ROI vs ROAS(ROI)

Long definition

The two metrics measure related but distinct things, and conflating them in reporting is a common cause of misallocated budget.

• ROAS = Revenue ÷ Ad Spend. Output is a multiple: 4.5x means $4.50 revenue for every $1 of ad spend. Common in paid acquisition, especially performance marketing. Ignores cost of goods sold, fulfillment, refunds, and ongoing service costs. A 4x ROAS on a 30%-margin product is unprofitable.
• ROI = (Gain - Cost) ÷ Cost. Output is a percentage: 200% means you got back twice what you spent, net of cost. Includes all costs you choose to count. The right number for budget allocation decisions.

For SEO, ROAS is rarely the right framing because there is no per-click ad spend. SEO costs are program-level: salaries, agency retainer, content production, tooling, technical implementation. The standard SEO-ROI framework:

1. Estimate organic revenue contribution (last-click, attribution-modeled, or incrementality-based).
2. Estimate program costs over the same window: people-cost (salaries × time allocation), tooling, content production, dev time for technical fixes.
3. ROI = (Revenue contribution × margin - Program cost) ÷ Program cost.

Three honest adjustments most SEO ROI calculations skip:

• Compounding asset value. A blog post ranking for two years contributes revenue across both. ROI on a single-month basis understates this; allocate cost over the asset's productive lifetime, or report ROI on a 12-month rolling basis.
• Brand and assisted contribution. Last-click ROI undercounts SEO. Use data-driven attribution or model an assisted-credit multiplier (typically 1.2-1.5x last-click revenue based on incrementality studies).
• Counterfactual cost. What would the same revenue have cost via paid acquisition? When organic ROI is reported alongside the avoided-paid-spend equivalent, the case for SEO investment becomes legible to non-SEO stakeholders.

When boards ask for "marketing ROI" they usually mean ROAS-like simplicity. Provide both numbers with definitions, not one number that obscures the math.

Common misconceptions

• "ROAS and ROI are the same metric with different names." They use different denominators (ad spend vs total cost) and different numerators (revenue vs profit). A 5x ROAS can correspond to negative ROI if margins are tight.
• "SEO has no cost so ROI is infinite." SEO has people cost, content cost, and opportunity cost. The "free traffic" framing is misleading. Treat the program's fully loaded cost as the ROI denominator.
• "You should optimize SEO for the highest ROI." Highest-ROI SEO often comes from already-ranking pages. Focusing only on these starves the pipeline of new content, and ROI collapses 12-18 months later when those pages decay. ROI is necessary but not sufficient for content portfolio decisions.
• "Organic ROI is computable to two decimals." It rests on attribution assumptions that are 20-40% noise. Report ROI as a directional range, and decisions become better — chasing a precise number papers over uncertainty in the underlying model.