The Search Query Performance report is part of Amazon Brand Analytics. For each search term, it shows how many times shoppers searched, and the impressions, clicks, cart adds and purchases for all of Amazon next to your brand's own count and share. It comes weekly, monthly or quarterly, never daily. Read the gaps between steps to see where shoppers leave you.
There is a search term where you look like you are winning.
Amazon puts you in front of shoppers on it more than on any other term. You are proud of that share. You spend more to hold it.
Then you read the next column across.
Shoppers saw you and clicked a rival. The term you were winning is the term you were losing, one step further down.
The Search Query Performance report is Amazon’s record of what shoppers searched for and what happened next. For each search term, it shows impressions, clicks, cart adds and purchases for the whole Amazon store, next to your brand’s own count and share of each. It is Amazon’s own view of where shoppers find you, and where they pick someone else.
It is also easy to misread. Shares get averaged, missing weeks get read as zero, and a changing list of terms makes share look like it fell when it rose. Below: what the report holds, a worked example, what it cannot tell you, and how to use it.
What Amazon records for every search
Search Query Performance (often shortened to SQP) sits inside Brand Analytics, a set of Amazon reports for brand owners. To see it, you need a Professional selling account and must be a brand representative for a brand in Amazon Brand Registry (Amazon Brand Analytics).
In Seller Central (Amazon’s seller dashboard), it lives under Brands, then Brand Analytics. It has two views:
- Brand view shows the top search terms that led shoppers to any of your brand’s products.
- ASIN view shows the same thing for one product. An ASIN is Amazon’s product code.
Amazon describes the dashboard as showing “your brand’s share of performance compared to overall query performance” (Amazon Brand Analytics guide).
Software can pull the same data, product by product, through the Selling Partner API, Amazon’s data connection for sellers. The report type is GET_BRAND_ANALYTICS_SEARCH_QUERY_PERFORMANCE_REPORT (SP-API analytics report types). The API needs a separate Brand Analytics role on top of Brand Registry.
Weekly, monthly, quarterly. Never daily.
The report comes by week, month or quarter. One request covers one period. A week must start on a Sunday and end on a Saturday. Amazon’s API docs list a limit of 17 months for monthly requests. There is no daily version.
Four steps, from shown to bought
Each row is one search term. Search volume is how many times shoppers searched it. Then come four steps, each with a total for all of Amazon and a count for your brand:
| Step | What it counts | Your share means |
|---|---|---|
| Impressions | Times a product appeared in the results | How often Amazon showed you |
| Clicks | Clicks on a product in the results | How often shoppers chose you to look at |
| Cart adds | Products added to cart from that search | How often you made it to the cart |
| Purchases | Products bought from that search | How often you won the sale |
Share at each step is your count divided by the total, shown as a percentage. The API also returns click rates, median prices and shipping-speed splits, but the four shares are where most of the reading happens.
A worked example: one search term, your brand vs everyone
Illustrative numbers only. A hand soap brand looks at one generic search term, “foaming hand soap refill”, for the week of August 16 to 22, 2026. Shoppers searched it 40,000 times.
| Step | All of Amazon | Your brand | Your share |
|---|---|---|---|
| Impressions | 1,200,000 | 60,000 | 5.0% |
| Clicks | 30,000 | 900 | 3.0% |
| Cart adds | 6,000 | 150 | 2.5% |
| Purchases | 3,000 | 60 | 2.0% |
A 2% purchase share on its own says little. It could be great or poor. The useful part is where the share drops between steps. Impression share is what Amazon gave you. Purchase share is what you did with it.
Read the drops, not the level. The dashed line is where each bar would sit if you kept your 5% share all the way to the sale. Illustrative numbers.
Shown, but not clicked
Amazon showed this brand in 5% of impressions, but it won only 3% of clicks. Shoppers saw the product and scrolled past it. That points at the search card: the main image, the price as shown, the star rating, the review count and any badge. Buying more impressions here pays Amazon to show more shoppers a card they already skip.
Clicked, but not bought
Click share of 3.0% falls to 2.0% purchase share. Some shoppers who clicked went on to buy someone else’s product. Check the basics first: did you hold the Buy Box (the main Add to Cart offer), and were you in stock all week? Then look at the detail page, the price per unit and the reviews.
What the gap is worth in dollars
If purchase share caught up to the 5% impression share Amazon already gives you, that is 5% of 3,000 purchases, or 150. You had 60. So the gap is worth about 90 purchases a week. At an $18 average price, that is about $1,620 a week. This is a calculated size, not a promise. It assumes demand and rivals hold still.
The opposite shape: bought more often than shown
Now picture a second term, “unscented hand soap”, where impression share is 1% and purchase share is 3%. Shoppers who see you buy at a high rate, but Amazon rarely shows you. This is the one shape where paying for more visibility might add sales. It is still a guess until the ad data shows how much of that 1% was already paid.
What it can’t tell you
Anything daily
Weekly is the smallest period. A week’s numbers are not seven equal days, so do not split them. Weeks also run Sunday to Saturday. Your ad and sales dashboards may run Monday to Sunday. If so, re-cut one to match the other before you compare them.
Weekly periods also cross month ends. For a month, use Amazon’s monthly report rather than adding up four weeks.
An average of shares is not a share
Every share is already a percentage. Averaging percentages across terms treats a tiny term like a huge one. Say you have two terms:
| Term | Total purchases | Your purchases | Your share |
|---|---|---|---|
| “hand soap” | 8,000 | 80 | 1.0% |
| “lavender hand soap refill 3 pack” | 20 | 8 | 40.0% |
| Simple average of shares | 20.5% | ||
| Share of the combined total | 8,020 | 88 | 1.1% |
The simple average says you win a fifth of the market. You win about 1%. Add up the counts, then divide.
One more unit trap. In an export, check whether 0.4 means 0.4% or 40%. Test one row by dividing your count by the total. Read it wrong one way and a brand that is nearly invisible looks dominant. Read it wrong the other way and a leader looks invisible.
A missing term is not zero share
Amazon describes the list as the top search terms that led shoppers to your products. A term that was there for five weeks and is gone in week six is not 0% share. It is unmeasured. Amazon does not say why a given term drops out. In the data we see, the terms that come and go are mostly small ones.
This also means the report cannot show one of your largest gaps: the searches where shoppers never reach your products at all.
A share measured over a different list of terms
A share is only comparable over the same set of terms. The list of terms changes as your catalog and visibility change. Illustrative numbers:
| Week 1 | Week 8 | |
|---|---|---|
| Terms you appear on | 50 | 80 |
| Total purchases, all terms | 100,000 | 150,000 |
| Your purchases, all terms | 2,000 | 2,600 |
| Your share, all terms | 2.0% | 1.7% |
| Total purchases, the 50 terms in both weeks | 100,000 | 102,000 |
| Your purchases, the 50 terms in both weeks | 2,000 | 2,250 |
| Your share, the same 50 terms | 2.0% | 2.2% |
A new product brought in 30 new terms where the brand is small. Share over everything fell. Share on the same terms rose. Both numbers are correct. Only the second one answers “are we gaining ground?”
Ad spend, profit or paid vs unpaid
The report has no cost column. It has no field for which impressions were ads. A rising share could be a better listing, or it could be more ad spend. It says nothing about margin either. Winning a term at a loss still looks like winning.
Who you are losing to
The report shows the market total and your brand. It never names another brand. The Top Search Terms dashboard names the three most-clicked products on each term, with their click and conversion shares. It is a different report. Coming fourth and coming four hundredth look the same there. Its shares are for one product at a time, not your whole brand, so never subtract one from the other.
Thin terms and event weeks
A term searched a few dozen times a week can swing tens of points on one shopper. List those terms, but do not trend them. Deal weeks such as Prime Day move demand for everyone. A share that rises in a deal week can be a real win, but read those weeks on their own, apart from the trend.
How to use it without fooling yourself
- Fix a stable list of terms. Pick the terms that matter for each product, write them down with the date, and keep the list the same for every period you compare. Only add terms at a planned review.
- Split your own brand name from generic searches. Someone typing your brand name already knows you. Your share there measures recall, and it should be high. Share on a generic term like “hand soap” measures market position. Never blend the two into one share.
- Decide what counts as your name once. Include misspellings and joined spellings (“brandname” and “brand name”). Leave out rival names that share one word with yours. A single word match is how a competitor’s product ends up counted as your own brand.
- Compare same-length, closed periods. Week with week, month with month. Wait for three closed periods before calling a trend. Two periods is a difference.
- Read the drop at each step for each term. Shown but not clicked means fix the search card. Clicked but not bought means check Buy Box, stock and the detail page. Bought more often than shown means test more visibility.
- Pair it with ad spend and conversion. Pull the search term report from Amazon Ads for the same terms and weeks. Spend only where cost per click fits inside your margin per order, which our post on break-even ROAS (sales divided by ad spend) walks through. Remember that recent ad sales numbers are not final, so compare ad weeks at the same age.
- Size each gap in orders, then dollars. Label it “calculated”, and note that it assumes demand holds.
- Save every pull. Amazon’s API docs cap monthly requests at 17 months. Store each report as it arrives, so your trend does not depend on what Amazon still serves.
- Set a date to check again. For example: “Re-read this term four weeks after the new main image goes live.”
Then pick one term and act on its shape. If shoppers see you and scroll past, the money goes into the search card, not into bids. If they click and buy elsewhere, check the Buy Box and the stock before you touch anything else.
The report is only as useful as the definitions around it: which terms are on the list, which spellings count as your name, which products form a family. That is the case for AI-ready data, where those rules are written down once instead of rebuilt in each spreadsheet.
Synthesis joins a brand’s Amazon data with its ad and store data in one warehouse per brand, isolated from every other brand. Definitions, such as which listings make up a product family, are written once and every answer shares them. Every number carries a label with its period and its caveats.
Questions people ask
Who can see the Amazon Search Query Performance report?
Brand owners. You need a Professional selling account and must be a brand representative for a brand enrolled in Amazon Brand Registry. Software that pulls it through the API also needs the Brand Analytics role.
What is impression share on Amazon?
Impression share is your brand’s impressions on a search term divided by all impressions on that term, for the same period. If a term had 1,000,000 impressions and your products had 50,000 of them, your impression share is 5%. It says how often Amazon showed you, not whether anyone clicked.
Is there daily Search Query Performance data?
No. Amazon offers it by week, month or quarter only. Weeks run Sunday to Saturday. Do not divide a week by seven to make a daily number.
Does the Search Query Performance report show competitors?
No. It shows the whole market and your brand, never another brand by name. For the products winning a search, use the Top Search Terms report. It names the three most-clicked products on each term.
Does Search Query Performance include sponsored ads?
The report has no field that splits paid from unpaid results, so you cannot tell from it how much of your share came from ads. Pair it with the search term report from Amazon Ads to see what you paid for on the same terms.
Why did my search share drop when sales went up?
Often because the list of search terms changed. The report lists the top terms that led shoppers to your products. A new product adds new terms, and share over a larger set can fall on its own. Compare only the terms that appear in both periods.