ROAS and ACoS explained: how to tell if your ads are making money

Here's a lesson I paid for. I was going through Amazon settlement reports for a few months of sales, and one period stood out: advertising spend had consumed nearly all of the revenue. Sales looked fine. The ad dashboard didn't look alarming. But when everything was added up, the ads had eaten what the products earned. One listing, a phone wallet, had made zero sales during that time while still racking up ad charges every day.

Nothing about it was hidden. I just wasn't comparing ad numbers to the one number that matters, my margin. This guide is the math I should have been doing.

The three numbers

ROAS = ad sales ÷ ad spend
ACoS = ad spend ÷ ad sales
TACoS = ad spend ÷ total sales

• ROAS (return on ad spend) says how many dollars of sales you got per ad dollar. 4x means $4 of sales for every $1 of ads. Walmart, Google, and most ad platforms report this.
• ACoS (advertising cost of sale) is the same ratio flipped: 25% means a quarter of ad-driven sales went to ads. Amazon reports this. ACoS = 1 ÷ ROAS.
• TACoS (total ACoS) divides ad spend by all your sales, including organic sales that didn't come from ads. It's the best single number for "are ads helping the business?"

Break-even ACoS: the number that matters

None of those numbers tell you whether you made money. For that, you need your margin before ads: what's left of the price after the item, marketplace fees, fulfillment, and shipping, as a percentage of the price.

break-even ACoS = margin before ads
break-even ROAS = 1 ÷ margin before ads

If you keep $7.50 on a $25 sale before ads, your margin is 30%. You can spend up to 30% of ad sales on ads and break even. At a 25% ACoS you make money. At 35% you lose money on every ad sale, no matter how good the sales volume looks.

What a "good" ACoS is

You'll see people say 15% or 25% or 30% is a good ACoS. The honest answer is that a good ACoS is anything comfortably below your break-even, and break-even is different for every product. A few practical targets:

• Launching a product: running near or even a little above break-even for a set time can make sense, because the sales build history, reviews, and organic rank. Decide the time limit and budget before you start.
• Established products: aim well below break-even, leaving real profit after ads.
• Low-margin products: be very careful. At a 15% margin, break-even ROAS is 6.7x, which is hard to sustain. A lot of cheap items simply can't be advertised profitably as singles. That's another reason to sell them as multi-packs or bundles.

Why TACoS matters

Ads can do two jobs: get the sale they're credited with, and push the listing up in organic search so it gets more sales on its own. ACoS only sees the first job. TACoS sees both.

If you increase ad spend and total sales grow faster, TACoS falls, which means the ads are lifting organic sales. If TACoS keeps climbing while total sales stay flat, the ads are just replacing sales you'd have gotten anyway, and you're paying for them.

The mistakes that drain ad budgets

Most ad blowups come from the same few mistakes. The first one was mine:

• Spending on a listing that couldn't convert. The phone wallet got clicks and no sales. Ads can't fix a listing people don't want to buy from. If a product gets clicks and no orders, fix the price, images, or reviews before spending more, or pause it.
• Budgets without a margin rule. A daily budget limits how much you spend, not how much you lose. You also need a rule like "pause anything above break-even ACoS for 14 days."
• Checking the ad console instead of the settlement. The ad console shows what ads did. The settlement report shows what you kept. Check both.
• Automatic campaigns left running unchecked. Auto campaigns are great for discovering search terms, but they also find irrelevant ones. Review the search term report and add negatives every week or two.

A simple routine that works

1. Calculate break-even ACoS for every advertised product and write it down next to the campaign.
2. Start with an automatic campaign at a small daily budget to find out what search terms convert.
3. Every week: pause keywords with lots of clicks and no sales, add irrelevant search terms as negatives, and move proven search terms into a manual campaign where you control bids.
4. Every month: compare total ad spend to your settlement or payout report, and look at TACoS.
5. Kill rule: any product that's above break-even ACoS for a month with no sign of improving gets its ads paused until the listing or price changes.

Walmart's Sponsored Search runs on the same math, it just reports ROAS instead of ACoS. Getting an ad account set up there has its own hurdles, which I cover in setting up Walmart Sponsored Search.

Questions people ask