What is ROAS?
Return on Ad Spend
Definition, formula, India benchmarks, and the channel-vs-blended distinction that decides if you're profitable.
ROAS, or Return on Ad Spend, is the revenue generated per unit of advertising spend. It is calculated by dividing the total revenue attributable to ads by the total ad spend over the same period. ROAS is most useful as a tactical channel metric; for blended profitability, use Contribution Margin or LTV/CAC instead.
ROAS = Revenue from ads ÷ Ad spend (often shown as a multiple, e.g. 4×).
Channel ROAS overstates efficiency; track blended ROAS for honesty.
"Good" ROAS is category-dependent — D2C beauty 2.5–4.5×, real estate 5–25×.
ROAS equals the total revenue attributable to ads, divided by the ad spend over the same period. It is typically expressed as a multiple (e.g., 4×) or a percentage (e.g., 400%).
ROAS = Revenue from Ads ÷ Ad SpendROAS vs ROI — the trap
ROAS is revenue-based; ROI is profit-based. A hero 4× ROAS at 25% contribution margin is only 1× ROI — every rupee of ad spend produces ₹1 of profit. Most "we hit 5× ROAS!" posts on LinkedIn are silent about gross margin. Always pair ROAS with contribution margin before celebrating.
Channel ROAS vs blended ROAS
Channel ROAS is what Meta, Google, and Performance Max report. Each platform claims credit for whatever conversion it can attach a pixel to — including conversions a customer would have made organically. Blended ROAS is total revenue divided by total paid spend, regardless of channel. Channel ROAS typically inflates 25–60% above blended.
India benchmarks (2026)
- D2C beauty (channel): 2.5–4.5× · (blended): 2.0–3.5×
- D2C fashion (channel): 1.8–3.2×
- Real estate (channel): 5–25× — high AOV, low purchase frequency
- F&B / restaurant (channel): 3–6×
- SaaS (B2B): not directly applicable — use CAC payback
Source: Frameleads aggregated client data plus Shopify Plus India 2025 outlook. Bands compress in saturated CPM regimes.
Common mistakes
- Optimising to channel ROAS while blended ROAS drifts down.
- Ignoring iOS 17 attribution drift (Meta over-reports by 8–22%).
- Confusing ROAS and ROI in board decks — never do this.
- Chasing high ROAS with narrow audiences that cap scale.
Worked example — Mumbai D2C beauty brand
A Mumbai D2C beauty brand spent ₹14L on Meta in October 2025; Meta-attributed revenue was ₹49L.
Channel ROAS = 3.5×.
But blended (after iOS attribution drift adjustment, organic-purchase deduction): 2.6×. With a contribution margin of 32%, net profit per ₹1 of ad spend is ₹0.32 × 2.6 = ₹0.83 — contribution-positive but tight.
Conclusion: lift AOV, lower COGS, or shift channel mix toward organic and email — don't double down on Meta until the blended number improves.
Frequently asked questions
What's a good ROAS?
It depends on category, gross margin, and stage. As channel ROAS, healthy bands in India 2026: D2C beauty 2.5–4.5×, D2C fashion 1.8–3.2×, real estate 5–25×, F&B 3–6×. As blended ROAS, expect 60–80% of channel ROAS once you account for organic-purchase deduction and iOS attribution drift.
Is ROAS the same as ROI?
No. ROAS is gross-revenue-based — it ignores margin, refunds, and overheads. ROI is profit-based and requires gross margin and overhead inputs. A 4× ROAS at 25% gross margin yields just 1× ROI on the ad spend.
Why is my channel ROAS higher than my blended ROAS?
Channel reports (Meta, Google) claim credit for assisted conversions and over-attribute. Blended ROAS divides total revenue by total spend across all paid channels and is honest about what acquisition is actually doing. Always operate to blended.
How is ROAS used differently for D2C vs B2B?
D2C uses ROAS as a primary tactical signal because the purchase is immediate and attributable. B2B SaaS rarely uses ROAS — sales cycles are 45–120 days; instead use CAC payback period and pipeline-sourced revenue.
Pair this with
Want this audited for your business?
Get a free 30-minute marketing audit from a senior Frameleads consultant. No slides — just an honest read of where your funnel leaks, with the three highest-leverage moves we'd make first.