Key takeaways: The search term report is the one report that shows what a shopper actually typed. Pull it weekly for the last 60 days, and run five sorts: zero-order spend, high-order terms not yet in exact match, terms converting in the wrong campaign, ACoS outliers, and new terms with rising impressions. Each sort has one action attached, so the report becomes a routine instead of a reading exercise.
The Amazon search term report lists the actual search queries, or the ASINs for product targeting, that triggered your ads, with the clicks, spend and sales each one produced. It is the only report where you see the shopper's words rather than your own keywords. Read it once a week on a rolling 60-day window, and every row falls into one of a handful of actions: negate, promote, move, re-bid or watch.
Where to find it and how to set it up
In Campaign Manager, go to Measurement and Reporting, then Sponsored Ads Reports. Choose Sponsored Products, report type Search term, and set the time unit to summary rather than daily so each term appears once. Set the period to the last 60 days. Sponsored Brands has its own search term report with fewer columns; Sponsored Display does not report search terms because it is not keyword-driven.
Schedule the report weekly so it is waiting on the same morning each week. Everything else in this article assumes the summary view of the last 60 days.
The columns that matter
The report has more than twenty columns. Seven carry the decisions.
| Column | What it tells you | Used for |
|---|---|---|
| Customer search term | What the shopper typed, or the ASIN the ad showed on | Everything |
| Targeting | The keyword or target that matched it | Whether the match was intended |
| Match type | Broad, phrase, exact, or auto group | Where a term is leaking from |
| Clicks | Volume of evidence | Whether there is enough data to act |
| Spend | Cost so far | Threshold for negatives |
| 7-day total orders | Sales the term produced | Promotion and negation |
| 7-day ACoS | Cost efficiency for the term | Bid adjustments |
The "7-day" in the column names refers to the attribution window, not the report period. An order counts if it happened within seven days of the click. This also means the most recent few days of any report are incomplete, because orders are still being attributed. Read the last week lightly.
The five sorts
Each week we run the same five sorts, in the same order, and each produces one kind of change.
Read this flowchart as text
- Search term report, 60 days, summary
- Sort 1: orders = 0, spend descending
- Negative exact where spend exceeds break-even per sale
- Sort 2: orders 2+, term not yet an exact keyword
- Add to Control campaign as exact. Negate in source campaign
- Sort 3: term converting in broad or auto while exact exists
- Negate in broad or auto. Exact campaign owns it
- Sort 4: ACoS far above or below target, 30+ clicks
- Adjust bid on the matching keyword 10–20%
- Sort 5: new terms with rising impressions, few clicks
- Watch list. Check relevance, revisit next week
Sort 1: spend with no orders
Filter orders to zero and sort spend high to low. The top of this list is where money leaves the account with nothing to show. Terms whose spend exceeds the break-even cost of one sale become negative exact candidates. Terms that are plainly for a different product become negatives regardless of spend. The rules for the threshold are in Negative Keywords on Amazon: How and When to Add Them.
Sort 2: winners not yet in exact match
Filter orders to two or more, then compare each term against the exact-match keywords already in the account. Terms that convert but exist only as broad or auto matches are ready for promotion: add them as exact keywords in the control campaign with a bid set from their own conversion rate, and add them as negative exact in the campaign they came from. This is how the account's keyword list grows from evidence rather than guesswork, and it is the routine in B14 · How a search term earns its own campaign.
Sort 3: right term, wrong campaign
Look for terms that already exist as exact keywords but are still receiving clicks through broad, phrase or auto matches. This means the negative from sort 2 was never added, or the term entered through a different campaign. Add the negative so the exact campaign, with its own bid and placement settings, owns the term.
Sort 4: ACoS outliers with enough clicks
Filter to terms with at least 30 clicks, then sort by ACoS. The top and bottom of the list are bid decisions. High ACoS on a converting term: lower the bid on the keyword that matched it. Very low ACoS with impression share to gain: raise the bid. Terms with fewer than 30 clicks are not ready for a bid decision, whatever their ACoS says.
Sort 5: new and rising
Sort by impressions and look for terms that have appeared recently with few clicks. These are the early signals of a new search behavior, a seasonal term or a competitor's new product name showing in your auto campaign. Most need nothing yet; a few need a relevance check. Keep them on a watch list.
Reading product-targeting rows
In auto campaigns and product-targeting campaigns, the search term column shows an ASIN rather than a phrase. The same logic applies. An ASIN that spends with no orders is a negative product target. An ASIN that converts is a candidate for its own product-targeting campaign, where you can bid on it deliberately. Look the ASIN up before acting; a competitor's page that converts for you is worth more than a complementary product that occasionally does.
What the report cannot tell you
It does not show searches where your ad was eligible but did not appear, so it cannot tell you what you are missing. Brand Analytics' Search Query Performance covers that gap for brand-registered sellers. It does not show organic performance, so a term that looks weak in ads may be carrying strong organic sales. And it does not know your margin; every threshold in this article has to be set from your own numbers.
Frequently asked questions
How often should I check the Amazon search term report?
Weekly, on a rolling 60-day window, with the report scheduled so it arrives on the same day each week. Daily checks react to incomplete attribution; monthly checks let wasted spend run for weeks.
What is the difference between a keyword and a search term?
The keyword is what you bid on. The search term is what the shopper typed. A broad-match keyword steel kettle can match the search term stainless steel electric kettle 1.7l. The report shows the second; your campaign shows the first.
Why do my search term report sales not match my campaign sales?
Attribution timing. The report attributes orders within seven days of a click, and recent days keep updating for up to a week. Compare the same date range a week later and the figures converge. Very recent data should be read as provisional.
Does the search term report show Sponsored Display data?
No. Sponsored Display targets audiences and products, not keywords, so there are no search terms to report. Sponsored Brands has its own search term report with a smaller set of columns.