Ecommerce Search Terms Need Margin Bands Before Negative Keyword Decisions

Blended ROAS can hide which search terms are truly wasteful and which ones still work for higher-margin products. Review queries by margin band before adding negatives.

What This Means: The Practical Takeaway

The same search term can be too expensive for low-margin products and still be acceptable for higher-margin inventory. That means one blended ROAS target is a poor rule for negative keyword decisions. Ecommerce teams should review queries against margin bands, not one account-wide average. This keeps profitable demand from being blocked just because another product tier could not support the same click cost.

Blended ROAS makes search term cleanup look simpler than it is.

That is the operational mistake. One query can touch several product lines with very different contribution economics, refund behavior, or shipping burden. When the team reviews that query only at account level, it sees one average outcome and is pushed toward one average decision.

That decision is often too blunt.

One Query Can Be Good For One Product Tier And Bad For Another

Search terms do not arrive with profit context attached.

In ecommerce accounts, one generic query can trigger products with very different margins, repeat-purchase value, and operational costs. A term that looks mediocre in aggregate may still be acceptable when it reaches a high-margin accessory, bundle, or replenishable line. The same term can be destructive when it routes into low-margin hero products where there is little room to absorb click costs.

That is why a flat negative keyword rule creates avoidable damage. The query may not be the real problem. The problem may be where the query is landing.

Blended Efficiency Metrics Hide The Wrong Kind Of Waste

Blended reporting is useful for pacing and portfolio management. It is weaker when the team is deciding whether a search term should be blocked.

If margin differences are buried inside the product mix, a query can survive because strong products are carrying weak ones. The reverse is just as common. Good demand gets blocked because the team judged the term by low-margin outcomes elsewhere in the account.

That creates two opposite errors:

- keeping a query live everywhere because one profitable product line saved it - blocking a query everywhere because one low-margin segment made it look bad

Both mistakes come from the same shortcut. The account is judging search terms without a product-economics layer.

Margin Bands Make Search Term Review More Honest

Margin bands are a practical way to add that missing layer.

The team does not need perfect SKU-level profit modeling on day one. It needs a small set of usable groups such as low margin, core margin, and high margin. Once those bands exist, the review process can ask a better question:

"Was this query bad everywhere, or bad only for one economic tier?"

That changes the next action.

- Some queries become negatives only for low-margin inventory. - Some queries stay live for higher-margin collections with tighter monitoring. - Some queries should route to different product groups or landing pages instead of being blocked. - Some queries expose pricing, merchandising, or feed-structure problems rather than query-quality problems.

This is where search term optimization becomes a business-control system, not just a cleanup task.

Product Labels And Value Rules Turn The Idea Into Workflow

Google Ads and Merchant Center already provide the raw building blocks for this kind of review.

Custom labels let ecommerce teams group products using business logic that Google does not infer on its own. Margin is one of the most useful labels because it turns product economics into something the account can segment. Product and listing groups can then be subdivided around those labels, and conversion value rules can help reporting or bidding reflect relative business value more honestly.

That does not replace human judgment. It makes human judgment more defensible.

Instead of arguing about whether a query is "good" in the abstract, the team can see how the same query behaves when it reaches low-, mid-, and high-margin product sets. Negative keyword decisions get narrower. Spend tolerance gets clearer. Performance conversations get less noisy.

Negative Keyword Scope Should Follow Contribution Logic

Negative keywords are supposed to prevent wasted spend. They should not erase acceptable acquisition paths for profitable inventory.

A margin-aware workflow helps the team avoid two common failures:

1. adding a global negative because low-margin products made the query look weak 2. leaving a query untouched because high-margin products made the query look safe

Those are opposite mistakes, but they come from the same flaw: treating blended economics as universal truth.

Once margin bands are visible, the right scope becomes easier to see. A term may deserve a tighter product-group exclusion, a different feed label, a routing change, or a separate threshold for review. The account stops pretending every click should be judged by the same level of tolerance.

How To Do It

Step 1: Export the search terms report with query, clicks, cost, conversions, conversion value, campaign, asset group or ad group, and product context where available. If product economics are not visible in the report, prepare a simple join back to feed labels, campaign structure, or a margin lookup so the review is not blind to product tier.

Step 2: Create three usable margin bands such as low, core, and high. Assign them through Merchant Center custom labels, feed rules, or campaign segmentation so every important product set can be mapped to one band. Keep the scheme simple enough that the team can maintain it.

Step 3: Review search terms inside each margin band before making exclusions. Mark whether the query is unprofitable everywhere, only weak for low-margin inventory, or still acceptable when it reaches higher-margin products. Do not make a global negative decision until that comparison is visible.

Step 4: Separate actions by economic tier. Low-margin failure may justify a scoped negative, tighter match control, or more restrictive routing. High-margin tolerance may justify keeping the term live, improving the landing page, or using a higher spend threshold before exclusion.

Step 5: Feed the same logic back into reporting and bidding. Use custom labels, product groups, and value-based settings so the account reflects different business value instead of pretending every conversion is equal.

Final check: Audit every new negative keyword that affects more than one product set. If the query only failed in one margin band, narrow the scope before it becomes an account-wide rule.

Sources

- [Google Ads Help: About the search terms report](https://support.google.com/google-ads/answer/2472708?hl=en-EN)

- [Google Ads Help: About conversion values](https://support.google.com/google-ads/answer/13064207?hl=en-EN)

- [Google Ads Help: Set up conversion value rules](https://support.google.com/google-ads/answer/10520348?hl=en)

- [Google Ads Help: Use custom labels for Shopping ads](https://support.google.com/google-ads/answer/6275295?hl=en)

- [Google Merchant Center Help: Product and listing groups](https://support.google.com/merchants/answer/3517331?hl=en)