Definition
Return on Ad Spend (ROAS) is a measure of campaign efficiency calculated as revenue attributable to advertising divided by advertising cost — typically expressed as a multiple (e.g. 4x means $4 in revenue per $1 spent).
ROAS is the cleanest single number for paid advertising performance, but only if the attribution is honest. Last-click ROAS dramatically overcounts the channel that gets the credit (usually search) and undercounts every channel that contributed to consideration earlier in the journey. Modern attribution models (data-driven, time-decay, marketing mix modelling) try to correct for this — none perfectly.
ROAS without margin context can mislead. A 5× ROAS on a 60%-margin SaaS product is hugely profitable; the same ROAS on a 15%-margin physical product loses money on every sale. Margin-adjusted ROAS (Profit on Ad Spend, POAS) is the honest version for non-software businesses.
Origin
Direct-response heritage from mail-order and infomercial advertising. Standardised in digital marketing through the 2000s with the rise of conversion tracking; now the default top-line KPI on most ad platforms.
How it works
- Define attribution window and model (last-click, data-driven, etc.).
- Track revenue from converted users back to the campaigns that touched them.
- ROAS = revenue / ad spend, computed per campaign / ad set / creative.
- Compare to break-even ROAS (1 / margin) — anything above is profitable on a unit basis.
- Optimise spend toward higher-ROAS audiences and creatives; pause or rework underperformers.
When to use it
Use when
- Every direct-response campaign with measurable revenue.
- As the headline KPI for paid retainers.
- Per-channel and per-creative — granular ROAS reveals where to redirect spend.
Skip when
- Brand-awareness campaigns. ROAS for those is misleading; CPM and brand-lift are the right KPIs.
- Without margin context. Top-line ROAS misses the cost of goods.
Key metrics
- ROAS itself.
- Margin-adjusted ROAS (POAS).
- Break-even ROAS for the business.
- ROAS by channel, audience, creative.
Examples
- The campaign hit 6.2x ROAS in month two and we doubled the budget.
- ROAS without margin context can mislead — sometimes break-even ROAS is fine, sometimes 5x isn't enough.
- We capped spend on the 3.2× ROAS audience and scaled the 8.4× one.
In practice at Makreate
Makreate Advertising campaigns are reported on ROAS at the campaign, ad set, and creative level — so spend follows what works. On a recent ecommerce engagement we found two creative concepts driving 8× ROAS while the broader campaign averaged 3.4×; we shifted 70% of spend to those concepts within two weeks and the account-level ROAS climbed from 3.4× to 5.6× without changing the budget.
Advertising →Common mistakes
- Targeting a flat ROAS goal across products with different margins.
- Trusting last-click attribution. It overcredits search and undercredits everything upstream.
- Ignoring the LTV adjustment. First-purchase ROAS understates campaigns that bring high-LTV customers.
- Optimising creative-level ROAS too aggressively before reaching statistical significance.
- Confusing ROAS with profit. ROAS doesn't account for COGS, fulfilment, refunds, or operations cost.
Frequently asked
What's a good ROAS?
Whatever exceeds your break-even — 1 / contribution margin. SaaS with 80% margins can be profitable at 1.5×; ecommerce with 25% margin needs 4×+.
ROAS or POAS?
POAS (Profit on Ad Spend) is more honest for goods-businesses. ROAS is fine for high-margin or subscription where margin is roughly constant.
How do I improve ROAS?
Better creative, tighter audiences, better landing page conversion, post-purchase upsell, and pruning the underperformers. Order matters — start where the data is most stable.