All Issues
Issue #1
April 8, 2026

Why Your Amazon Advertising Data is Lying to You

Most advertisers trust their ACoS. Here's why that number is almost always wrong — and what to look at instead.

advertisingdata-integrity

Every Amazon advertiser knows ACoS. Advertising Cost of Sale. The number every dashboard shows front and center.

There's just one problem: it's almost always wrong.

The Attribution Gap Nobody Talks About

Amazon's default attribution window is 14 days. That sounds reasonable — until you realize it means a customer who clicked your Sponsored Product ad on March 1st and bought on March 14th gets counted.

But a customer who clicked March 1st and bought March 15th? Not counted.

That's not a bug. That's by design. And it means your ACoS is measuring a different population of buyers than the ones you're actually converting.

The result: You're optimizing for the fast converters and ignoring the deliberate buyers. In most categories, the deliberate buyers are more valuable.

What to Look at Instead

Three metrics that give you a more honest picture:

1. Total ROAS (not just attributed ROAS) Pull your total ad spend against total revenue for the same period — not just attributed revenue. If your attributed ROAS is 4x but your total ROAS is 1.8x, your attribution model is flattering you.

2. New-to-Brand % This is buried in Amazon's campaign reports but it's gold. It tells you how much of your ad spend is acquiring genuinely new customers vs. retargeting existing ones. A healthy new-to-brand percentage is above 60% for growth-stage brands.

3. Search Term Impression Share If you're winning on your top terms but your impression share is low, you're leaving budget on the table in the moments that matter most.

The Honest Benchmark

Across MixShift's customer base of Amazon brands and agencies:

  • Average ACoS as reported by Amazon: 28%
  • Actual blended ACoS when total ad spend is divided by total revenue: 41%

That 13-point gap is the attribution flattery tax. Almost every brand is paying it without knowing it.

What This Means for Your Strategy

Stop optimizing for ACoS alone. Build a reporting layer that sits above Amazon's native numbers and shows you:

  • Total spend vs. total revenue (not just attributed)
  • Week-over-week trend, not just current snapshot
  • Category-level breakdown so you know which product lines are genuinely profitable

That's exactly what MixShift's Report Center is built to show you. But even if you're not using us — build this view manually. It'll change how you allocate budget.


That's Issue #1. If this was useful, forward it to one Amazon brand operator you think would benefit. The list grows when the content earns it.

Weekly AI Shifts

Get this every week.

Join Amazon brand owners and agency operators who read Weekly AI Shifts every Tuesday.

No spam. Unsubscribe anytime.