ROAS (return on ad spend): what it measures and how to calculate it

ROAS is the ad metric most brands quote first. It tells you how much revenue your ads were credited with for each dollar spent. It does not tell you whether that revenue made money.

What is the ROAS formula?

ROAS = ad-attributed revenue ÷ ad spend

Amazon Ads calculates it as ad revenue ÷ ad spend. Its worked example: $50 of spend that brings in $100 of ad revenue is a ROAS of 2.

The same result appears in three formats:

• Ratio: 2.0, as Amazon Ads shows it. Shopify writes the same result as 2:1.
• Percentage: 200%. Google Ads states target ROAS this way, so a goal of $5 of conversion value per $1 spent is 500%.
• Inverse: a 50% ACOS, Amazon's advertising cost of sales.

How do ROAS and ACOS relate?

They are the same two inputs, flipped. Amazon Ads calls ROAS the inverse of ACOS. Divide 1 by ACOS as a decimal to get ROAS, or 1 by ROAS to get ACOS.

ROAS	ACOS	Pays for itself on attributed sales only if margin before ads is above
1.0	100%	100% (not possible)
2.0	50%	50%
2.5	40%	40%
2.8	35.7%	35.7%
3.0	33.3%	33.3%
4.0	25%	25%
5.0	20%	20%

TACOS is the wider view: the same ad spend measured against total sales, organic included.

What is a good ROAS?

Amazon Ads says there is no definitive good ACOS, and the same holds for its inverse. Shopify ties an acceptable ROAS to your margins, operating costs and what the ads are meant to achieve. What every SKU has is a break-even point.

Amazon Ads says ACOS has to stay below profit margin for a campaign to remain profitable. Flip that, with margin measured before ad spend, and you get:

Break-even ROAS = 1 ÷ margin before ad spend

• A $30 product that clears $9 before advertising has a 30% margin. Break-even ROAS is 1 ÷ 0.30, about 3.33.
• A $30 product that clears $12 has a 40% margin. Break-even ROAS is 2.5.

So a 2.8 ROAS loses money on the first product and makes money on the second. Whether 2.8 is good depends on margin after marketplace fees, fulfillment, landed cost and returns. That margin-based target is what margin ACOS expresses, and contribution margin is where its input comes from.

Some SKUs run below break-even on purpose, to launch a product or win new customers. That should be a decision with an end date, not a default.

What are the common ROAS mistakes?

1. One target for the whole catalog. A single "3x" goal overfunds thin-margin SKUs and starves high-margin ones.
2. Margin from the wrong layer. Price minus product cost ignores fees, storage and returns, so break-even ROAS looks lower than it really is.
3. Treating attributed revenue as caused revenue. Platforms assign credit with an attribution model, and Amazon Ads notes that a last-touch model credits only the final touchpoint before a sale. Some of those buyers, especially on branded searches, may have bought anyway.
4. Comparing channels at face value. Attribution rules differ by platform, so the same 4.0 on two channels can credit very different sets of sales.
5. Ignoring stock. A high-ROAS SKU with two weeks of cover is where scaling hurts most. Extra spend pulls sales forward and can run it out before the next shipment lands.

Which ops gap does ROAS connect to?

Inventory-aware ad spend. ROAS is computed from ad data alone and carries no stock signal, so a SKU can post a strong ROAS right up to the day it runs out. The fix is to judge each SKU's ROAS against its own break-even and its days of cover together.

Ecomsellertool Growth OS is built to run this on your own accounts. Reporting gives a live P&L per SKU, the input for each SKU's break-even ROAS. Inventory and replenishment gives days of cover per SKU and shipment plans. Margin-aware advertising moves budget toward deep-stock SKUs and away from products about to stock out. On Amazon, bid and budget changes can run through the Amazon Ads API, which Amazon says supports automated bid and budget optimization with no additional API fee. This builds on the margin-aware ad metrics and agents we built for Amazify. After handover, your data and the custom modules we build are yours; the Growth OS base is licensed to you.

Where should you start?

If one ROAS target covers your whole catalog, or ad spend keeps flowing to SKUs that are about to run out, start with the free Quick Scan in the Ops Gap Diagnostic. It works from five standard Amazon reports you upload, including your Sponsored Products advertised product report, and returns your top three operations gaps, each with a yearly dollar estimate. Ad spend on SKUs whose days of cover has fallen below an agreed minimum is one of the gaps it checks. The Quick Scan covers Amazon; the paid Deep Diagnostic covers every channel you sell on.