Which numbers to trust

The ads console shows you a wall of metrics. You don’t need most of them to make decisions. There’s an order to it.

ACOS comes first. It’s ad spend divided by ad sales, so a $25 spend that brings in $100 of sales is a 25% ACOS. (ROAS is the same ratio flipped, $100 ÷ $25 = 4. Use whichever you like, they always agree.) ACOS goes into almost every decision you make.

The only time you don’t use ACOS is when you don’t have one, because there are no orders yet, or so few that the number means nothing. With one order, ACOS is mostly luck.

Then clicks. With no orders, clicks tell you how much of a chance something has had. That’s what you use to decide when to pause a keyword or negate a search term.

Then impressions. With no clicks, impressions tell you whether the ad is even showing.

Everything else (click-through rate, cost per click, impression share) helps you understand a decision, but shouldn’t make it. If a call is close, they can break the tie.

Target ACOS

ACOS needs something to compare against, otherwise it’s just a number. That’s your target ACOS.

A good starting point is break-even. Take what the product sells for, subtract product cost, Amazon’s fees and shipping, and what’s left as a share of the price is your margin. If your margin is 40%, a 40% ACOS breaks even on the ad sale. Every $1 of ads brings back $1 of profit to cover it.

That sounds like you’re not making money, and on the ad sale alone you’re not. But you’re getting the sale and the ranking that feeds the organic side, which is the point of lesson 1. From there, go lower if you want the ads themselves profitable, or higher if you’re launching and buying ranking on purpose.

Your target can change by product and over time. Just make sure every product has one.