The Sponsored Products search term report shows the actual words shoppers typed before clicking your ad, next to the keyword or target that matched them, with spend, clicks, orders and sales for each. Sort it by spend, then sort terms into four buckets: winners to move into exact match, promising terms to keep, waste to add as negatives, and terms that don’t have enough data yet. Doing this weekly is the single most useful habit in Amazon PPC.
The Amazon search term report tells you exactly what shoppers typed before they clicked your Sponsored Products ad, and what each of those searches cost and earned you. To read it, sort by spend, then decide for each search term whether it’s a winner to target directly, waste to block with a negative keyword, or something that needs more data. That weekly habit does more for most ad accounts than any bid tweak.
The report looks intimidating the first time you open it. It’s a spreadsheet with a couple dozen columns and hundreds of rows. But you only need about eight of those columns, and the decisions come down to a few simple rules.
How do you download the search term report?
- Open the Amazon Ads console. You can get there from Seller Central through the Advertising menu.
- Go to Sponsored ads reports (under the reporting or measurement section of the left menu).
- Click Create report. Choose Sponsored Products as the ad product and Search term as the report type.
- Pick the time unit (Summary is easiest to read) and a date range. Last 30 days is a good default.
- Run the report and download it when the status shows it’s ready. You can also schedule it to repeat.
Amazon only keeps a limited window of search term history, so save a copy each month if you want to look back further later. Screens move around; as of September 2026, if you can’t find it, search “search term report” in the Ads console help.
Sponsored Brands has its own search term report. Sponsored Display doesn’t, because it isn’t keyword-driven.
Which columns actually matter?
Hide or ignore everything except these:
| Column | What it tells you |
|---|---|
| Campaign Name / Ad Group Name | Where the spend happened, so you know where to make changes. |
| Targeting | The keyword or product target you bid on. In auto campaigns this shows the auto targeting group, like close match or substitutes. |
| Match Type | Broad, phrase or exact for keywords. Tells you how loosely the search was matched. |
| Customer Search Term | What the shopper actually typed. If it looks like an ASIN, the click came from a product page. |
| Clicks | How many times shoppers clicked your ad from that search. |
| Spend | What those clicks cost. |
| 7 Day Total Orders | Orders attributed to those clicks within the report’s attribution window. |
| 7 Day Total Sales | The sales value of those orders. |
From these you can calculate everything else: conversion rate (orders divided by clicks), cost per click, and ACoS (spend divided by sales). The report includes some of these already, but it helps to know how they’re built.
One quirk worth knowing: “Total” sales include orders for other products from your brand bought after the click, not just the advertised item. There are separate columns for advertised SKU units if you want to see the difference.
What’s the difference between Targeting and Customer Search Term?
This is the part that clicks everything into place. Targeting is what you told Amazon to bid on. Customer Search Term is what the shopper typed.
With an exact match keyword, the two are basically the same (Amazon allows close variants like plurals). With broad or phrase match, and with auto campaigns, one target can match dozens of different searches. You might bid on “bamboo cutting board” in broad and find you paid for “cutting board set,” “bamboo plates” and “wood cutting board for meat.” Some of those are great. Some are not your product at all.
That gap between what you bid on and what you paid for is where most wasted spend hides.
How do you spot wasted spend?
Sort the whole sheet by Spend, highest first. Then filter for rows with zero orders. You’re looking for three patterns:
- Irrelevant terms. Searches that clearly aren’t your product, like a different material, a different use, or a competitor’s unrelated item. These don’t need much data. If it’s obviously wrong, negate it.
- Relevant terms with lots of clicks and no orders. The search looks right, but it isn’t converting. Before negating, compare its clicks to your usual conversion rate. If you normally convert 1 in 10 clicks and a term has 25 clicks and no orders, that’s a real signal. If it has 4 clicks, it’s noise.
- Converting terms with an ACoS far above break-even. These sell, but you pay too much for each sale. The fix is usually a lower bid on an exact match target, not a negative.
For how to judge “enough clicks,” and how negative exact and negative phrase differ, see how to use Amazon negative keywords.
How do you spot winners?
Now sort by orders, or filter for ACoS below your target. Look for search terms that:
- Have several orders, not just one lucky sale.
- Have an ACoS at or below your target.
- Came from an auto, broad or phrase campaign, and aren’t yet targeted as exact match anywhere.
Those are your harvest. Add each one as an exact match keyword in a manual campaign, where you can set its bid directly. Then add it as a negative exact in the campaign it came from, so the two campaigns don’t compete for the same search and muddy your data.
A worked example
Here are four rows from an example report for a $30 bamboo cutting board. These numbers are made up for illustration, and the seller’s target ACoS is 30%.
| Search term | Clicks | Spend | Orders | Sales | ACoS | Action |
|---|---|---|---|---|---|---|
| bamboo cutting board large | 60 | $60 | 8 | $240 | 25% | Winner: move to exact |
| cutting board set | 40 | $44 | 2 | $60 | 73% | Lower bid or watch |
| bamboo plates | 15 | $15 | 0 | $0 | – | Irrelevant: negate |
| wood chopping block | 5 | $6 | 0 | $0 | – | Too early to judge |
In this example, “bamboo cutting board large” gets its own exact match keyword with a bid you control. “Bamboo plates” is a different product, so it becomes a negative exact right away. “Cutting board set” sells, but at 73% ACoS the seller is losing money on each ad sale, so they’d lower the bid on it or target it in exact with a smaller bid. “Wood chopping block” stays untouched until it has more clicks.
What should you do with it each week?
- Download the last 30 days, not just the last 7, so each term has enough data.
- Sort by spend. Negate obvious irrelevant terms.
- For relevant zero-order terms, compare clicks to your normal conversion rate before negating.
- Sort by orders. Move proven winners into exact match and negate them in the source campaign.
- Write down what you changed, so next week you can see whether it helped.
Keep your target in mind while you do this. If you don’t know your break-even ACoS yet, work it out first with ACoS vs. TACoS. Otherwise you’re sorting terms into “good” and “bad” without a line to measure against.
Shelfmate reads your search term data every week, flags likely waste and harvestable winners against your own margin, and proposes negatives and new exact match keywords. Nothing changes in your account until you approve it. More on the ads side of Shelfmate.
You don’t need to love spreadsheets to do this well. You just need to look at the same eight columns every week and make a few small, boring decisions. That consistency beats clever strategy almost every time.
Frequently asked questions
Where do I find the search term report on Amazon?
In the Amazon Ads console (also reachable from Seller Central under Advertising), open Sponsored ads reports, create a report, choose Sponsored Products, and pick Search term as the report type. Set a date range and download the file when it’s ready.
What is the difference between a keyword and a search term on Amazon?
A keyword is what you bid on. A search term is what the shopper actually typed. A broad or phrase match keyword like “dog bowl” can match many different search terms, and the report shows each one separately.
Why doesn’t my search term report show every search?
The report generally lists search terms that led to at least one click, not every search where your ad appeared. Some rows for product targeting show an ASIN instead of words, meaning the shopper clicked your ad on that product’s page.
How often should I review my Amazon search term report?
Weekly is a good rhythm for most small brands. Look at a longer window, such as the last 30 days, when deciding on negatives, so each term has enough clicks to judge fairly.