The average e-commerce brand spends 12-15% of revenue on advertising (Statista, 2025). Whether that spend is an investment or a waste depends almost entirely on one number: ROAS. Return on Ad Spend tells you how many dollars you earn for every dollar you put into advertising. A 4x ROAS means $4 back for every $1 spent. Simple math, but the way brands misuse it causes real problems.
ROAS is the default performance metric across Google Ads, Meta Ads, TikTok Ads, and most paid media platforms. It's the number your agency puts in the monthly report, the number your CFO asks about, and the number that determines whether your ad budget grows or gets cut next quarter.
How ROAS Is Calculated
ROAS = Revenue from Ads / Ad Spend
If you spend $10,000 on Google Ads and those ads generate $50,000 in revenue, your ROAS is 5.0x (or 500%). Some platforms express ROAS as a ratio (5:1), others as a percentage (500%), others as a multiple (5x). They all mean the same thing.
Important: ROAS measures revenue, not profit. A 5x ROAS sounds strong, but if your profit margin is 15%, you're only netting $7,500 on that $50,000 in revenue. Subtract the $10,000 in ad spend and you've actually lost $2,500. This is why ROAS alone can be misleading.
ROAS vs. ACoS vs. CPA: Which Metric to Use
Different platforms and teams prefer different metrics. Here's how they compare:
| Metric | Formula | Expressed As | Used By | What It Tells You |
|---|---|---|---|---|
| ROAS | Ad Revenue / Ad Spend | Multiple (e.g., 4x) or percentage (400%) | Google, Meta, TikTok, agencies | Revenue efficiency of ad spend |
| ACoS | (Ad Spend / Ad Revenue) x 100 | Percentage (e.g., 25%) | Amazon sellers | Cost efficiency (inverse of ROAS) |
| CPA | Ad Spend / Number of Conversions | Dollar amount (e.g., $45) | Lead gen, SaaS, service businesses | Cost per individual conversion |
| TACoS | (Ad Spend / Total Revenue) x 100 | Percentage (e.g., 10%) | Amazon brands (strategic view) | Ad dependency relative to total business |
Quick conversions: ROAS of 4x = ACoS of 25%. ROAS of 2x = ACoS of 50%. The formula is ROAS = 1 / (ACoS / 100).
ROAS Benchmarks by Platform and Industry
What counts as "good" ROAS depends heavily on your platform, industry, and business model. These benchmarks come from aggregated agency data (WordStream, Databox, and platform reports from 2024-2025):
| Platform / Channel | Average ROAS | Strong ROAS | Notes |
|---|---|---|---|
| Google Search | 2x-4x | 6x+ | High-intent traffic. Branded terms inflate averages. |
| Google PMax | 3x-5x | 7x+ | Varies widely by product feed quality. |
| Meta (Facebook/Instagram) | 2x-3x | 5x+ | Signal loss from ATT makes tracking harder. Actual ROAS often higher than reported. |
| Amazon Sponsored Products | 3x-5x | 7x+ | First-party purchase data makes attribution reliable. |
| Amazon DSP | 2x-4x | 5x+ | Upper-funnel. Harder to attribute directly. |
| TikTok Ads | 1.5x-3x | 4x+ | Younger platform. Strong for awareness, weaker for direct response (improving). |
Industry matters too. Apparel and fashion brands often see 3-5x ROAS. Electronics and consumer tech run closer to 2-3x due to thinner margins and higher competition. Subscription businesses can accept lower initial ROAS (even 1-2x) because lifetime value justifies higher acquisition costs.
Why ROAS Alone Is Not Enough
ROAS has three blind spots that trip up even experienced advertisers:
1. It ignores profit margins
A 4x ROAS on a product with 60% margins is very profitable. A 4x ROAS on a product with 20% margins means you're losing money after accounting for COGS, shipping, and platform fees. Always calculate your break-even ROAS: 1 / profit margin. If your margin is 40%, your break-even ROAS is 2.5x.
2. It doesn't account for attribution problems
Platform-reported ROAS is based on each platform's attribution model. Google claims credit for a sale if the user clicked a Google ad within the attribution window. Meta claims credit for the same sale if the user also saw a Meta ad. Both platforms report the full revenue. Your actual blended ROAS is lower than any single platform reports. Use a tool like Triple Whale, Northbeam, or Rockerbox for cross-platform attribution.
3. It conflates new and returning customers
Running retargeting ads to existing customers will produce high ROAS (8x, 10x, sometimes 20x). But many of those customers would have purchased anyway. Prospecting campaigns that acquire genuinely new customers will have lower ROAS (2-3x) but drive actual growth. Blending both into one ROAS number hides this distinction.
How to Improve ROAS
- Tighten targeting. Broad audiences waste spend on low-intent users. Use first-party data, lookalike audiences built from your best customers, and in-market signals to focus on high-conversion-probability users.
- Improve your landing page. A 1% increase in conversion rate can improve ROAS by 25-40%. Test page speed, product imagery, social proof placement, and checkout flow.
- Test more creative. On Meta and TikTok, creative is the biggest ROAS lever. Ads fatigue in 7-14 days. Brands that refresh creative weekly consistently outperform those that run the same ads for months.
- Segment by intent. Set different ROAS targets for prospecting (lower target, 2-3x) versus retargeting (higher target, 5x+). Holding prospecting to the same ROAS standard as retargeting kills growth.
- Fix your product feed. For Shopping and PMax campaigns, your product feed quality directly determines which searches trigger your ads. Complete titles, accurate categories, and high-quality images improve impression quality and click-through rate.
Example: Benchmarking ROAS Across Platforms
A fitness equipment brand running $80K/month across Meta, Google, and Amazon tracked ROAS separately by platform and found stark differences: Google Shopping delivered 5.2x ROAS, Amazon Sponsored Products delivered 3.8x, and Meta prospecting campaigns delivered 1.9x. The industry benchmarks align with this pattern — TrueProfit 2025 data shows Google Shopping averages 5.0-6.5x ROAS while Meta's median is 2.2x overall. But when they ran incrementality tests, Meta's true contribution was higher than platform-reported ROAS suggested, because many Meta-exposed users searched on Google before purchasing. The lesson: ROAS benchmarks are useful starting points, but cross-platform incrementality testing reveals the real picture.
Frequently Asked Questions
What ROAS should I target?
Start with your break-even ROAS (1 / profit margin). If your profit margin is 50%, your break-even is 2x. Then add your growth and profit goals on top. Most e-commerce brands target 3-5x blended ROAS, but the right target depends entirely on your margins, customer lifetime value, and growth stage.
Why is my platform-reported ROAS different from actual results?
Every platform over-reports because they each claim credit for overlapping conversions. A customer might click a Google ad, see a Meta retargeting ad, then purchase. Both platforms count the full sale. Cross-platform attribution tools solve this, but as a rule of thumb, expect your true blended ROAS to be 20-40% lower than individual platform reports suggest.
Is a 10x ROAS always better than a 3x ROAS?
Not necessarily. A 10x ROAS usually means you're spending conservatively, targeting only the easiest-to-convert audiences (brand searches, warm retargeting). You're likely missing a much larger pool of potential customers who would convert at 3-4x ROAS. The goal is maximum profitable revenue, not maximum ROAS. Spending $100K at 3x ROAS ($300K revenue) often beats spending $20K at 10x ROAS ($200K revenue).
Read next:
- Amazon ACoS for the Amazon-specific version of ad efficiency metrics
- TACoS for measuring ad spend against total revenue (not just ad-attributed)
- AI-Powered Advertising for how machine learning optimizes ROAS automatically