The labelizer playbook: tier your feed by margin, watch ROAS jump

Most brands start Shopping with one campaign, one product group, one Target ROAS. It's fine at 50 SKUs. It collapses at 2,000.

The problem: Google bids on expected margin, but doesn't know your actual margin. It bids on expected conversion velocity, but not your actual velocity. Without labels, every SKU looks identical to the auction, and your budget flows to whatever converts fastest regardless of whether you make money on it.

Hero SKUs: high margin (60%+ gross) and high 30-day velocity. These are the products where every extra click is likely profit. Bid aggressively.

Workhorses: mid margin (30-60%) or mid velocity. Steady state products, the body of the catalog. Bid for a healthy but not heroic ROAS.

Bleed SKUs: low margin (<30%) or low velocity (fewer than 3 sales in 30 days). Cap their spend or exclude them.

Those three buckets are enough to win. You can split further later.

Export your full product catalog with margin and 30-day conversion count. In Google Sheets, write a tier formula: if margin >= 60% and velocity >= 10 sales, tag 'hero'. If margin < 30% OR velocity < 3, tag 'bleed'. Everything else is 'workhorse'.

Push the tier back to Merchant Center as a custom label (custom_label_0 works). Rebuild your Shopping campaigns around those labels: one campaign per tier, different Target ROAS each.

Do this manually for a one-shot restructure. Do it daily if you want the tiers to stay current. That's what the labelizer automates.

When the jewelry brand restructured into three tiers with per-tier Target ROAS (4.0 / 3.0 / 2.2), two things happened fast.

Hero SKUs got more budget. Google stopped starving the 60%-margin bestsellers to fund a low-margin product that converted at a similar rate.

Bleed SKUs stopped eating budget. Products that were losing money on every sale got capped, freeing up spend for the middle.

Within a week, blended ROAS was up 38%, hero SKUs were carrying more volume, and the team didn't have to touch a single bid manually.

If your Shopping campaigns aren't tiered by margin and velocity, that's usually the single highest-ROI change in the account. The only real decision is whether to do it once manually or run it daily with a tool like the labelizer.

Either way: stop treating your catalog as one undifferentiated blob. Your top 10% of SKUs deserve a different strategy than your bottom 10%.