The Search Query Performance report tells you where you lose the shopper, not which keywords to copy.

The Search Query Performance report is the only place Amazon tells you where in the funnel you lose the shopper — and almost everyone uses it as a keyword-mining list, which throws away the one thing it does that no other report can. Most sellers open SQP, sort by search volume, copy the top queries into a broad-match campaign, and close the tab. That is using a funnel diagnostic as a keyword scraper. The report's real job is to show you whether a query is failing at the click, at the cart, or at the purchase — because each failure has a different fix, and only one of them is a bid problem.

What the Search Query Performance report actually is

SQP lives inside Amazon Brand Analytics, so it is available only to Brand Registry-enrolled sellers. For every search query that led to your brand, it reports Amazon's own first-party funnel: the total search query volume, then impressions, clicks, cart adds, and purchases — each shown both as the market total across all brands and as your brand's count and share of it. It is the closest thing Amazon gives a seller to the shopper's actual path from typing a query to paying for a product.

Two boundaries matter before you trust a single number in it. First, the impression counts on the search results page include both organic placements and Sponsored Products ads — but not Sponsored Brands, Sponsored Display, or anything served on a product detail page. So SQP is a search-page view of demand, not a whole-account view. Second, it is query-level, not campaign-level: it tells you how a search term performed for your brand, not which campaign or match type served it. It is a demand map, not a bidding console — and confusing the two is the first mistake.

The number nobody reads: click share versus purchase share

The volume column is the one everyone sorts by, and it is the least useful. The column that earns the report its keep is the gap between your click share and your purchase share on the same query. Click share is the fraction of all clicks on that query that came to your listing; purchase share is the fraction of all purchases on that query that were yours. When those two numbers diverge, the report is pointing at money.

A query where you hold a high click share but a much lower purchase share is a listing that wins attention and loses the sale — a price, image, review, or availability problem the auction cannot fix. A query where you hold low click share but a high purchase-to-click rate is the opposite: shoppers who do reach you buy, and the constraint is visibility, which is a bid-and-relevance problem. Pouring spend into the first case buys more expensive clicks that still do not convert. Pulling spend from the second starves a query that was already earning its place. The volume ranking cannot tell those two situations apart; the funnel shares can.

Where SQP stops, and the bid begins

SQP is diagnostic, not prescriptive. It tells you a query converts well for you; it does not tell you what that click is worth or what you should pay for it. That answer comes from margin, not from search volume.

Break-even ROAS = 1 ÷ margin. An ASIN carrying 35% margin after COGS and fees breaks even at roughly a 2.9x ROAS — a 35% ACoS. SQP can tell you a query has a strong purchase share and is worth pursuing; only the break-even math tells you the ceiling bid that keeps the pursuit profitable. A high-converting query bid above break-even still loses money on every sale.

This is why a report that looks like a targeting cheat sheet is really an input to a bidding decision. SQP narrows the field to queries that deserve attention; the break-even calculation — the same one behind what counts as a good ACoS in 2026 — decides how much each one is worth. Skip that second step and the report just tells you which expensive mistakes to make faster.

The EU wrinkle: there is no "your" Search Query Performance report

SQP is per-marketplace, and that is not a formatting detail — it is the whole point for a European seller. The query volumes, the click shares, and the purchase shares on amazon.de describe German shoppers searching German terms; amazon.fr, amazon.it, and amazon.es each generate their own, unrelated funnel. A term that dominates purchase share in Germany can sit mid-table in Italy because the shopper phrases the need differently, the competitive set is different, and the price sensitivity is different.

Read one marketplace's SQP and act on it everywhere and you repeat the error behind the documented +28% ACoS penalty on cross-marketplace copy shipped without per-region tuning — the one we unpack in EU-native is not a translation layer. The queries are not translations of each other. Pull SQP per marketplace, and treat each as its own demand map. This is exactly the asymmetry Mirox's Semantic layer is built around: German queries matched against German shopper intent, not routed through English first.

What SQP feeds, and what feeds off it

The report is most valuable when it is wired into the rest of your PPC hygiene rather than read in isolation. The high-click, low-purchase queries it surfaces are prime candidates for the search-term review that drives your negative keyword list — if a query pulls clicks and never converts, it belongs on the negative list before it belongs in a campaign. The high-purchase-share queries are the ones worth graduating from broad discovery into tighter match types where you can price them deliberately. And the queries where your click share is healthy but growing are the branded-demand signal that Sponsored Brands is meant to build. SQP does not do any of that work — it tells you which queries each lever should point at.

How to use the Search Query Performance report

1. Pull it per marketplace, never blended. Export DE, FR, IT, ES and the rest separately; a query's funnel in one market says nothing about its funnel in the next.
2. Sort by the click-share-to-purchase-share gap, not by volume. The divergence between the two is where the report points at a fixable problem — volume just tells you how big the query is.
3. Split the diagnosis before you touch a bid. High click share and low purchase share is a listing problem — price, images, reviews, availability. Low click share and high conversion is a visibility problem — the only one a bid actually solves.
4. Price the visibility queries on margin, not on volume. Take the queries where you convert and set the ceiling bid from break-even ROAS, so pursuing a strong query never costs more than the sale is worth.
5. Route the rest to hygiene. Feed the clicks-without-purchases queries to your negative-keyword review; graduate the high-purchase-share queries into tighter match types where you control the bid.
6. Re-pull on a fixed cadence. Query funnels move with season, price, and competition — a quarterly snapshot read once is a photograph of a moving thing.

The one-line version

The Search Query Performance report is a funnel diagnostic, not a keyword list. Its value is the gap between your click share and your purchase share on a query — that gap tells you whether you are losing the shopper at the ad, the listing, or the checkout, and only the last of those is a bid problem. Read it per marketplace, act on the divergence rather than the volume, and let margin — not search volume — decide what any of it is worth paying for.

See the conversion and margin inputs attached to a single bid, or read how TACoS versus ACoS keeps a strong-converting query from being judged on the wrong number.