ROAS vs ACoS, against your margin

ROAS vs ACoS is one calculation written two ways: sales divided by ad spend, or ad spend divided by sales. What matters is the line you judge either against, set by your margin, and the two measures, TACoS and MER, that catch what both miss.

ROAS vs ACoS: what is the difference?

ROAS, return on ad spend, is the sales an ad platform credits to your ads divided by what you spent. ACoS, advertising cost of sales, is the same two figures the other way up: ad spend divided by ad-attributed sales, shown as a percentage.

So each is the reciprocal of the other. To convert, divide 100 by the ACoS percentage to get ROAS, or divide 100 by ROAS to get ACoS as a percentage. As a campaign grows more efficient, ROAS rises and ACoS falls.

ROAS is sometimes written as a multiple and sometimes as a percentage, so check which a report uses before comparing. Neither figure is profit: both count revenue, often including GST, before the cost of the goods, the fees and the returns.

Which number does each platform report?

Amazon’s advertising reports lead with ACoS; Google Ads and Meta report ROAS, or conversion value against cost; other marketplaces’ consoles use their own names, such as ROI, for what is usually ROAS. Read each console’s definition before you put two in one table.

The names are the smaller problem. Each platform decides which sales count, and the differences can matter more than the ratio:

• Attribution window. Each platform credits sales within its own period after a click, or a view, and the defaults differ.
• Which products. Marketplace consoles can credit an ad with other products bought after the click, not only the one advertised.
• Which value. Marketplace consoles usually report the price the buyer paid, GST included; Google and Meta record whatever value your tracking sends, with or without GST and shipping.
• Which orders. Most platforms count an order when it is placed, before cancellations, returns and RTO, return to origin: orders that come back undelivered.

So two campaigns showing the same ROAS on two platforms may not be equally good. Put them on one basis, as set out below, before they share a report. For Meta specifically, our guide to Meta ads for D2C brands covers measurement in detail.

How do you find your break-even ROAS?

Divide your selling price, net of GST, by what one order leaves you after every cost except advertising. The result is your break-even ROAS, below which an ad-won sale loses money; the same margin, as a share of price, is your break-even ACoS.

What one order leaves you is its contribution before advertising. Work it out per product or product group, because margins differ, from these lines:

• Selling price, net of GST.
• Landed cost of the product: its cost delivered to you, with packaging.
• Marketplace commission and fixed fees, or payment-gateway charges on your storefront.
• Shipping out, and reverse shipping on returns.
• The expected cost of cancellations, returns and RTO, spread across delivered orders.

Illustration: one product in round numbers, to show the arithmetic only. These are not anyone’s real costs, not a benchmark and not a quotation.

The campaign clears its break-even line, so its spend earns back more contribution than it costs. One below the line loses money on every sale credited to it, however good its ROAS looks on its own.

Two cautions. Use contribution net of returns and RTO, or the break-even line sits too low. And compare like with like: if a console’s sales include GST, take GST out before you compare them with a break-even worked net of it.

What is a good ROAS or ACoS?

A good ROAS is one above your own break-even line; a good ACoS is one below it. No general figure exists, because break-even follows margin, which differs by category, by price point and even by size.

• Fashion. Returns and RTO cut deep into contribution, so break-even sits higher than the margin on paper suggests.
• Low-priced items. Fixed fees and shipping take a larger share of a small order, which pushes break-even up; bundles help.
• Consumables. Repeat orders can justify a lower first-order target, if your own data shows buyers return.
• Heavy or bulky goods. Shipping both ways makes every return expensive, so work out break-even after returns, not before.

So a ROAS figure quoted without a category and a margin, whether as a benchmark or a promise, tells you nothing about your account.

What TACoS tells you that ACoS cannot

TACoS, total advertising cost of sales, divides your ad spend on a marketplace by all your sales there, paid and organic. ACoS shows what ads cost against the sales credited to them; TACoS shows how much of the whole business advertising carries.

Read the two together, over weeks rather than days:

• ACoS steady, TACoS falling. Organic sales are growing faster than ad sales. Ads may be lifting the listing’s organic rank, or the product is selling on its own.
• ACoS steady, TACoS rising. The business leans more on paid sales each month. Check organic rank, ratings, price and stock before raising budgets.
• ACoS falling, TACoS rising. Ads may be buying sales you would have had anyway, often on your own brand terms. Test pausing them.
• Both rising. Spend is growing faster than sales of any kind. Cut back to the terms that clear break-even.

Not every console shows TACoS, so build it from the advertising report and the marketplace’s sales report for the same dates. A launch can run a high TACoS for a while by design; a mature product should not.

Why MER keeps every platform honest

MER, marketing efficiency ratio, is total revenue from your own order system divided by total marketing spend across every platform. It cannot say which campaign worked, but no platform can inflate it, so it checks the ROAS each platform claims.

Platforms overlap. Google and Meta can each claim the same storefront order within their own windows, so adding their attributed sales overstates what you sold. If every platform reports a healthy ROAS while MER falls, the platforms are counting the same sales.

Two refinements make MER sharper. Count revenue net of GST, cancellations, returns and RTO; and, where your data allows, work out a new-customer MER from first orders alone, because returning buyers often order without any ad.

Blended ROAS, a term you will also meet, is usually MER under another name. Our article on attribution after Consent Mode v2 explains why it belongs first in a weekly report.

Where does CAC fit beside ROAS?

CAC, customer acquisition cost, is ad spend divided by the new customers it won. ROAS judges orders; CAC judges customers. For a product bought once, the two tell the same story; for one bought again and again, CAC is the number to plan with.

A storefront selling a consumable can accept a first-order ROAS below break-even when its own records show what a customer’s later orders leave, and how soon they arrive. Without that history, treat every first order as if it were the last.

Marketplaces rarely let you reach a buyer again, so repeat value there shows up mostly as organic sales and reviews, which is another reason to read TACoS beside ACoS.

Which number should drive each decision?

Use each number for the decision it can answer, and none for the decisions it cannot.

How should you set ROAS and ACoS targets?

Start from break-even and decide how much contribution each campaign must leave after its spend, which depends on what the spend is meant to buy. A target is a business decision about profit now against growth later, written down before the campaign runs.

• Your own brand terms. Cheap to win, and many of those buyers would find you anyway, so hold them to a ROAS well above break-even, and test whether pausing them loses sales.
• Prospecting. Finding new buyers costs more, so accept a target nearer break-even, judged on new customers as well as ROAS.
• New products. A launch may run below break-even for a set period to earn sales history and reviews. Agree the period and the ceiling first, and treat the spend as an investment with an end date.
• Repeat-purchase products. A thinner first order can pay if buyers come back, but only if your own data shows they do.

Set targets per channel and per product group, not one for the account. A single target starves the campaigns that find new buyers and overfeeds the ones that harvest existing demand. Above them all, set the MER the business needs.

A lower ACoS is not always better. Cutting spend on terms that clear break-even lowers ACoS and loses profitable sales with it; the aim is the most contribution after ad spend, not the smallest ratio.

A ROAS means nothing until you know the ROAS at which you break even.

Why does platform ROAS overstate what you keep?

Because platforms count gross sales at the moment of order, and you keep net sales after delivery. GST, cancellations, returns, RTO and overlapping attribution all sit between the two, and on cash-on-delivery orders the gap can be wide.

1. Agree one value basis, net of GST and shipping charges, and send that value from your tracking.
2. Take out cancellations, returns and RTO at each channel’s own rate, from your order system.
3. Compare each platform’s purchases with your orders for the same dates, every week.
4. Set targets on the reconciled figure, never on the console’s.

Getting the value right at source is tracking work, covered on our page on conversion tracking setup. If RTO is the larger gap, start with our guide on how to reduce RTO.

Common mistakes with ROAS and ACoS

• Comparing an ACoS from one console with a ROAS from another without converting either.
• Using one target for products whose margins differ.
• Judging campaigns on sales that include GST against a break-even worked without it.
• Adding Google’s and Meta’s attributed sales together as if they were separate orders.
• Cutting a campaign on ACoS alone when TACoS shows it is lifting organic sales.
• Chasing a lower ACoS by cutting terms that were still profitable.
• Reading a launch product’s ACoS as if it were a mature product’s.
• Ignoring RTO, so cash-on-delivery campaigns look better than the money they bring in.

How we report ad returns each week

We are an operator, not an agency: we run brands of our own as well, so we read these ratios as a seller paying for the clicks. Client reports put ROAS, ACoS, TACoS and MER on one basis.

That basis is revenue net of GST, cancellations, returns and RTO, and each target is set per channel from your margin, with written rules for when spend is cut back. Our performance marketing services page sets out the method.

The campaigns themselves run under Google Ads and Meta Ads management, and sponsored ads inside the marketplaces sit alongside marketplace account management, so bids and listings are planned in the same weekly review.

The terms used here, from RTO to TCS, are defined in plain English in our ecommerce glossary.