August 1st, 2026   •   5 minute read   •   Issue #165

"How do you analyze your Facebook Ads data, and how do you know when to scale, refresh the creative, or turn an ad off?"

That question landed in my inbox this week. It's a good one, and it comes up constantly for authors running Facebook Ads to sell books on Amazon. The honest answer starts with this:

Facebook cannot tell you if your ads are working.

Facebook knows the click happened. Everything after that click, whether the reader landed on your Amazon page, bought the book, borrowed it on Kindle Unlimited (KU), read it, went on to buy book two, all of that lives outside Facebook. Meta is guessing at what happens next, and usually flattering itself in the process.

If you're only reading Ads Manager, you're reading one third of the story.

You need three dashboards side by side to get a fuller picture of what the ads are actually doing. Four if you want the bonus view.

Facebook Ads Manager Tells You About The Click

This is the top of the funnel. You're asking whether people are stopping the scroll and clicking through.

And here's a shift most authors miss.

Use the Unique Outbound versions of these metrics wherever Facebook offers them:

→ Unique Outbound CTR (Click-Through Rate)

→ Unique Outbound CPC (Cost-Per-Click).

→ Unique Outbound Clicks.

"Unique" means unique people. If one reader clicks your ad three times, that's one unique click, not three. A much closer read on what's actually happening.

→ Unique Outbound CTR. Aim for 2% or higher. Below that, the creative isn't earning its impressions.

→ Unique Outbound CPC. Watch the trend, not the absolute number. A single-day spike is noise. Costs drifting up week after week is a signal that the ad is fatiguing.

→ Frequency. Once you push past 2.5, the same people are seeing the ad again and again, CTR usually starts sliding and CPC usually begins to climb.

Facebook metrics tell you whether the ad is doing its job of getting a qualified click. That's all.

Amazon Attribution Is A Signal, Not A Scoreboard

Amazon Attribution links let you see how many of your Facebook clicks turned into detail-page views, sales and Kindle Unlimited (KU) Page Reads on Amazon. Useful. But treat it as a directional signal, not a source of truth.

In my experience, Amazon Attribution is roughly 70% accurate on sales, and only 30 to 40% accurate on KU page reads. The numbers you see there are an under-count, not the full picture.

Attribution is still useful for head-to-head ad decisions.

If you're running two ads against each other, and one has some attributed sales while the other has zero, you can turn off the zero one with confidence. That comparison Attribution handles well, even at its accuracy level.

Likewise, if you see in Attribution that one ad has 6 sales and 3,500 page reads, and another has 1 sale and 0 page reads, the first ad is a candidate for scaling, whereas the second ad is not.

KDP Reports Is Where The Truth About Royalties Lives

Not Facebook. Not Attribution. KDP.

Line up your ad spend for a period against the royalties earned on the advertised book (and, ideally, the whole series or catalog) in the same period. That's your read on whether the ads are paying for themselves, and whether they're starting to compound into profit.

It's simple, but it's the only view that includes everything the ads have set in motion. Sales. KU page reads. Follow-on books. All of it.

If your KDP royalties are climbing faster than your ad spend over a few weeks, the ads are doing their job (you're profitable), whatever Attribution says.

Author Central Is Your Bonus View

Author Central shows you the Amazon Bestseller Rank (BSR) trend for the advertised book, and for every other book of yours published on Amazon.

When the ads are working, you'll see BSR improve not just on the advertised book, but on the next book in the series, and sometimes on unrelated titles that share a reader. That's read-through and sell-through showing up in real time. A quiet sanity check on top of the KDP numbers.

So What Do You Actually Do With All Of This?

Three decisions.

Scale when the whole picture holds. Unique Outbound CTR healthy. Unique Outbound CPC stable or drifting down. Frequency still low (below 2.5). Attribution showing at least some sales and/or page reads. KDP royalties tracking upward against your spend. Nudge budgets up by around 20% at a time and give each change a few days to settle. Big jumps confuse the algorithm and reset what it has learned.

Refresh the creative when Facebook engagement droops. Unique Outbound CTR sliding. Unique Outbound CPC rising as a sustained trend, not a spike. Frequency climbing. The audience is fatiguing on that hook or image, not on the book. Time to test some new creative.

Turn it off when the head-to-head in Attribution shows one ad earning sales and its sibling showing none, or when a refresh doesn't revive it. Some ads just don't find their reader. That's data, not failure.

One last thing. Give any new ad seven days minimum before you judge it, and don't touch it in that window.

Poking at it every day teaches the algorithm nothing except that you're impatient and prevents it from learning who your readers and where to find them.

Let it breathe.

That's it for this week. Thanks for reading. See you next Saturday.

To Your Success
– Matt