"Is our ROAS any good?" is the question every founder asks their marketing lead. The honest answer: it depends on what you sell, where you sell it, and which platform you're spending on. A 2.5:1 ROAS is a celebration for a B2B SaaS company and a slow-motion bankruptcy for an e-commerce brand.
Here's what the numbers actually look like in 2026, and how to read them.
First, what ROAS doesn't tell you
ROAS — Return On Ad Spend — is a vanity metric in isolation. The formula is dead simple:
ROAS = Revenue Attributable to Ads ÷ Ad Spend
But "revenue attributable" hides a war. Is that last-click revenue? First-touch? Multi-touch? Including or excluding returns? Gross or net of platform fees? Two agencies reporting "5x ROAS" on the same account can mean wildly different things.
Before benchmarking, lock down three things:
- Attribution window. Google Ads defaults to 30 days post-click. Meta defaults to 7 days. LinkedIn defaults to 30 days. Inconsistent windows → inflated ROAS on whichever platform claims credit first.
- Net vs. gross. Returns, refunds, chargebacks, and platform fees eat 10–25% of "revenue" before it hits your bank. Real ROAS uses net.
- Marginal vs. blended. Blended ROAS divides total revenue by total ad spend. Marginal ROAS is what each additional dollar earns. Marginal is what matters for scaling decisions.
Now the numbers.
E-Commerce ROAS benchmarks (B2C)
E-commerce is the most ROAS-sensitive category because the unit economics are tight and the customer journey is short. Anything below 2.5:1 on a paid channel usually means you're losing money once you factor in COGS, fulfilment, and customer service.
| Sub-vertical | Channel | "Good" ROAS | "Excellent" ROAS |
|---|---|---|---|
| Fashion & apparel (UK/UAE) | Google Shopping | 3.5–4.5x | 6x+ |
| Beauty & cosmetics | Meta Ads | 2.5–3.5x | 5x+ |
| Home goods | Google Shopping | 4.0–5.0x | 7x+ |
| Luxury retail (Dubai) | Meta + Google | 4.0–5.5x | 8x+ |
| Subscription DTC | Meta (cold) | 1.8–2.5x | 3.5x+ |
| Marketplace sellers | Meta + Google | 3.0–4.0x | 5.5x+ |
A few things stand out. Luxury retail in the UAE outperforms apparel by 30–40% because the average order value masks the higher CAC. We saw a Dubai watch retailer push 8.2x blended ROAS in their first 6 months — but their AOV was AED 4,200 and their cold-acquisition CAC was AED 380. Reframe the same ratio for a £45 t-shirt brand and the math falls apart.
Subscription DTC looks weak on initial ROAS, but first-order ROAS isn't the right metric there. LTV:CAC ratios above 3:1 are healthy even at sub-2x initial ROAS, because the customer pays off over months 2 through 12.
B2B / Lead-gen ROAS benchmarks
B2B is harder to benchmark because the value of a lead depends on close rate, deal size, and sales cycle. The right metric isn't ROAS — it's cost per qualified lead (CPQL) and revenue per lead acquired (RPL).
That said, when clients insist on ROAS reporting:
| Sub-vertical | Channel | "Good" first-touch ROAS | "Excellent" ROAS |
|---|---|---|---|
| B2B SaaS (mid-market) | Google Search | 2.0–3.0x | 4.5x+ |
| Professional services | LinkedIn Ads | 1.5–2.5x | 3.5x+ |
| Enterprise sales | LinkedIn + Google | 1.2–2.0x | 3.0x+ |
| Local services (UK trades) | Google Search | 4.0–6.0x | 8x+ |
| Coaching & courses | Meta + Google | 2.0–3.5x | 5x+ |
The interesting line is local services — a UK plumbing or roofing business running Google Search can hit 6–8x ROAS routinely because the intent signal is razor-sharp ("emergency plumber London") and AOV is high (£600+ per job). If your local-services account is below 4x, the problem is almost always negative keywords or geo-targeting.
Channel-specific notes
Google Search Ads — Highest intent. Should run the highest ROAS of any paid channel. If your search ROAS is below 3x and you're not selling commodities at razor-thin margins, your match types or keyword selection is wrong.
Google Shopping / Performance Max — Watch the asset-group breakdown. Google bundles search, display, YouTube, and Discover into Performance Max and reports a single number. The blended ROAS hides which placement is actually working. Ask for the asset-group performance report monthly.
Meta Ads — Cold-audience ROAS dropped meaningfully post-iOS 14. Conversion API + server-side tracking recovers about 70% of the visibility loss. If you're not running CAPI in 2026, your reported ROAS is wrong by 15–30% in either direction.
LinkedIn Ads — Stop chasing direct-click ROAS. LinkedIn's strength is influence over time for B2B. Plan for 90-day attribution windows and pipeline-velocity metrics, not 7-day click-through revenue.
Region matters more than people admit
We work across UK, UAE, and India, and the same campaign architecture produces very different ROAS curves region to region.
- UK: higher CPCs (especially in finance, legal, insurance — £8–25 per click on Google Search), but conversion rates and AOVs match. ROAS expectations: ~4x for e-commerce, ~2.5x first-touch for B2B.
- UAE: lower CPCs in most verticals, but elevated CAC for the luxury segment because the audience is small and brand-conscious. ROAS expectations: 4.5–6x for luxury e-commerce, 3x for general retail.
- India: dramatically lower CPCs (often ₹4–₹40 per click), but lower AOVs and aggressive competition on price. ROAS expectations: 5–8x for D2C e-commerce, 2–3x for SaaS.
Don't import US benchmarks blindly. A "good" Meta ROAS in California ad agency reports (typically 2.0–2.5x for DTC) is a bad one in the UAE, where ad inventory is cheaper and trust signals more important.
The diagnostic checklist when ROAS is below benchmark
If your number is under the "good" column for your category, run through this before making any changes:
- Attribution model. Switch from last-click to data-driven attribution in Google Ads. Most accounts see ROAS look 10–25% better just from this change because previously-undercredited assist clicks get the right weight.
- Negative keywords. Pull the Search Terms report. If 15%+ of spend is on terms that don't match buyer intent, that's where the leak is.
- Landing page conversion rate. If your account-level conversion rate is below 3% on intent-rich keywords, the problem isn't the ads — it's what happens after the click. Page load speed under 2 seconds is non-negotiable.
- Ad creative refresh cadence. Meta ads decay fast — typical creative starts losing CTR within 3 weeks. If you haven't refreshed creative in 30+ days, performance decay alone could explain a 20–30% ROAS drop.
- Match type audit. Broad match in Google has improved with smart bidding, but on accounts under £3K/month it still leaks money to irrelevant queries. Phrase match is the safer default for smaller budgets.
Stop optimising for ROAS alone
Here's the contrarian point. ROAS is a constraint, not an objective. The actual objective is profit per period.
A campaign at 4x ROAS spending £1,000 generates £4,000 in revenue. A campaign at 2.5x ROAS spending £10,000 generates £25,000. The second one is mathematically "worse" on the ratio but generates 6× more profit at the same gross margin.
The right question isn't "what's a good ROAS?" — it's:
- What's my marginal ROAS at this spend level?
- How much can I scale before marginal ROAS drops below my break-even threshold?
Most accounts hit a clear inflection where marginal ROAS falls off a cliff. That inflection — not the headline number — is where you stop scaling. The number we typically see is 5–8x current spend before marginal returns soften meaningfully, but it depends on audience size and channel.
Want to know where yours sits?
We run free 24-hour audits on Google, Meta, and LinkedIn accounts. We pull six months of data, benchmark you against your category, and identify the specific leaks — usually 3–6 things, ranked by how much money each is costing you per month.
Black Arrow Technologies builds and runs performance ad campaigns for businesses across the UK, UAE, and India. We've reduced CAC by 42% for Dubai real estate, hit 3.2x ROAS for UK fintech, and put £4 of return on every £1 spent for luxury retail. We don't launch a single campaign without conversion tracking in place.