No — a separate campaign for every SKU is usually the wrong move, and it's the mistake that follows right after the right one: getting product-level tracking working at all. Once you can see which product sold, the tempting next step is a one-to-one campaign for every item in the catalog. That's manageable for a store with a dozen products. For a store with a few hundred, it turns into a structure nobody can actually manage, and a bidding algorithm that never sees enough sales on any single campaign to learn from.
What you need instead is a small number of groups, built around margin and demand, not one campaign per item. A handful of bestsellers that carry the account deserve their own visibility. Everything else can be grouped by how well it sells and how much it earns you when it does, so the budget and the bidding both have enough volume to work with.
Why one campaign per product breaks down
Shopping and Performance Max campaigns both rely on the platform's automated bidding (Google's system that adjusts what it pays for a click or a sale based on the conversion data it's collected) to find the right price for each auction. That system needs a steady flow of conversions to learn from. Split a catalog of three hundred products into three hundred campaigns, and most of them get a handful of sales a month, if that. The algorithm never accumulates enough signal on any one campaign to bid well, so performance stays flat or erratic on the exact products you split out to give more attention.
There's also a plainer problem: nobody has time to manage three hundred campaigns. Bids drift, budgets sit unspent on slow movers, and the one or two products actually worth close attention get lost in the list.
Group by what the product actually is to your business
The grouping that works is built on two questions: how well does this sell, and how much does it earn you when it does. Not every product answers those the same way, and treating a clearance item the same as a bestseller wastes budget on one of them.
A few groups usually cover a whole catalog:
- Bestsellers — the handful of products driving most of your revenue. These can justify their own campaign or a tightly scoped group, because there's enough sales volume for the bidding to learn from and enough revenue at stake to watch closely.
- High-margin, lower-volume — products that don't sell as often but are worth more each time they do. Grouped together, they generate enough combined volume for the algorithm to work with, while keeping bids high enough to compete for them.
- Low-margin or clearance — items where the priority is moving inventory, not maximizing profit per sale. These can run on tighter budgets and lower bid ceilings without dragging down the account's overall numbers.
- Everything else — the long tail. Grouped into one or two broad campaigns instead of ignored, so it still generates some volume without needing individual attention.
This isn't a permanent structure. Products move between groups as they sell better or worse, go on promotion, or go out of stock. The point isn't to get the grouping perfect once — it's a structure with few enough pieces that someone can actually watch and adjust as reality changes.
Product-level tracking is what makes the grouping possible
None of this works without knowing, at the product level, what's actually selling and at what margin. Tracking and analytics infrastructure built around a current product feed and platform conversion tracking is what turns "campaign performance looks fine" into "this specific product's cost per sale is too high for its margin." Without that visibility, grouping by margin and demand is a guess dressed up as a strategy. You're back to reading account-level numbers and hoping they mean what you think they mean — the same problem last-click reporting causes when figuring out which ads actually drove a sale.
Feed hygiene matters here too. Google Shopping and other platforms build their campaigns directly off your product feed, so a feed with stale prices, missing availability, or thin product data undermines any grouping structure sitting on top of it, no matter how well-designed.
Where to start
If your account is either one campaign holding every product or a campaign for each one, neither extreme is serving you. Pull a sales-by-product report, rank by revenue and margin, and let that number, not your product count, decide how many groups you actually need. For an online store, that's the difference between a structure someone can actually watch and one that just looks thorough on paper.