SEO ROI: How to Calculate It, With Examples

SEO ROI (return on investment) is the value your SEO work produced, minus what it cost, divided by what it cost. The formula is simple: ROI = (value from SEO − cost of SEO) ÷ cost of SEO × 100. The hard parts are deciding what “value from SEO” includes, over what time window, and how to count AI answers that influence buyers without sending a click.

This guide is for owners, marketing leads and finance people who need a defensible SEO ROI number. It walks through the formula, a worked example with clearly hypothetical numbers, attribution windows in GA4, and a practical way to value AI visibility. What SEO costs in the first place is covered in SEO pricing.

On this page

• The SEO ROI formula
• What counts as value from SEO?
• A worked SEO ROI example
• Attribution windows and why they change the answer
• How do you value AI citations that never click?
• How long before SEO pays back?
• Common SEO ROI mistakes

The SEO ROI formula

The standard SEO ROI formula divides net return by cost. Semrush’s guide to SEO ROI gives it as (revenue from SEO − cost of SEO) ÷ cost of SEO, multiplied by 100 for a percentage, with the example of $200,000 in revenue on $40,000 of cost producing a 400% ROI. We recommend two refinements before you use it with a finance team.

Refinement 1: count incremental value only. If organic search already produced 20 leads a month before the work began, those 20 are not a return on the new spend. Subtract the baseline, or at least show it.

Refinement 2: use gross profit, not revenue. A 400% ROI on revenue can be a loss once you subtract the cost of delivering what you sold. Finance teams compare investments on profit, so give them profit.

On the cost side, include everything: agency or freelancer fees, internal staff time at a loaded rate, content production, tools, and developer hours spent on SEO fixes. Leaving internal time out is the most common way SEO ROI gets inflated.

What counts as value from SEO?

Value from SEO is any business result you can trace, with reasonable confidence, to organic search or AI answers. For ecommerce that is mostly orders. For lead generation it is qualified leads turned into customers. Semrush suggests valuing a lead as customer lifetime value multiplied by the lead-to-customer conversion rate: if a customer is worth $20,000 over their lifetime and 15% of leads become customers, each lead is worth $3,000.

Decide three things before calculating:

1. The conversion events that matter. Purchases, qualified form fills, booked calls. Newsletter sign-ups and page views are not value on their own.
2. Revenue window. First purchase only, twelve months, or lifetime value. Lifetime value flatters SEO; first purchase undercounts it. Pick one and say which.
3. The source of truth. GA4 key events, your CRM, or your order system. CRM data beats analytics for lead businesses because it knows which leads became customers.

For ecommerce the data is usually cleaner, because orders carry revenue and GA4 can attribute them. Margin still varies by product, so pull gross profit from your order system rather than applying one average margin to all organic revenue. For lead businesses the weak link is the handoff from form to CRM. If leads arrive by phone, use call tracking numbers on the site, or you will undercount organic by however many people prefer to call.

Before any of this, make sure the inputs are trustworthy. Search Console should be verified for the whole domain, GA4 key events should fire once per real conversion, and the CRM should record the original source. A short monthly report that shows those inputs next to the ROI figure makes the number much easier to defend; the structure is in the SEO report template.

A worked SEO ROI example

The example below uses hypothetical numbers for a service business, chosen to show how the refinements change the answer. They are not benchmarks.

Now run the formula three ways:

• Naive (all organic leads, revenue-style): 360 leads × 15% × $2,500 = $135,000. ROI = ($135,000 − $42,000) ÷ $42,000 = 221%.
• Incremental, first-year profit: ($45,000 − $42,000) ÷ $42,000 = 7%.
• Incremental, three-year customer value (if the average customer stays three years at $2,000 gross profit a year, so $6,000 each): 18 × $6,000 = $108,000. ROI = ($108,000 − $42,000) ÷ $42,000 = 157%.

Same business, same year, three defensible-sounding numbers from 7% to 221%. The honest report shows the incremental first-year figure as the headline and the lifetime figure as context, and explains both. A reader who sees only “221%” is being misled, even if unintentionally.

Attribution windows and why they change the answer

Attribution decides which channel gets credit for a conversion, and the lookback window decides how far back a touchpoint can be credited. SEO touches often happen early, when someone is researching, so short windows tend to undercount it.

In GA4, according to Google’s help page on attribution settings, the lookback window for acquisition key events (first visit and first open) can be 7 or 30 days, with 30 as the default, and for all other key events it can be 30, 60 or 90 days, with 90 as the default. Google also notes that the first click, linear, time decay and position-based models are no longer available as of November 2023, which leaves data-driven attribution and last click.

Practical settings for SEO ROI:

• Keep the 90-day lookback for key events unless your sales cycle is very short.
• Compare data-driven and last-click reports side by side. If organic’s share drops sharply under last click, SEO is doing early-journey work that last click hides.
• For B2B or long sales cycles, rely on CRM first-touch and opportunity source fields, not GA4 alone. GA4’s windows are shorter than many sales cycles.

How do you value AI citations that never click?

You value AI visibility that does not click by combining what you can measure with what you can reasonably estimate, and by keeping the estimate separate from the core ROI number. Google says in its documentation on AI features in Search that clicks from results pages with AI Overviews tend to be higher quality, with users more likely to spend more time on the site, but many AI answers produce no click at all. The click loss itself is covered in AI Overviews and the zero-click problem.

Search Console’s generative AI performance report shows impressions for AI Overviews and AI Mode by page, country, device and date, though Google notes it is still rolling out. It does not give revenue, so it belongs in the evidence, not the ROI arithmetic. Setting up the GA4 side is covered in tracking AI traffic in GA4 and Search Console, and the metrics worth reporting are in the AI visibility metrics that matter.

A simple, honest way to put a number on it: take conversions where the customer named an AI assistant as their source, value them the same way as other conversions, and report that as “self-reported AI-influenced value” next to the core ROI. It will undercount, because many people will not remember or say. That is better than an inflated model nobody can check.

How long before SEO pays back?

SEO usually pays back over quarters, not weeks. Semrush’s guide quotes Google’s Maile Ohye: in most cases, SEOs need four months to a year to implement improvements and then see potential benefit. Costs are front-loaded while returns build, so ROI measured at month three is often negative and at month twelve positive for the same program.

Two habits keep this from becoming an excuse:

• Agree leading indicators up front. Indexed pages, impressions for target queries, rankings for commercial terms, and AI visibility on a fixed prompt set should move before revenue does. The ones worth tracking are in SEO KPIs.
• Report cumulative ROI and trailing twelve-month ROI. Cumulative ROI shows whether the program has paid back yet. Trailing twelve months shows whether it is still earning.

Common SEO ROI mistakes

Most inflated or deflated SEO ROI figures come from a handful of mistakes:

• Counting the baseline as return. Organic traffic you already had is not a result of new work.
• Revenue instead of profit. Makes SEO look better than paid channels reported on margin, or worse, depending on who is presenting.
• Leaving out internal time. The cheapest SEO is the one where staff time is ignored.
• Branded search as SEO success. People searching your name found you because of everything else you do. Separate branded and non-branded.
• Short attribution windows for long buying cycles.
• No time window stated. “SEO ROI is 300%” means nothing without “over what period.”
• Ignoring AI answers entirely, or the opposite, claiming every brand search as AI-driven.

Making SEO ROI a number you can defend

SEO ROI is worth calculating carefully because it decides budgets. Use incremental gross profit, include every cost, state the window, keep attribution settings consistent, and report AI visibility as measured evidence plus a clearly labeled estimate. Done that way, the number survives a finance review, and it tells you which parts of the program to fund next.

If you want help setting up the measurement for both organic and AI search, and a program built around the numbers it produces, that is what our AI SEO service does.