Your Amazon ACOS Is Lying to You (If Your Books Are in Kindle Unlimited)

You open the Amazon Ads console, sort by ACOS, and start pausing every campaign above 70 percent. It feels responsible. It feels like cutting waste.

For a Kindle Unlimited author, it is often the single most expensive mistake you can make.

Here is the uncomfortable truth: the ACOS number Amazon shows you is not your real ACOS. For any book enrolled in KDP Select, that number is structurally incomplete, and it is incomplete in a direction that makes profitable campaigns look like failures. Authors pause winners every single day because of it.

Let me show you exactly where the number breaks, and how to calculate what is actually true.

What ACOS actually measures

ACOS stands for Advertising Cost of Sale. The console computes it like this:

ACOS = Ad Spend / Attributed Sales Revenue

If you spend 10 dollars and Amazon attributes 20 dollars of sales to those clicks, your ACOS is 50 percent. Simple enough.

The problem is hiding inside the words "Attributed Sales Revenue." The ad console only counts one kind of revenue: a direct purchase of the exact book you advertised. That is the entire universe it can see.

Now think about what actually happens when a Kindle Unlimited reader clicks your ad.

The two kinds of revenue ACOS cannot see

1. KENP page reads from borrows. A KU reader clicks your ad and borrows the book. They do not buy it. They read it. Those pages turn into KENP royalties in your KDP dashboard at roughly 0.0045 dollars per page (the rate was 0.004521 per page as of September 2025). A 400 page book read in full is worth about 1.80 dollars to you. The ad console never sees a cent of it, because no "sale" happened. Your spend still counts. The revenue does not.

2. Series read-through. A reader discovers Book 1 through your ad, finishes it, and reads or buys Books 2, 3, 4, and 5. Every one of those downstream books generates revenue that your Book 1 ad created. The console attributes none of it back to the campaign that started the chain.

So for a series author in KU, the ACOS number is missing borrow revenue and missing read-through revenue. Both are exactly the revenue your ads are working hardest to produce. The metric is blind to your best outcomes.

What this does to your decisions

Picture a Book 1 campaign with a console ACOS of 180 percent. By the "pause anything over 70 percent" rule, it is dead on arrival.

But this is Book 1 of a five book series, in Kindle Unlimited. Once you fold in the KENP borrows that ad drove, plus the readers who went on to devour Books 2 through 5, that "180 percent loser" might be running at a true ACOS of 45 percent. It could be one of the most profitable things in your entire account.

You would never know, because the console is showing you one slice of a much larger pie and calling it the whole pie.

This is not a rounding error. For KU heavy series authors, the gap between console ACOS and real ACOS is routinely the difference between "pause it" and "pour fuel on it."

How to calculate your real ACOS

You can do this by hand. It is tedious, but it is worth understanding before you let any tool do it for you.

Step 1: Find your read-through revenue per ad click. During an advertising period, look at your KDP dashboard. Estimate the average KENP pages a single ad-driven borrow generates, including pages read across the rest of the series. Multiply by the current KENP rate.

KU value per borrow = avg KENP pages per borrow x KENP rate

Step 2: Estimate your series read-through. If Book 1 grosses 20 dollars and the full series grosses 100 dollars per buyer cohort, each Book 1 conversion is really worth far more than the Book 1 price. A common rule of thumb is 50 to 60 percent read-through from Book 1 to Book 2, tapering through the series.

Step 3: Build the true revenue figure. Add it all up for the campaign:

True Revenue = Direct Book 1 sales
             + KENP royalties from ad-driven borrows
             + downstream series sales and borrows (read-through)

Step 4: Recompute.

Real ACOS = Ad Spend / True Revenue

That fourth number is the one that should drive your Scale, Hold, Fix, or Pause decision. Not the console number.

A worked example

Say a campaign spent 100 dollars this month.

• Console attributed sales: 55 dollars. Console ACOS: 182 percent. Looks terrible.
• Ad-driven KU borrows generated an estimated 18,000 KENP pages. At 0.0045 per page, that is 81 dollars of borrow revenue.
• Series read-through added roughly 40 dollars of downstream sales and borrows.

True revenue is 55 + 81 + 40 = 176 dollars. Real ACOS is 100 / 176, which is about 57 percent.

Same campaign. Same spend. One number says burn it down. The other says it is comfortably profitable and probably worth scaling. The only thing that changed is whether you counted the revenue your ad actually produced.

Why most authors never do this math

Because doing it by hand, per campaign, every week, across a whole catalog, is miserable. The KENP rate shifts monthly. Read-through changes per series. KDP and the ad console live in separate dashboards that do not talk to each other. So most authors either skip it entirely and trust the lying number, or they spreadsheet it once, get overwhelmed, and give up.

That gap is exactly why we built TrustACOS. You connect your numbers, and it folds KENP borrows and series read-through back into every campaign automatically, then hands each one a plain verdict: Scale, Hold, Fix, or Pause. No spreadsheet archaeology. Just the real number and what to do about it.

You can try the math on a single campaign right now, free and without an account, using the TrustACOS calculator. Put in a campaign you were about to pause. There is a real chance it has been quietly profitable the whole time.

The one thing to take away

Console ACOS is not wrong because Amazon is hiding something. It is incomplete by design, because the ad system genuinely cannot see borrow royalties or downstream series revenue. That is not a metric you should be making pause decisions on.

Before you cut another "high ACOS" campaign, calculate the real one. The winner you are about to kill might be the best ad you have.