ACoS vs ROAS: two ways to see the same spend and when to use each

ACoS and ROAS are the same information seen from opposite sides: both compare how much you spent on ads against how much sales that spend produced. The only difference is which way you divide. ACoS (Advertising Cost of Sales) is spend ÷ sales, expressed as a percentage: “of every dollar sold through ads, how much went back into advertising.” ROAS (Return on Ad Spend) is sales ÷ spend, expressed as a multiple: “for every dollar I put into ads, how many dollars of sales did I get back.” They are exact mathematical inverses. If your ACoS is 25%, your ROAS is 4 (because 1 ÷ 0.25 = 4). If your ROAS is 5, your ACoS is 20%. They don’t measure different things: they measure the same thing on two opposite scales.

So which one should you use? The short rule: use ACoS when you’re thinking about profitability and spend limits —it’s easy to line up a 25% ACoS against your 30% margin and see you’re still ahead—; use ROAS when you’re thinking about return and scale, or when you’re reporting to someone from a marketing and performance background. Amazon Ads shows ACoS by default; a lot of digital advertising platforms and agencies speak in ROAS. Being able to translate between the two instantly, without a calculator, keeps two people from arguing over the same number while believing they’re talking about different things.

The real problem for a multichannel seller isn’t the formula: it’s that each marketplace hands you its own version under its own label. Amazon reports ACoS. Your Google Shopping or Meta campaign reports ROAS. MercadoLibre gives you Product Ads spend on its own logic. And you end up, once again, copying figures into a spreadsheet just to get them onto the same scale before you can compare channel against channel. By the time the table is built, the data is already yesterday’s and the bids kept running. This article gives you the mental conversion so ACoS and ROAS stop being two languages and become a single number you can read in either direction.

the same division, flipped

Let’s start concrete. Say you spent $1,000 on advertising and those ads generated $5,000 in attributed sales. The ACoS is 1,000 ÷ 5,000 = 0.20, or 20%: one fifth of that revenue went into ads. The ROAS is 5,000 ÷ 1,000 = 5: every dollar invested brought back five dollars of sales. Same spend, same sales, two numbers saying the same thing.

The relationship is fixed and simple: ACoS = 1 ÷ ROAS, and ROAS = 1 ÷ ACoS. A 10% ACoS equals a ROAS of 10. A 50% ACoS equals a ROAS of 2. A 100% ACoS equals a ROAS of 1 (you spent exactly what you sold). And watch the intuition: with ACoS, lower is better; with ROAS, higher is better. They run in opposite directions. A seller who drops ACoS from 30% to 20% is doing the same thing as one who lifts ROAS from 3.3 to 5. It’s the same move, told from both sides of the coin.

Glossary: ACoS is ad spend divided by the sales attributed to those ads; it reads as a percentage and lower is better.

when ACoS is more useful

ACoS shines when the conversation is about profitability. Because it’s a percentage of revenue, it compares directly against other percentages you already use: your gross margin, your category fee, your break-even. If you know that after fees, fulfillment, product cost, and tax you keep 30% margin on the sale price, then any ACoS below 30% leaves you profit and any ACoS above it eats that profit. That mental comparison —ACoS against margin— is instant, because both live in the same unit.

That’s why Amazon Ads reports in ACoS and why most sellers think in ACoS: the daily operating question is “is this campaign still profitable?”, and ACoS placed next to margin answers it with no translation. When you want to define the spend ceiling that still leaves profit, ACoS is the natural language, which is why it pays to calculate that limit carefully, as we cover in target ACoS: how to calculate the spend limit that still leaves profit. ACoS speaks the same language as your P&L.

when ROAS is more useful

ROAS wins when the conversation is about return and scale. A ROAS of 6 lands differently than “a 16.6% ACoS”: the multiple communicates “every dollar gives me six back” immediately, which is the frame anyone from performance marketing or anyone managing budget across several channels thinks in. If you’re deciding where to put the next dollar of investment —Amazon, Meta, Google Shopping— comparing ROAS across channels is more intuitive than comparing ACoS, because a higher ROAS shouts “this returns more” without your having to flip the logic in your head.

ROAS is also the language of agencies and of almost all advertising outside Amazon. If you work with someone optimizing your campaigns, or if you sell on your own channels alongside the marketplaces, you’ll hear ROAS constantly. The key is not getting lost in translation: when the agency celebrates a ROAS of 4 and you have a 28% margin, that ROAS of 4 equals a 25% ACoS, just under your break-even. The number that sounds great in marketing may be brushing the red in your accounting. Without putting them on the same scale, you never notice.

the detail almost nobody adjusts: gross vs net

Here’s the trap neither ACoS nor ROAS solves on its own: both are calculated, by default, on gross sales. Neither a 20% ACoS nor a ROAS of 5 knows what the product cost you, what you paid in fees, in fulfillment, or in tax. Both are measures of spend efficiency over revenue, not of profit. A ROAS of 5 can be fantastic on a 45%-margin product and a loss on a 15%-margin one, exactly like a 20% ACoS. Switching from ACoS to ROAS gives you no new information about profitability; it only changes the scale.

That’s why “net” versions exist: target ACoS or break-even ACoS, and their mirror the target ROAS, which are calculated against margin rather than gross sales. But for that you need to know your real net margin, SKU by SKU and channel by channel. And that’s where the multichannel seller gets stuck: margin lives in one catalog, the fee in another tab, the 3PL fulfillment in an email. Converting ACoS to ROAS is a few seconds of arithmetic; cross-referencing them against real margin to know whether you’re actually winning is what you normally rebuild by hand every week.

Glossary: real net margin is what’s left after ALL costs —product, fees, shipping, tax, and advertising—, not just price minus cost; it’s the benchmark you judge an ACoS or a ROAS against.

a conversion table to memorize

You don’t need a calculator to move between the two: just anchor a few pairs in your head. ACoS 10% = ROAS 10. ACoS 20% = ROAS 5. ACoS 25% = ROAS 4. ACoS 33% = ROAS 3. ACoS 50% = ROAS 2. ACoS 100% = ROAS 1. With those six reference points you can interpolate almost any real case. If someone says “let’s go with a ROAS of 4.5”, you already know it’s between 20% and 25% ACoS, around 22%.

The practical payoff is avoiding misunderstandings. In a meeting where one person reports ACoS and another reports ROAS, the instant mental conversion keeps a bad decision from getting made by comparing apples to oranges. And when you set dynamic pricing —raising in peak season, lowering to defend the Buy Box— ACoS and ROAS move together because both depend on the sale price, not just on spend. That’s why it pays to read advertising and pricing in one place; a price move scheduled in your automatic price calendar shifts the denominator of your ROAS and the implicit numerator of your margin at the same time.

one scale, in real time

The real value isn’t knowing the formula —you learned that in two paragraphs— but not having to translate by hand every time. When Amazon gives you ACoS, Meta gives you ROAS, and MercadoLibre gives you its spend loose, every cross-channel comparison starts with a stretch of normalizing in Excel. And because campaigns don’t wait, by the time the table is done the bids have changed, stock has moved, and conversion isn’t the one in your sheet.

A single source of truth solves exactly that friction: it shows you each campaign’s ACoS and ROAS in both scales at once, already cross-referenced against real net margin per SKU and per channel, updated today and not yesterday. So the question stops being “what unit was this number in?” and becomes “did this spend leave me a profit?”. With the conversion handled underneath, ACoS and ROAS stop being two reports in two languages and become a single reading of the same dollar invested —seen, as convenient, as a percentage or as a multiple— always against your real margin and availability.

Glossary: real available stock is sellable inventory net of reservations and in-transit; if it drops, your conversion falls and both your ACoS and your ROAS worsen without the campaign having changed.