How to check what your Amazon ads AI tool spent

Download your search term report

The search term report shows what shoppers searched for before they clicked your ad, with the cost and the orders. It is the one file that shows where the ad money went.

Amazon describes it as the report of shopper searches “that resulted in at least one click” (Reports for book advertising).

In the Amazon Ads console, open Measurement & Reporting, then Sponsored ads reports, then Create report (How to navigate the advertising console). Pick Sponsored Products and the search term report, choose at least the last 30 days, choose Daily if it asks for a time unit, and download it as a CSV file.

Amazon is also rolling out a newer report builder, labelled Reporting (Beta), with its own Create report button (Amazon’s announcement). There you pick the columns yourself.

Which columns to tick

Newer custom report: Needed: Date range, Campaign ID, Campaign name, Target ID, Targeting match type, Search term, Clicks, Total cost, Purchases. Recommended: KENP read, Sales, Budget currency. That report has no keyword column, so duplicate keywords aren’t checked.

Older report: Needed: Campaign Name, Customer Search Term, Targeting, Clicks, Spend, Orders (7 or 14 day), and Date (or Start Date and End Date). Recommended: Campaign ID, Ad Group Name, Match Type, KENP Read, Country or Marketplace, Sales, and Currency.

These are the columns Billevera’s free check reads, so a file with them works for the last step too. If you can’t find the report, search the Amazon Ads help for “search term report”.

Work out your break-even per sale

Your break-even is what one sale earns you before ads: your royalty per book, or your margin per product. An ad that costs more than that for each sale costs more than the sale brings in.

Amazon’s own measure is ACOS, short for advertising cost of sales: ad spend divided by ad sales, times 100. Amazon says your ACOS needs to be lower than your profit margin to stay profitable (Amazon’s ACOS guide).

Comparing cost per sale with your royalty asks the same question in dollars, which is easier to check on a calculator.

Example with made-up numbers: a book earns you $3.50 a sale. A campaign that spent $14.00 for 2 sales paid $7.00 a sale, twice your break-even.

Look for spend with no orders over at least 14 days

Sort the report by spend. Look at the campaigns, keywords and targets that spent at least your break-even over 14 days or more with no orders.

A few days is too short to judge. A few clicks is too, because some clicks with no sale are plain chance: 5 clicks with no sale says little; 70 says much more.

If your report covers fewer than 14 days, download a longer one before you change anything.

Compare cost per sale with your break-even

For each campaign with orders, divide spend by orders. A cost per sale above your break-even is worth a look.

Wait for at least 2 orders: one sale says too little.

To skip rounding, multiply instead of dividing: is the spend more than your break-even times the orders?

Find search terms that never sold

Add up what each search term cost across all your campaigns. A term that cost at least your break-even with no orders is a candidate negative keyword.

A negative keyword tells Amazon not to show your ad for that search. A tool may run the same search term in several campaigns, so a term can look cheap in each one and still add up.

Read each term before you act on it. You know your book or product better than any tool does.

Check each campaign runs where you sell

Make sure every campaign runs in the marketplaces you chose and advertises the books or products you meant.

A tool may copy a campaign into another country’s marketplace, or advertise an old title you no longer push. In the older report, the Country or Marketplace column shows where each row ran.

If your books are in Kindle+ (formerly Kindle Unlimited), count page reads

Page reads in Kindle+ earn royalties but are not orders. Keep the KENP read column, or a book that earns from page reads can look like it never sold.

KENP stands for Kindle Edition Normalized Pages. KDP’s help pages now call the program Kindle+, and say your payment depends on your share of the total pages read across it (Royalties in Kindle+). The money per page changes from month to month, so page reads can’t simply be added to orders.

Treat a search term with page reads and no orders as too early to tell, not as a dud.

Judge discovery campaigns by what they find

Automatic (discovery) campaigns are meant to find new search terms. Judge them by the terms they find that go on to sell, not only by their own cost per sale.

In an automatic campaign, Amazon matches your ad to searches and products similar to yours, in four groups: close match, loose match, substitutes and complements. Amazon suggests reading the automatic campaign’s search term report to find what to target in manual campaigns, adding the top performers there, and letting an automatic campaign run about two weeks first (Amazon’s targeting guide).

So an automatic campaign with a high cost per sale may still have done its job. Ask: did a search term it found move into a manual campaign and sell there?

Billevera’s free check doesn’t make this distinction yet. It applies the same rules to automatic campaigns, and only leaves their automatic targets out of the duplicate-keyword check. Read its lines on these campaigns with this step in mind.

Let the free check do the sums, if you like

Billevera’s free check reads the same report in your browser and lists the campaigns and search terms worth a look, with the money and the rows behind each one.

You give your break-even per sale and your marketplaces, and tick a box if your books are in Kindle+. It uses 14 days and 10 clicks as minimums and marks anything with less as too early to tell. It also checks the tool against your own spending rules. Your files are not uploaded.