Sellers pay hundreds of dollars a month for tools that estimate search volume and guess at click behavior. Meanwhile, Amazon hands brand-registered sellers the actual numbers – real impressions, real clicks, real purchases, per search query – and most of them have never opened the report.
It is called Search Query Performance, it lives inside Brand Analytics, and it is the closest thing to ground truth you will ever get on Amazon.
What Search Query Performance Actually Is
Every third-party tool works from the outside in. It scrapes search results, models traffic, and estimates what shoppers probably did. Search Query Performance works from the inside out. It is Amazon telling you, query by query, how the search funnel actually played out.
For each search query where your brand appeared, you get the full funnel: how many impressions the query generated, how many clicks, how many cart adds, how many purchases. And critically, you get your share of each stage.
That last part is what makes the report different from everything else in Seller Central. You are not just seeing your own numbers. You are seeing your numbers relative to the entire market for that query. When Amazon says you had 4% click share on a query with two million impressions, that is not an estimate. It is the ledger.
Where to Find It – and Who Gets Access
The report sits in Seller Central under Brands > Brand Analytics > Search Analytics > Search Query Performance. In Vendor Central it lives under Brand Analytics as well.
The catch: you need Brand Registry. This is brand-level data, and Amazon only shows it to the brand owner. Resellers do not get it. Sellers who skipped trademark registration do not get it. If you are still on the fence about Brand Registry, this report alone justifies the trademark cost – never mind A+ Content, Vine, and Sponsored Brands.
Once inside, you can view the data at two levels:
- Brand view – all queries where any of your ASINs appeared, aggregated across the brand.
- ASIN view – the same funnel filtered down to a single product.
You can pull data weekly, monthly, or quarterly, and export everything to CSV. Do the export. The Seller Central interface is fine for browsing, but the real work happens in a spreadsheet where you can sort, compare periods, and calculate rates Amazon does not show you directly.
How to Read the Funnel
Each row in the report is one search query, and each query gets four funnel stages:
- Impressions – how many times products appeared in results for this query, and how many of those were yours.
- Clicks – how many clicks the query generated in total, and how many landed on your products.
- Cart adds – same split, one step deeper.
- Purchases – the end of the funnel, again split between market total and your share.
For every stage you get the total market number, your absolute number, and your share as a percentage. Amazon also includes median prices and shipping-speed data at each stage, which quietly answers a question sellers argue about constantly: what price are people actually clicking and buying at for this query – not what competitors list at, but what converts.
The share columns are the point
Absolute numbers flatter you; share numbers tell the truth.
A query that brings you 200 purchases a month sounds healthy – until you see the market bought 8,000 times and your purchase share is 2.5%. Same data, completely different conclusion.
The diagnostic move is to read the shares as a sequence – and to be clear, no particular number is “healthy.” What matters is that the shares hold roughly level through the funnel. If you get 4% of impressions, 4% of clicks, and 4% of purchases, the listing is doing its job at every step; the same is true at 12% or 30%. The absolute level is a visibility question. The shape is a listing question.
The interesting rows are the ones where share collapses between stages. Where it collapses tells you exactly what is broken.
Diagnosis One: Seen but Not Clicked
Pattern: impression share 12%, click share 3%.
Amazon is showing you. Shoppers are scrolling past you. This is a search-results problem, and only three things exist in search results:
- Main image – the usual suspect, and the highest-leverage fix on Amazon. If your hero shot does not communicate the product instantly at thumbnail size on a phone, nothing else matters.
- Price – compare your price against the median clicked price in the report. If the market clicks at $19 and you sit at $27, you know why the click share is gone. Sometimes the answer is a price move; sometimes it is repositioning so you are not competing on that query at all.
- Title – less about keywords here, more about relevance at a glance. If the query is “insulated lunch bag for men” and your title reads like a generic cooler bag, shoppers self-filter you out.
Reviews play a supporting role – the star rating and review count show in results – but image and price do most of the damage.
What this diagnosis is not is a PPC problem. Raising bids on a query where shoppers already see you and do not click just buys more impressions to be ignored in. Fix the click-through problem first, then spend.
Diagnosis Two: Clicked but Not Bought
Pattern: click share 10%, purchase share 4%.
Shoppers choose you from search, land on your page, and leave. The search results did their job. Your listing did not. Now you are looking at everything that lives on the product page:
- Secondary images and A+ Content – the shopper clicked because the promise looked good. Something on the page failed to confirm it. Weak lifestyle images, missing dimension infographics, and thin A+ are the usual gaps.
- Reviews – both count and content. Read your recent negative reviews for this product; shoppers do, right before deciding.
- Price, again – but differently. In search, price filters clicks. On the page, price gets weighed against the full offer. Check the median purchase price for the query. If people click you at your price but buy competitors at a lower one, your page is not justifying the premium.
- Variations and availability – missing sizes, slow delivery promise, or a suppressed Buy Box will bleed conversion invisibly.
One more check before you rebuild the listing: make sure the query actually matches the product. If you show up for a loosely related query, shoppers will click out of curiosity and bounce out of reality. Low purchase share on an irrelevant query is not a problem worth solving.
Finding the Queries You Are Bleeding On
Here is a simple monthly routine that turns the report into decisions:
- Sort by headroom, not raw market size. The prize on any query is total market purchases times the share you are not capturing. A 2% share of a query with 10,000 monthly purchases is a huge opportunity; a 40% share of the same query is a position to defend, not a growth lever – and both matter more than a perfect score on a query with 50 purchases. Sort by market purchases as the first pass, then read your own share before deciding where the money actually is.
- Flag the funnel breaks. For each high-volume query, compare impression share, click share, and purchase share. Mark every row where share drops by half or more between stages.
- Compare against last period. A query where purchase share slid from 9% to 5% over two months is a competitor doing something – new listing, price cut, review surge. The report will not tell you who, but it tells you where to look.
- Feed it into PPC. Queries with strong purchase share and weak impression share are your best bid-up candidates: proven converters where you simply are not visible enough. Queries with high impression share and terrible purchase share are where you cut or fix before spending another dollar.
- Feed it into the listing. The top queries by purchase share are, by definition, the terms Amazon’s customers use when they buy you. If those exact phrases are not in your title and bullets, put them there.
That last point deserves emphasis. Most sellers build keyword lists from tools that estimate what people search. This report shows what people searched and then bought your product after searching. There is no cleaner input for a title rewrite.
The Limitations You Need to Know
The report is the best free data on Amazon. It is not perfect, and pretending otherwise leads to bad decisions.
- Queries are aggregated and truncated. Amazon groups similar queries and only shows those above a volume threshold. Your long tail – often a meaningful chunk of total sales – is invisible here. Do not conclude a keyword is worthless just because it does not appear.
- Brand view hides ASIN problems. At brand level, a strong product can mask a weak one on the same query. Whenever a brand-level number surprises you, drop to the ASIN view before acting. The aggregate is for spotting; the ASIN view is for diagnosing.
- It counts search, not everything. Purchases attributed here come from the search path for that query window. Browse traffic, external traffic, and repeat purchases through order history live outside this report. Your total sales will not reconcile against it, and they are not supposed to.
- Impression counting is generous. An impression means your product appeared in results – including positions almost nobody scrolls to. High impression share on page three is not the same asset as high impression share above the fold. Cross-check with actual rank before celebrating.
- Data lags. Weekly data arrives a few days after the week closes. This is a strategic report, not a live dashboard. Use it monthly for direction, not daily for reaction.
Closing
Third-party tools estimate the market from the outside. Search Query Performance is Amazon opening the books.
It tells you which queries matter, where your funnel breaks, and whether the problem is your main image or your product page – before you spend a single dollar on ads to paper over it.
It is free. It is sitting in Brand Analytics right now. Open it.