The Ultimate Guide to ROAS in 2026

Share

Return on ad spend is the revenue your ads produce divided by what you spent on them. Spend $1,000, attribute $3,000 in sales, and your ROAS is 3.0x. It is the most quoted number in ecommerce advertising and one of the easiest to misread.

What a typical ROAS looks like by platform

These are median figures from the BenchMarketing dataset. The middle half of accounts sits inside the range shown; the rest are above or below it.

Amazon is reported differently: the source we use publishes an average rather than a median, at 3.26x. See Amazon Ads benchmarks for the detail.

Two patterns stand out. Channels that catch people while they are shopping (Shopping, Amazon) return more per dollar than channels that interrupt people while they are scrolling. And the spread inside each platform is wide: a Meta account at 1.5x and one at 4.5x can both be run competently, for products with very different margins.

The ROAS you need is set by your margin, not by a benchmark

A benchmark tells you what other accounts achieve. It cannot tell you whether a campaign makes you money. For that you need your break-even ROAS:

Anything above break-even contributes profit before overheads; anything below loses money on every sale unless the customer comes back. Work it out once per product line with the break-even ROAS calculator, then use the benchmark to judge whether hitting it is realistic on a given channel.

Why the ROAS in your ad account is optimistic

Each platform counts the sales it can connect to its own ads, inside its own attribution window. Meta's default is a 7-day click and 1-day view window; Google Ads credits conversions using its own model and conversion window. When a customer sees a Meta ad, searches your brand on Google and buys, both platforms can claim the sale. Add the platforms' ROAS figures together and you will usually count more revenue than you actually made.

Three habits keep the number honest:

• Compare platform ROAS with a blended figure: total revenue divided by total ad spend over the same period. That is your MER, and it cannot double count.
• Keep the attribution window the same when you compare periods or channels.
• Report prospecting and retargeting separately. Retargeting audiences already intended to buy, so their ROAS flatters the whole account.

When a lower ROAS is the right call

A campaign at 1.8x can be worth more than one at 5x. Brand search and retargeting post high ROAS because they reach people who were close to buying anyway; cutting their budget often loses little revenue. Prospecting campaigns post lower ROAS because they find new customers, and a new customer who buys three times is worth more than the first order suggests. If you know your customer lifetime value, judge prospecting against it rather than against first-order ROAS. Our guide to lifetime value models covers how.

How to improve ROAS without starving growth

1. Fix tracking first. Missing purchase events or duplicate tags move ROAS more than any optimisation.
2. Split campaigns by margin. A product at 60% margin can carry a lower ROAS target than one at 20%.
3. Cut placements and audiences that never convert before you cut budget overall.
4. Raise average order value with bundles and thresholds; ROAS rises without any change in media cost.
5. Refresh creative on a schedule. On social platforms ROAS usually decays as the same people see the same ads.

Check where your own number sits with the ROAS calculator, or compare it against your industry on the ROAS benchmark page.

See where you stand

Compare your own numbers with these benchmarks for your industry.

Related pages

• ROAS benchmarks by channel and industry
• Break-even ROAS calculator
• MER benchmarks

More ecommerce guides

• Pinterest Ads for Ecommerce | ROAS Benchmarks and Creative Best Practices 3 min read
• Ecommerce Conversion Rate Benchmarks by Industry 3 min read
• Optimizing Product Feeds for Google Shopping 3 min read