How to Calculate Marketing ROI (And Why Most Get It Wrong)

The naive formula (don't use this alone)

Naive ROI = Revenue from campaign / Cost of campaign

$50,000 revenue from $10,000 spend = 5× ROI

Three problems: revenue isn't profit, attribution may be wrong, and you haven't counted incremental sales.

The accurate formula

True Marketing ROI = (Incremental gross profit − Marketing cost) / Marketing cost × 100

Where:
  Incremental = sales that wouldn't have happened without the campaign
  Gross profit = revenue minus COGS (not revenue alone)
  Marketing cost = ad spend + agency fees + creative + tools

Why use gross profit, not revenue

A $50 sale at 40% gross margin only generates $20 of profit. If marketing cost was $10 per acquired customer, the naive ROI is 5× ($50/$10) — but the real ROI is just 1× because only $20 of that revenue is yours. Always use gross profit, not revenue.

The incrementality problem

Most attribution tools (Google Ads, Facebook, etc.) claim credit for sales they didn't cause. A customer searches your brand name on Google because they saw a billboard, clicks a paid ad, and converts — Google Ads attributes 100% of that sale to the click. In reality the billboard caused the search.

The gold standard test: run a holdout. Turn off the campaign in 20% of geographies (or for 20% of audience) for 4–8 weeks. Compare sales in the test vs control. The difference is the true incremental revenue.

A worked example: Facebook Ads campaign

Facebook reported 6× ROAS ($60k/$10k). True ROI was 68% — still positive, still worth doing, but a much more sober number.

Customer LTV vs first-order ROI

For businesses with repeat purchases or subscriptions, first-order ROI alone can mislead you the other way. A subscription business that loses $5 acquiring each customer at first order but makes $200/year for 3 years has a fantastic LTV-ROI even though first-order ROI is negative.

LTV-ROI = (Lifetime gross profit per customer − CAC) / CAC × 100

Where CAC = Customer Acquisition Cost (marketing cost ÷ new customers)

Time-adjusting ROI

A campaign that returns 30% in 3 months is annualised at over 200% — far better than a 30%-over-2-years campaign. Compare campaigns on annualised ROI when the time horizons differ.

Use the ROI Calculator — entering an investment period in months produces the annualised figure automatically.

Benchmarks

• Email marketing: often 30–40× ROI (very cheap, high engagement)
• SEO (steady state): 5–10× ROI but with 6–12 month lag
• Paid search: 2–5× ROAS typical, 1.5–3× after gross-margin adjustment
• Paid social: 1.5–4× ROAS, often 0.5–2× true ROI after incrementality
• Brand / out-of-home: not directly measurable; budget as a percentage of revenue

Where most teams go wrong

1. Reporting ROAS as ROI. ROAS uses revenue; ROI should use profit.
2. No holdout test. Attribution platforms over-report by 30–70%.
3. Ignoring brand traffic. Paid ads on your own brand name often steal credit from organic traffic.
4. Counting one-off costs only. Forget creative production, agency retainer, tools.
5. Optimising on last-click. Last-click attribution favours bottom-of-funnel ads but starves top-of-funnel that fed the funnel.

Bottom line

• True marketing ROI uses gross profit, not revenue; incremental sales, not attributed.
• Holdout tests are the gold standard for measuring incrementality.
• For repeat-purchase businesses, track LTV-ROI not just first-order.
• Annualised ROI lets you compare campaigns of different durations.