Shopping feed gaps and Performance Max overlap
A home-hardware retailer lists 12,000 SKUs in Merchant Center. About 2,300 products lack GTINs, 1,100 are disapproved for promotional overlay text in the main image, and 800 have missing product type. Standard Shopping is split into high-priority best sellers, medium-priority seasonal items, and low-priority long tail. A Performance Max campaign serves all products and now takes 70% of Shopping-related spend, while standard Shopping impression share is 35%. Margins are thin on power tools and healthy on accessories. Free listings are enabled but unmanaged. Last quarter, accessory revenue grew 6% while power-tool revenue fell 4%. The feed updates daily. The team wants to pause Performance Max to 'stop cannibalizing' standard Shopping and move budget back to high-priority Shopping. The account has a shared budget.
1. Which feed issues should be fixed first, and why? How would you prioritize against pausing Performance Max?
Read the worked answer
Fix feed eligibility before changing campaign structure. The 1,100 disapproved products have zero ability to serve, so they are the clearest loss. Remove promotional overlay text from main images and resubmit. Next, address the 2,300 missing GTINs, which can limit product matching, eligibility, and Performance Max asset quality; use manufacturer GTINs where available and correct the feed source. Then fix the 800 missing product types, because product type drives listing groups, reporting, and asset-group structure in Performance Max.
Pausing Performance Max before fixing the feed would likely reduce reach without solving the underlying inventory visibility problem. The feed updates daily, so fixes can be validated quickly. Also claim and manage the free listings; unmanaged free listings can distort the picture of paid coverage.
Prioritize by revenue at risk and ease of fix: disapprovals first, GTINs second, product type third, free listings alongside. Only after those changes should the team judge whether Performance Max overlap is a real problem or a symptom of poor product data.
2. Evaluate the proposal to pause Performance Max. What structure and bidding approach would you recommend instead?
Read the worked answer
The cannibalization claim is not proven. Standard Shopping priority only orders standard Shopping campaigns against one another; it does not establish priority over Performance Max. Pausing Performance Max could remove incremental reach and conversion volume rather than transfer it cleanly to standard Shopping. The 35% impression share also suggests there is room to grow, not necessarily that one campaign is stealing from the other.
Recommend a controlled test instead of a blanket pause. Keep Performance Max running, but isolate brand traffic with brand exclusions where appropriate, add account-level negative keywords, and use search themes or asset groups that reflect margin tiers. Keep standard Shopping for control and for best-seller coverage, but do not rely on campaign priority as a margin lever.
Segment by economics: accessories have healthy margin, so they can support a higher target ROAS and broader coverage; power tools need a lower bid or stricter tROAS because margin is thin. Use separate budgets or portfolio strategies aligned to margin. Run a geo holdout or experiment pausing Performance Max in matched markets to measure incremental profit. If standard Shopping recovers most of the lost volume profitably, adjust; if not, keep both with clearer roles.
3. How would you allocate budget and measure incremental profit by category rather than reported revenue?
Read the worked answer
Start with margin-based labels in the feed: for example, custom label 0 for margin tier, with power tools in a low-margin tier and accessories in a high-margin tier. Build campaign or asset-group reporting around those labels so budget can be compared on gross profit, not revenue. Set target ROAS by tier rather than one blended target; low-margin power tools should face a higher required return or a lower bid.
The revenue pattern matters: accessories grew 6% while power tools fell 4%. That may be a healthy mix shift, not a failure, but only margin data can confirm it. Measure incremental profit with a geo holdout or experiment, comparing markets with the new allocation against matched controls. Include returns, shipping cost, and free-listing influence.
Allocate budget first to categories where incremental profit per dollar is highest, not to the largest revenue line. Keep a small test budget for long-tail SKUs and review query and product-level performance. Do not pause Performance Max solely on last-click revenue; use the holdout result and margin tiers to decide.