AdModifier.

Which of your search terms are burning money?

Drop in your Amazon search-term report. You'll see every term that took 10+ clicks and produced no orders — the same threshold we use on client accounts — and what it cost you over the window you exported.

Choose your search-term report or drag it here · .csv
🔒 Your file never leaves your browser. Nothing is uploaded.
  1. Amazon Ads → Reporting → search-term report, last 60 days
  2. If it downloads as .xlsx: open it and File → Save As → CSV
  3. Drop it above

What this is actually checking

One rule, applied to every row in your file: 10 or more clicks, zero orders, over the window you exported. That is the same threshold we negate on in client accounts.

The reason it is ten and not two: most search terms convert somewhere in the single-digit to low-double-digit percentages, so a few clicks without a sale is ordinary variance rather than evidence. Ten clicks and nothing is where we stop calling it bad luck. It is a working threshold, not a law — raise it for an expensive product that converts slowly, lower it for a cheap impulse buy.

The second list is the opposite problem: terms with 2 or more orders that are still sitting inside an auto or broad campaign, where you are paying discovery prices for a keyword that has already proved itself.

Getting the right file, in about a minute

  1. Amazon Ads console → Reporting → Create report
  2. Report type: Search term. Not "Targeting", not "Advertised product" — those do not contain customer search terms, so the calculation cannot run
  3. Campaign type Sponsored Products, time unit Summary, last 60 days
  4. If it downloads as .xlsx, open it and File → Save As → CSV

Sixty days is a deliberate choice. Shorter and a slow-converting product looks like waste; much longer and you are judging bids you have already changed.

What to do with the list

Negate as exact, in the ad group the spend happened in. A phrase or broad negative will take out matches you have not looked at, and a campaign-level negative is a blunter instrument than you need.

Negate at the source as well as harvesting. If you move a converting term into its own exact-match ad group and leave it live in the auto campaign, the two compete and you pay twice for the same customer.

Change one thing, then wait. Negations show up in days; bid changes take longer. Doing both at once means you will not know which one worked.

What this will not tell you

Honest limits, because a tool that only lists its strengths is selling something:

Questions people actually ask

Why 10 clicks and zero orders?

Because one or two clicks with no sale tells you nothing - most terms convert somewhere between 5% and 15%, so a handful of clicks without an order is ordinary variance. Ten clicks with nothing is the point where we stop calling it bad luck on a client account. It is a working threshold, not a law of nature: raise it for a high-priced product that converts slowly, lower it for a cheap impulse buy.

Does my file get uploaded?

No. The parsing happens in your browser using the File API, so the report never travels to a server. If you open your network tab while running it you will see no upload. The only thing this site ever receives is an email address, and only if you type one in.

Which report do I need?

The Sponsored Products search-term report from Amazon Ads - not the targeting report and not the advertised-product report. It is the only one that lists the actual customer search terms alongside clicks, spend and orders, which is what makes the calculation possible.

My export is .xlsx, not .csv

Open it and use File then Save As, choosing CSV. Amazon changes which format it hands you depending on the report and the console you are in; the numbers are identical either way.

It says my wasted spend is small. Is something wrong?

Probably not - it means your negatives are doing their job, which is the good outcome. A tight account is a real result and worth knowing. In that case the money is more likely sitting in bids, placements or the listing itself, none of which this file can see.