The Amazon Search Term Report: Five Cuts Worth Making
The search term report is the only place Amazon tells you what people actually typed before clicking your ad. Everything else — keywords, targets, bids — is what you guessed they might type.
That makes it the highest-value report in the account, and the most commonly misread.
Search terms are not keywords
The distinction trips up a lot of sellers.
A keyword is what you bid on. A search term is what a shopper typed. Broad and phrase match mean the two are often nothing alike — you bid on cat mom shirt and Amazon serves you for funny gifts for cat lovers, crazy cat lady mug and cat mom alike.
An automatic campaign has no keywords at all. Amazon chooses the search terms, and the report is the only way to see what it chose.
So the report answers two different questions: which of my keywords are matching to nonsense? and what did Amazon discover that I should be bidding on deliberately?
Getting the report
In the ads console: Measurement & Reporting → Sponsored Products → Search Term Report, choose a date range, download.
Two things worth knowing before you rely on it.
Conversions lag. Amazon attributes an order to a click for up to 7 days after that click. A term pulled today is missing orders that haven't landed yet, so a 7-day report always looks worse than the same week will look next week. Pull at least 30 days for anything you plan to act on. Low-volume terms are hidden. Amazon suppresses search terms below a privacy threshold. Your report's clicks won't quite reconcile with your campaign totals, and that's expected rather than a bug.The five cuts worth making
Open the file and do these in order. It takes about twenty minutes.
1. Spend with zero orders
Sort by spend, descending. Filter to orders = 0.
Then apply a threshold rather than a feeling. If a term converted at your normal rate, the chance of seeing zero orders across n clicks is (1 − CVR)^n. At a 10% conversion rate:
- 10 clicks → 35% chance it's just luck. Leave it.
- 20 clicks → 12%. Watch it.
- 40 clicks → 1.5%. Negate it.
Cutting at 10 clicks feels productive and kills good terms about a third of the time. Use your own conversion rate — at 5%, double every figure.
2. Converting terms above your break-even
These are the expensive ones because they look like they're working.
A term at 25% ACoS is a win on a private-label product with 40% margin and a loss on a $16.99 shirt paying a $2.02 royalty, where break-even is under 12%. Sort by ACoS descending among terms that did convert, and compare each to your own break-even rather than to a benchmark. The free break-even calculator gives you the number.
3. Winners to promote
Terms with orders and an ACoS comfortably under break-even belong in a manual campaign where you set the bid deliberately.
The half people skip: when you promote a term to manual, add it as a negative exact in the automatic campaign that found it. Otherwise the auto campaign keeps serving that term at its own bid and you never gain the control you built the manual campaign for. Promotion and negation are one action with two halves.
4. Intent mismatches
Some terms are wrong regardless of volume — negate on sight at one click:
free,template,svg,png,mockup— people making a shirt, not buying oneused,second hand,wholesale,bulk- Formats you don't sell:
audiobook,hardcover,pdf - Competitor brand names you can't legally target
5. High impressions, no clicks
Rarely acted on, and it isn't a bidding problem. If a term shows thousands of impressions and almost no clicks, shoppers are seeing your listing and choosing something else — your image, title or price is losing the comparison. Negating it hides the symptom; fixing the listing is the actual work.
Reading it for an automatic campaign
Auto campaigns bid across four targeting types — close match, loose match, substitutes and complements — and they behave nothing alike. Substitutes puts you on a competitor's product page beside something the shopper is already considering. Complements bids you into a basket they've already assembled.
The search term report doesn't split these out for you, but the bulk file does. Before concluding an auto campaign is failing, check whether three targeting types are fine and one has never converted. Pausing that one type is usually the fix, not pausing the campaign.
At scale, use the bulk file instead
For a handful of campaigns the console report is enough.
Past that, the search term report becomes awkward: it tells you what was searched but not what to change, and every negative has to be added by hand, one at a time, in the interface.
The bulk operations file carries every campaign, ad group, keyword, target and placement in one sheet. You can sort the whole account by spend, apply your click threshold across every term at once, write the negatives directly into the file, and upload it back in a single action. An afternoon of console clicking becomes about ten minutes.
That's the workflow PPC Optimizer Pro automates — it applies your thresholds against your real break-even, with royalty attributed per design and returns subtracted, and returns a bulk file with the negatives and bid changes already written.
A repeatable routine
Every two to four weeks:
That's the whole report. Twenty minutes, and it's usually the highest-return twenty minutes available in an Amazon ads account.
Keep reading
Written by
PPC Optimizer Pro Team
The PPC Optimizer Pro Team consists of Amazon sellers and developers who built this tool after years of managing Sponsored Products campaigns manually. We share data-driven strategies to help sellers reduce wasted ad spend and improve ACOS.