- Report organic revenue and contribution, not rankings and traffic.
- Separate branded demand (which marketing creates) from non-branded demand (which SEO wins).
- Use the paid-search equivalent and incremental margin to value the channel, and track forecasts against actuals.
Why SEO ROI is so often misreported
Most SEO reports answer the wrong question. They show rankings, traffic and tasks completed, the inputs, rather than the business outcome. CEOs then face an unhelpful choice: trust the channel on faith, or cut it.
A better model treats organic search like any other investment: what did we put in, what incremental value came out, and what's the expected return on the next dollar?
The five numbers to ask for
- Organic revenue and gross profit, attributed with a consistent, documented model.
- Non-branded share: revenue from searchers who weren't already looking for you. This is where SEO creates new demand capture.
- Paid-search equivalent value: what the same non-branded clicks would cost in paid search. A useful, conservative proxy for the asset's value.
- Investment: people, agencies, tools and engineering time allocated to SEO.
- Forecast vs. actual: the plan the team committed to, and how they performed against it.
A simple ROI framework
| Component | How to calculate |
|---|---|
| Incremental organic gross profit | Non-branded organic revenue × gross margin, versus baseline trend |
| Total SEO investment | Team + agencies + tools + allocated engineering and content cost |
| ROI | (Incremental gross profit − investment) ÷ investment |
| Payback period | Months until cumulative incremental gross profit exceeds cumulative investment |
Because organic assets keep producing after the investment is made, measure ROI over 24–36 months, not a single quarter. Most of the value arrives in years two and three.
Proving causality
Attribution shows correlation. To build confidence that SEO work causes results, mature teams use SEO split tests (changing a template on a random subset of similar pages and comparing against a control), staged rollouts, and pre-registered forecasts for major initiatives.
The KPI tree for the board
Boards don't need 40 metrics. They need a tree:
- Outcome: organic revenue / contribution and share of new customers
- Drivers: non-branded traffic, conversion rate, share of search versus competitors, AI visibility
- Health: indexation of priority page types, technical health score, risk indicators
- Initiatives: the 3–5 bets underway, each with expected impact and status
If your SEO report wouldn't survive your CFO's first question, it isn't a report. It's a status update.
Questions to ask your team this quarter
- What is our non-branded organic revenue, and how has it trended over three years?
- What did we forecast last year, and how did we do?
- What are the three initiatives that will move the number most, and what do they need from other teams?
- How exposed are we to AI answers, and what are we doing about it?
- What are we doing that we should stop?
Frequently asked questions
What is a good ROI for SEO?
It varies widely by business model and maturity. Because organic assets compound, well-run programs often show returns that improve significantly in years two and three; evaluate over 24–36 months.
How long does SEO take to show ROI?
Early indicators appear within months; payback for substantial programs typically occurs within 6–18 months depending on starting authority, competition and execution speed.
Should branded search count toward SEO ROI?
Report it, but separate it. Branded demand is largely created by brand and product; non-branded demand is where SEO creates incremental value.