Key takeaways
- SEO ROI is revenue produced by organic search minus what it cost, divided by that cost, expressed as a percentage.
- The formula is trivial and the inputs are not, which is why almost every disputed return figure is an argument about what was counted rather than about the arithmetic.
- Branded search should be excluded from the revenue side, because people typing your company name would mostly have arrived anyway and counting them turns existing demand into a claimed result.
- Valuing organic clicks at what the same clicks would cost in paid search gives a defensible floor when revenue tracking is incomplete, and it makes the durability difference visible, since paid clicks stop the day the budget does.
- A return figure is only useful if the method behind it stays fixed, because a number that moves when the attribution rules change teaches a finance team to distrust the whole programme.
SEO ROI is the revenue organic search produced, minus what you spent to produce it, divided by that spend. The arithmetic takes one line. The difficulty is entirely on the input side: valuing traffic that does not convert on the first visit, and attributing revenue to search without quietly borrowing it from other channels. Get the inputs right and the figure survives a finance review. Get them wrong and it survives only a slide.
A search engine optimization company that reports against revenue rather than positions makes the honest version of this easier to assemble, mostly because the baseline and the attribution rules get set before the work starts rather than afterwards.
The SEO ROI formula, and the two inputs that decide whether it is real
The calculation: revenue from organic search, minus the cost of the work, divided by the cost of the work, multiplied by one hundred.
Two inputs decide whether the result means anything. The cost side has to include everything — retainers or salaries, content production, tooling, and the developer hours spent implementing fixes. Leaving out internal time is the most common way a return figure gets inflated, because it is the cost nobody invoices.
The revenue side has to be traceable to organic search specifically. That means excluding branded queries, excluding traffic that would have arrived through another channel anyway, and being explicit about which conversion actions you are counting. Understating the cost or overstating the revenue produces the kind of number that looks impressive in a review and collapses the moment somebody asks how it was built.
How to value organic traffic when the sale is not immediate
Most businesses do not close on a first visit, so last-click revenue alone will undercount search badly. Two methods hold up.
Revenue attribution is the stronger one where tracking is clean: follow organic visitors through to conversions and to the value of those conversions in analytics and the CRM. It answers the question directly rather than by proxy.
The paid-equivalent method is the fallback, and it is more useful than it sounds. Value the organic clicks at what the same clicks would cost you in paid search for the same queries. It gives a floor rather than a ceiling, it is easy to defend, and it lands well with anyone who already buys ads, because it is denominated in a currency they use daily.
The paid comparison also makes the durability difference visible in a way a percentage does not. A paid click stops arriving the moment the budget pauses. A page that ranks keeps producing the same clicks in its second year at no additional media cost, which is the entire reason the two channels have different return shapes rather than different return sizes. Our SEO vs PPC comparison works through that tradeoff in full.
Attribution without flattering yourself
Four rules keep the revenue side honest, and each of them pushes the number down rather than up.
- Separate branded from non-branded organic. Judge the work on the terms you did not previously rank for. Counting people who searched your company name credits the programme with demand it did not create.
- Use assisted conversions rather than last click alone. Search often opens a relationship that email or direct closes, and a last-click model hands the whole credit to whoever was standing closest at the end.
- Set a baseline before the work starts. Record organic traffic, leads and revenue for the period beforehand. The return is the lift above that baseline, not the total that follows it.
- Compare year over year rather than month over month wherever you have the history, so a seasonal pattern does not get read as a result.
The target is a figure you would put in front of a finance team unprompted. If the spend side of the equation is the part you are unsure about, SEO pricing covers it.
The payback curve, and why quarter one tells you nothing
The return does not arrive evenly. It arrives late and then accelerates, because the assets built early keep producing after the spend that created them has stopped.
| Period | What is happening | Shape of the return |
|---|---|---|
| Opening quarter | Technical remediation, research, first pages published | Entirely investment, no return yet |
| Middle of year one | Positions climbing on the less competitive terms, traffic building | Approaching the point where the two sides meet |
| Late year one | Content maturing, links compounding, leads growing | Turning positive as earlier work starts producing |
| Year two onward | Existing assets keep earning against lower new spend | Strongest, because you are harvesting rather than planting |
The exact timing depends on your starting authority, the competitiveness of the category, and how much content and link work the budget supports, which is why borrowed timelines are worth less than your own baseline. What is reliably true is the shape: judging the programme in its first quarter is judging an orchard in the month it was planted. how long SEO takes covers the timeline question on its own terms.
A worked example, start to finish
The figures below are illustrative and invented for the purpose of showing the mechanics. Substitute your own; the point is the order of operations, not the numbers.
Take a hypothetical B2B services company. Assume an average annual contract value of $90,000, a qualification rate of half on inbound leads, and a close rate of one in five on qualified leads. Assume it spends $4,000 a month, so $48,000 across the year, covering delivery, content and tools.
- Organic leads across the year: 60, ramping from a couple a month early to roughly eight a month by the end.
- Qualified at half: 30.
- Closed at one in five: 6 new customers.
- Revenue: 6 multiplied by $90,000, so $540,000.
- Return: $540,000 minus $48,000, divided by $48,000.
Then discount it deliberately. Assume only half of those customers are clearly search-driven and the rest would have found the company another way. Revenue drops to $270,000 and the return roughly halves. It is still comfortably positive, and it is now a figure the sales team would recognise rather than dispute, which is worth more than the larger version.
The second-year point is the one worth keeping. If the technical foundation and most of the content already exist, the spend can fall while the leads continue, so the return improves without any new investment. That is the mechanic behind the whole channel, and it is invisible in any single-year calculation.
The six mistakes that inflate the number
The same errors recur, and all of them push in the flattering direction.
- Counting branded search. If people already knew the name, ranking for it is not new revenue.
- Omitting internal time, content production and tooling from the cost side.
- Judging at the end of quarter one, when the curve has barely started.
- Using last click only, which buries every assist.
- Treating traffic as revenue. A page can triple its visits and produce nothing if the intent behind those visits is wrong.
- Having no baseline, which leaves you able to assert a lift but not to show one.
Avoid those and the figure becomes something you can act on, which is the only reason to calculate it. For the metrics that explain movement over time rather than summarising it annually, our guide to SEO KPIs is the companion to this.
Report it the same way every month
Pick the attribution rules, the baseline and the branded split once, then hold them fixed. A figure that moves because the method changed is worse than no figure, because it trains everyone reading it to discount the next one.
Write the definitions down where the whole team can see them: what counts as a conversion, which queries count as branded, what is included in cost, and which comparison period is used. Then produce the number on the same cadence with the same definitions, and record what changed each time. Consistency is what converts a return figure from a talking point into an input to a budget decision.
Related terms
You may see this topic described with related searches like how to measure roi on seo, how to measure seo roi, roi seo, seo roi calculator, and seo roi forecast. Those phrases are useful when they clarify what the reader needs next, but they should still point back to one clear plan.
Frequently asked questions
How is the return on organic search different from the return on paid search?
Paid search is easy to calculate and resets to zero the moment spending stops, because the traffic is rented for the duration of the campaign. Organic search is harder to attribute and compounds instead, because pages and links keep producing after the work that created them is finished. The practical consequence is that the two channels should be compared across a multi-year horizon rather than a single quarter, since a one-quarter window shows paid at its best and organic at its worst.
Can you measure the return on organic search without perfect tracking?
Yes, using the paid-equivalent method. Value your organic clicks at what the same clicks would cost in paid search for the same queries, which gives a defensible floor rather than a precise figure. Pair it with a clean baseline of traffic, leads and revenue captured before the work started, and you can produce something credible while you fix the conversion tracking in parallel.
Should branded search be counted as a search win?
Generally no. People typing your company name have already decided to find you, so counting that traffic credits search with demand created elsewhere, usually by sales, word of mouth or another channel. Split branded from non-branded in Search Console and judge the work on the non-branded side. The exception is when branded volume itself is rising as a documented result of the content, in which case report the two separately rather than merging them.
Is paying someone for SEO worth it?
SEO is worth it when your buyers search before they buy and the site can turn qualified visits into leads or sales. It is not worth funding blindly; the work should be tied to priority pages, ranking movement, and revenue.
Is SEO worth it for small business?
SEO is worth it when your buyers search before they buy and the site can turn qualified visits into leads or sales. It is not worth funding blindly; the work should be tied to priority pages, ranking movement, and revenue.
Is SEO worth it for small businesses?
SEO is worth it when your buyers search before they buy and the site can turn qualified visits into leads or sales. It is not worth funding blindly; the work should be tied to priority pages, ranking movement, and revenue.
Is Google SEO worth it?
SEO is worth it when your buyers search before they buy and the site can turn qualified visits into leads or sales. It is not worth funding blindly; the work should be tied to priority pages, ranking movement, and revenue.