What's a Good ROAS by Industry? (2026 Benchmark Data)

There's no single "good ROAS" number — return on ad spend depends heavily on platform and vertical. Across 18,000 ecommerce brands, median ROAS runs 1.93x on Meta versus 3.68x on Google Ads (Triple Whale, 2025 data) — nearly double, on the same underlying business. Importing a single benchmark from a different platform or category will misjudge whether your campaigns are actually working.

This post covers what a good ROAS looks like by platform and ecommerce vertical, what we could and couldn't verify for fintech and subscription/app businesses, and the mistakes that come from benchmarking against the wrong number. For the underlying regional CPC and CAC data, see our CAC and CPC benchmarks for Southeast Asia and India.

Key Takeaways

• Median ecommerce ROAS is 1.93x on Meta versus 3.68x on Google Ads, across 18,000 brands — comparing return on ad spend across platforms without accounting for this gap will misread which one is actually underperforming (Triple Whale, 2025).
• Google Ads ROAS varies by vertical from 2.12x (Health & Wellness) to 4.30x (Travel & Luggage).
• There's no verified, disclosed-methodology ROAS benchmark specifically for fintech or subscription/app businesses — we checked and couldn't find one. Revenue-per-install data is the best available proxy, and we show how to build a target from your own unit economics below.
• Most tracked verticals declined year-over-year — a benchmark from even a year ago may already be stale.

Why "Good ROAS" Depends on Platform and Vertical

In 2026, the honest answer is that ROAS reflects the underlying auction and customer intent on each platform, not a universal efficiency score. Google Ads captures more high-intent, bottom-funnel search traffic — someone typing "buy running shoes size 10" is closer to a purchase than someone scrolling past a discovery ad — which is a large part of why median Google Ads ROAS (3.68x) runs nearly double Meta's (1.93x) across the same 18,000-brand sample (Triple Whale, 2025). Neither number says one platform is "better." They measure different jobs: Google mostly harvests demand that already exists, while Meta creates and captures demand earlier in the funnel, where a sale attributed today often started with an impression weeks ago. That's why brands running both platforms successfully almost never hold them to the same target, and why a blended cross-platform ROAS is close to useless as a diagnostic.

Practical implication: if you're running both platforms, don't hold Meta and Google to the same ROAS target. A 2x ROAS might be strong for Meta-driven awareness/consideration traffic and disappointing for Google Search's higher-intent demand — judge each platform against its own realistic range, not a blended number.

What's a Good ROAS for Ecommerce?

It depends heavily on category. Google Ads ROAS ranges from 2.12x for Health & Wellness to 4.30x for Travel & Luggage — nearly double, within the same platform (Triple Whale, 2025 data).

Vertical	Google Ads ROAS (2025)	Year-over-year change
Travel & Luggage	4.30x	−21.10%
Consumer Electronics	3.02x	declined
Pets & Animals	2.84x	+2.51% (only improving vertical)
Health & Wellness	2.12x	−15.64%
Ecommerce median (all verticals, Google Ads)	3.68x	declined
Ecommerce median (all verticals, Meta)	1.93x	declined

Table data: Triple Whale, Google Ads & Facebook Ads Benchmarks, 2025 data, retrieved 2026-07-09.

Worth noting: Pets & Animals was the only tracked vertical where ROAS improved year-over-year (+2.51%) in the Triple Whale dataset, while most other verticals declined — Travel & Luggage fell 21.10%, Health & Wellness fell 15.64%. A vertical-average ROAS from even 12 months ago may no longer reflect current auction conditions in that category.

Common mistake: treating a vertical average as a fixed target rather than a moving baseline. Re-check your category's benchmark at least annually, because in a dataset where most verticals declined double digits year-over-year, the direction of change matters as much as the number itself.

What's a Good ROAS for Fintech or Subscription/App Businesses?

Here's the honest gap: we could not find a credible, disclosed-methodology ROAS benchmark specifically for fintech or subscription/app businesses, the way Triple Whale publishes for ecommerce. We looked. It's a genuine data gap in the industry, not something we're papering over with an imported number.

What we do have is proxies. Revenue per install at Day 14 in India/SEA sits at $0.06-$0.09, roughly a fifth of the $0.39 median in North America (RevenueCat, State of Subscription Apps 2025) — that's monetization efficiency, not ROAS directly, but it's real and dated. A Kantar study commissioned by Meta found Meta ads delivered 1.8x ROAS in Thailand, the Philippines, Indonesia, and Vietnam (July 2024), though it excludes Singapore and India and, being Meta-commissioned, carries an obvious incentive to favor Meta's own platform.

Practical implication: for fintech and app/subscription businesses, build your ROAS target from your own unit economics — CAC payback period, LTV, and revenue-per-install — rather than importing an ecommerce benchmark that doesn't reflect your monetization model.

How to Calculate Your Own ROAS Target Instead

A benchmark tells you what the market looks like; your break-even ROAS tells you what you need, and it takes one line of arithmetic: divide 1 by your contribution margin. If you keep 40% of each sale after cost of goods, shipping, and payment fees, your break-even ROAS is 1 ÷ 0.40 = 2.5x — anything below that loses money on first purchase regardless of what any vertical average says. Add a profit cushion on top (most operators target 20-50% above break-even), and that's your working target.

Run that arithmetic against the benchmark table above and something useful falls out. A Health & Wellness brand with typical supplement-style margins of 60-70% breaks even at just 1.4-1.7x, so the vertical's "low" 2.12x average ROAS is actually comfortably profitable. An electronics retailer on 15-20% margins needs 5.0-6.7x just to break even — meaning the Consumer Electronics average of 3.02x implies the median advertiser in that vertical loses money on first purchase, presumably betting on warranty attach rates, accessories, or repeat purchases to recover it. The verticals with the highest ROAS averages are not automatically the healthiest ones; the margin structure behind the number decides that.

Two adjustments matter for subscription and app businesses. First, replace "revenue per sale" with revenue over your payback window (Day 30, 60, or 90 — whatever your cash position tolerates), since first-transaction revenue understates what a subscriber is worth. Second, sanity-check the result against a regional revenue-per-install figure like RevenueCat's: if your target implies each install must generate 4x more Day-14 revenue than the regional median, your plan needs either an exceptional product or a different channel mix.

Does Your ROAS Benchmark Account for How Campaigns Are Actually Structured?

Increasingly, no — because more ecommerce budget runs through automated campaign types that target their own ROAS number, not a manually-set bid. A study of 4,000+ Performance Max campaigns across 500+ advertiser accounts found PMax median ROAS targets rose to a 4.7x-6.0x range, with actual campaign performance achieving 95-116% of that target — meaning advertisers are increasingly setting the target and letting the algorithm hit close to it, rather than discovering what ROAS a campaign naturally produces (Smarter Ecommerce, State of Performance Max 2025, authored by Mike Ryan, April 2025).

That changes what a "benchmark" even means. If most advertisers in a category are running PMax with a self-selected target ROAS, an average ROAS across that category partly reflects what advertisers asked for, not a market-clearing efficiency number. A vertical average of 3x might mean the auction genuinely supports 3x, or it might mean most advertisers in that category simply set their target at 3x and the algorithm mostly delivered it.

So when you benchmark against a vertical average, ask what's actually behind it. A number produced by open auction dynamics tells you what the market supports; a number produced by aggregated target-setting mostly tells you what other advertisers asked for — and only the first one says anything about what happens if you push harder.

Common Mistakes When Comparing Against Benchmarks

1. Blending ROAS across platforms into one number. A blended 2.5x tells you nothing about whether Meta or Google is actually underperforming — measure each platform against its own realistic range.
2. Importing a benchmark from a different vertical. A 3x ROAS is mediocre in Travel & Luggage and strong in Health & Wellness, based on the same dataset — category context changes what "good" means.
3. Treating ROAS as the only efficiency metric that matters. ROAS doesn't account for margin, so a lower-ROAS channel selling higher-margin products can be more profitable than a higher-ROAS channel selling thin-margin ones. Pair ROAS with contribution margin before reallocating budget.
4. Using a stale benchmark. With most tracked ecommerce verticals showing ROAS declining year-over-year, a number from a year-old report may already overstate what's currently achievable.

Frequently Asked Questions

What is a good ROAS for ecommerce?

It depends on platform and vertical: median ecommerce ROAS runs 1.93x on Meta versus 3.68x on Google Ads across 18,000 brands, with vertical-level Google Ads ROAS ranging from 2.12x (Health & Wellness) to 4.30x (Travel & Luggage).

Is a 2x ROAS good or bad?

It depends entirely on platform and category. A 2x ROAS is below the ecommerce median for Google Ads (3.68x) but close to the Meta median (1.93x) — the same number means something different depending on which platform and vertical you're comparing it to.

What's a good ROAS for a fintech or subscription app business?

There's no verified industry-wide benchmark for this — we checked. Build your target from your own unit economics instead: CAC payback period, LTV, and revenue-per-install data, rather than importing an ecommerce number that doesn't reflect a subscription monetization model.

Why did my ROAS drop even though nothing changed in my campaigns?

Most tracked ecommerce verticals saw ROAS decline year-over-year in 2025 data, likely reflecting broader auction competition rather than anything specific to your account. Compare against a current benchmark, not one from a year ago.

Should I use the same ROAS target for Meta and Google Ads?

No. Median ROAS differs by roughly 2x between the platforms in ecommerce data, reflecting different customer intent at the point of ad exposure. Set platform-specific targets rather than one blended number.

Does Performance Max change what a "good ROAS" benchmark means?

Yes, partially. Since PMax campaigns increasingly target a self-selected ROAS number and the algorithm delivers close to it (95-116% of target in one 4,000+ campaign study), a vertical average may reflect what advertisers asked for as much as what the auction naturally supports. Treat category averages as a starting reference, not proof of what's achievable.

How do I calculate my break-even ROAS?

Divide 1 by your contribution margin. A brand keeping 40% of each sale after cost of goods, shipping, and payment fees breaks even at 1 ÷ 0.40 = 2.5x ROAS. Set your working target 20-50% above that break-even figure, and for subscription businesses, use revenue over your payback window rather than first-transaction revenue.

Conclusion

"Good ROAS" is a category- and platform-specific number, not a universal target. Ecommerce benchmarks vary nearly 2x between Meta and Google Ads and by a similar margin across verticals — and once you run the break-even arithmetic against your own margins, a "low" vertical average can turn out to be comfortably profitable while a "high" one quietly isn't. For fintech or subscription businesses, no credible industry benchmark exists at all, so build the target from your own unit economics instead.

How these benchmarks were compiled: every figure above traces to a named, dated source — Triple Whale's 18,000-brand dataset, RevenueCat's State of Subscription Apps 2025, Smarter Ecommerce's State of Performance Max study, and the Meta-commissioned Kantar study (caveats stated inline) — each verified directly against the source as of July 9, 2026. Where no credible benchmark exists, as with fintech and subscription ROAS, we've said so rather than substituting a plausible-looking number. Coact is a performance marketing agency working with ecommerce and app businesses across Singapore, India, and Indonesia.

Continue Learning

• Real CAC Benchmarks for Southeast Asia & India (2026) — the regional CPC and revenue-per-install data underlying several figures in this post
• The Complete Guide to Performance Marketing in 2026 (publishing soon)
• Performance marketing agency in India
• Talk to our team about building a ROAS target grounded in your own unit economics.