What Does ROAS Mean on Amazon?

The Basic Definition

ROAS stands for return on ad spend. On Amazon it means the same thing it means everywhere else: take the revenue your ads generated, divide it by what you spent on those ads, and you get a ratio. A ROAS of 4 means every ad dollar brought back four dollars in sales.

Two things trip people up. First, ROAS is a ratio, not a percentage — when someone says "400% ROAS" they mean the same 4x, just written differently, and mixing the two formats inside one report is where confusion starts. Second, ROAS is revenue-in, revenue-out. It never touches your product cost, your referral fee, or your margin. A 4x ROAS on a product with thin margins can lose money; a 2x ROAS on a high-margin product can be very profitable. The number tells you about media efficiency, not profit.

On Amazon specifically, "revenue" in that formula comes from attributed sales — clicks, and for DSP sometimes views, that Amazon's system credits to your ad within a set window. That attribution model is the single biggest reason two people can look at the "same" campaign and get different ROAS numbers.

A Worked Example, With Real Numbers

Say you spent $2,000 on Sponsored Products last month and Amazon's reports show $10,000 in attributed sales from those ads. ROAS = $10,000 ÷ $2,000 = 5. You got $5 back for every $1 spent. That's the whole calculation — no adjustment for fees, no averaging across campaigns unless you choose to add spend and revenue together first.

The number changes shape once you're running at real scale, across many campaigns and ad types. Across 30 of our advertisers in July 2026, the portfolio delivered 6.04x return on ad spend — measured across the whole book, not cherry-picked from the best line item. That same book ran 78.4 million impressions at a $4.00 CPM and a blended $1.42 cost-per-click (the $0.41 CPC people quote in this category is online-video only, not the whole book), for a blended cost per acquisition of $5.49 across 57,137 attributed purchases.

Notice what a portfolio figure hides that a single-campaign figure shows: your best campaign might run at 12x while your worst runs at 1.5x, and both are true at once. If someone quotes you one ROAS number for "the account," ask whether it's a blend, and ask what's inside the blend before you decide anything is working or broken.

What Counts as a Good ROAS

There's no universal good number, and anyone who gives you one without asking about your margin is guessing. Amazon's own advertising guidance describes a ROAS around 2x as roughly average, with brands typically aiming for 3x to 4x — but that range assumes a margin structure that may not be yours.

The number that actually matters is your break-even ROAS: 1 divided by your margin percentage after Amazon's fees and cost of goods. If your margin after fees is 25%, you need at least a 4x ROAS just to break even on the ad, before it's contributed anything. A 5x ROAS on that product is good. The same 5x on a 50%-margin product is very good. A 3x ROAS that looks "below benchmark" might already be well above your break-even.

Category also moves the number — categories with high repeat-purchase rates can run a lower ROAS profitably because the first sale isn't the only one being paid for.

ROAS vs ACOS vs TACoS

ROAS, ACOS, and TACoS all describe the same relationship between spend and revenue, just from different angles — and using the wrong one for the question you're actually asking is a common, avoidable mistake.

• ROAS asks: what did I get back?
• ACOS asks: what percentage of ad revenue did I give up to get it?
• TACoS asks: how much of my whole business, not just ad sales, is being funded by ads?

ROAS and ACOS are mathematically the same information written two ways — ACOS is roughly the inverse of ROAS, expressed as a percentage. TACoS is the one people skip, and it's the one that catches a brand quietly becoming ad-dependent: TACoS can stay flat or worsen even while ROAS looks fine, because ROAS only ever looks at ad-attributed sales, never at what happened to organic sales alongside them.

When the ROAS Number Is Bad News

When the ROAS number is bad news, there are three different problems, and they need three different fixes.

The number itself is wrong. This usually means an attribution mismatch — you're comparing a 7-day-click Sponsored Products number against a DSP number that includes view-through credit, or you're summing campaign-level ROAS figures that double-count the same sale across sponsored ads and display. The fix is to reconcile in one place, at the shopper level, before arguing about the result.

The setting is already on. We've had clients ask us to "fix" a low ROAS on a campaign that was already running with tight, ROAS-favoring bids — the ceiling had already been reached, and the real lever was budget or targeting, not bidding.

The fix didn't work. This is a mistake we've made ourselves: cutting spend on a campaign with a low direct-attributed ROAS, because the dashboard said it wasn't earning its keep, only to see repeat-purchase revenue soften months later. Last-click and even view-through attribution can't tell you what a shopper would have done without seeing the ad — they can only tell you what happened after. That's a real limit, not a settings problem, and no amount of bid tweaking fixes it. Proving whether an ad actually changed behavior needs a holdout or a matched control, not another read of the same dashboard.

Where This Fits for reMKTR

reMKTR runs Amazon DSP as a managed service, inside the Full Circle group, which has managed more than $500M in Amazon spend across 100+ brands including HexClad, Ridge, BK Beauty, Beardbrand, Epic Gardening, The Woobles and Walkize. We hold real DSP seats — 109 live right now — and reconcile spend and sales in Amazon Marketing Cloud, because that's the only place display and sponsored ads stop double-counting each other. Whether or not you ever work with us: if a dashboard hands you one ROAS number that blends sponsored and DSP line items, ask how it handles the overlap before you trust it.