What is a good ROAS?

The only benchmark that means anything is yours

Break-even ROAS is 1 divided by gross margin. That is the whole formula, and it is the number that decides whether a campaign lives.

• 20% margin: break even at 5.0x
• 25% margin: break even at 4.0x
• 30% margin: break even at 3.33x
• 40% margin: break even at 2.5x
• 50% margin: break even at 2.0x
• 60% margin: break even at 1.67x
• 70% margin: break even at 1.43x

Gross margin here means contribution margin: revenue after cost of goods, shipping, payment processing, fulfilment and expected returns. Most people compute it on cost of goods alone. That makes break-even look lower than it is, which makes marginal campaigns look profitable, which is how a quarter disappears. There is a break-even ROAS calculator if you want the arithmetic done for you.

Why industry benchmarks are close to worthless here

Margin structure varies more inside a category than between categories. A supplement brand at 85% margin and a hardware reseller at 15% are both ecommerce, and a published "ecommerce average" describes neither of them.

Published benchmarks are also almost always platform-reported ROAS, which double-counts across channels and includes view-through conversions. Averaging numbers that are individually inflated does not produce a useful target. We are not going to invent one, and you should be suspicious of pages that do.

Platform ROAS is not the same number as profit

• It is self-reported by the platform selling you the ads.
• It usually includes view-through conversions, so a scroll-past can be credited as a sale.
• It overlaps with every other channel claiming the same order. See why Meta reports more revenue than Shopify.
• It rarely accounts for discounts, refunds or cancellations.
• It does not separate new customers from returning ones. A 6x built on customers who would have repurchased anyway is not 6x of growth.

What to measure instead

Use break-even ROAS as the floor and treat it as a kill line rather than a target. Use blended performance, total ad spend against total store revenue, to answer whether the business is actually working. Blended cannot be double-counted, because there is only one denominator.

Then use contribution margin per order to decide how hard to push, and reach for incrementality testing when the spend is large enough to justify the effort. Definitions for all of these are in the glossary, and Muze reads the spend side straight from Meta, Google and Amazon while pulling revenue from Shopify, so the blended number comes from one place instead of four exports.

Related questions

Is 3x a good ROAS?

Only if your break-even is below 3x. At a 33% contribution margin, 3x is exactly break-even, which means the campaign is doing precisely nothing for the business.

What is a good ROAS for ecommerce?

There is no single number, and any page giving you one is guessing. Work out 1 divided by your contribution margin and use that as your floor.

Should I use ROAS or MER?

Both, for different jobs. Per-campaign ROAS is for deciding what to scale or cut inside a platform. Blended MER, total spend against total revenue, is for deciding whether the business is working.

Does ROAS include shipping and tax?

It depends what value your purchase event sends to the platform. If it sends the order total including tax and shipping, your reported ROAS is inflated relative to the revenue you actually keep. Check the event before you trust the ratio.

What is break-even ROAS?

The ROAS at which a campaign covers its own cost and contributes nothing. It equals 1 divided by your gross contribution margin.

Keep reading

Other answers

See every answer