the same budget.

My workflow: label, split, fund, revisit

The routine I settle into on every account is short and repeatable. I pull the performance data, define the bestseller threshold on volume and efficiency together, and set a feed rule that writes the selling-rate label from that logic. Then I split the flagged products into their own campaign with a dedicated budget and a target that fits their economics, leaving the rest of the catalog in the lower tiers on tighter constraints. The value comes from what happens after the build.

Because the label is a claim about current performance, I revisit it on a cadence that matches the catalog: weekly for fast-moving stores, monthly for stable ones. Each review promotes rising products and demotes ones that cooled off, and since the movement is just a label change the feed rule handles, the campaigns re-sort themselves without a rebuild. That discipline is what separates a feeder strategy that keeps working from a one-time segmentation that's stale within a month.

The takeaway is short. Treat the feed as a budget instrument, not a static file: identify bestsellers from real sales and efficiency data, tag them with a dedicated selling-rate custom label applied by feed rules, split them into their own budget so they can't be starved, and revisit the label often enough that it always reflects what's actually selling. Do that and your spend compounds on your winners instead of leaking across a catalog that was never equal.