How Do You Calculate SEO ROI?

How Do You Calculate ROI on SEO?

Calculate ROI on SEO by adding up everything the SEO work cost over a period, counting the customers organic search won beyond what it was already bringing, turning them into gross profit, and dividing the profit left after the cost by the cost. Why we use gross profit rather than revenue is on our page about how to measure digital marketing ROI; what’s particular to SEO is the baseline, the branded searches and the timing.

1. Choose two periods, the last twelve months and the whole time since the program began, because SEO’s costs come before its results and its pages keep working after they’re paid for.
2. Add up the full cost of the SEO work for each period: agency or consultant fees, writers, the developer hours spent on technical fixes, tools, and the time your own experts spend briefing writers and reviewing pages.
3. List the customers whose first contact came through organic search, from your CRM. Google Analytics counts a visit as Organic Search when it arrives from a non-ad link in organic results, including Google’s AI Overviews and AI Mode,¹ and the lead record should carry that source beside the buyer’s answer to “How did you hear about us?” and any calls your call tracking ties to organic visits.
4. Take out the customers who searched for your name, because they knew you before the search, as the next section explains.
5. Subtract the baseline, the number of non-branded organic customers you were winning in the same months before the SEO work started.
6. Turn the customers left into gross profit on the same basis you use for your other channels, first-year or lifetime, so SEO can sit beside them.
7. Subtract the cost, divide by the cost, and write down the period, the basis and what you counted, so next year’s figure is built the same way.

The Channel Selection step shows how to set up the CRM as the source of truth that steps 3 to 5 depend on.

How Do You Prove the ROI of SEO?

Prove the ROI of SEO by showing the gain in non-branded organic customers over a baseline you recorded before the work began, because organic search was bringing in customers before anyone was paid to improve it, and an SEO report that counts all of them claims credit for rankings the company already had.

Search Console’s branded filter splits searches that include your name from those that don’t,² but only as totals for the site, and Google Analytics can’t break that query data down by its own visitor-level dimensions,³ so no report tells you which search a particular customer typed. In the accounts we review, two clues settle most cases: the first page the customer landed on, since organic visitors who arrive on the home page or contact page have usually searched for the company, and what the buyer says at intake about whether they already knew your name.

The baseline needs the same split, taken from the period before the work started:

• Search Console’s history, exported now, because Search Console keeps only the last 16 months of data,³ and the branded filter has data only from its introduction on March 11, 2025.²
• The non-branded organic customers in your CRM for the four quarters before the program, judged with the same two clues.
• The same quarter a year earlier for each quarter you report, so a busy season isn’t mistaken for SEO.

Customers who first landed on pages the program created are the easiest to credit, since those pages brought nobody before, and where the stakes justify it, an incrementality test on one service line or region gives stronger proof than any before-and-after comparison.

How Long Does SEO Take to Show a Positive ROI?

SEO tends to show a negative ROI in its first months, because the cost is paid before the pages rank and Google says some changes take several months to show in search results,⁴ so judge it on cumulative cost against the cumulative gross profit of the customers it added, and call the point where the second passes the first its payback.

Your own ceiling depends on how long customers stay, how early their profit arrives and how much cash you can tie up, which the Channel Selection page shows how to work out. Which shows results more quickly, SEO or search advertising explains why we don’t put a fixed number of months on SEO, and how long SEO takes gives the checkpoints to read before the customers arrive.

What Is a Good ROI for SEO?

A good ROI for SEO is one where the cost per added customer, over the life of the program, sits below the most you can afford to spend to win a customer and compares well with what paid search costs on the same searches, and no published benchmark is reliable enough to set that target for you.

The SEO ROI figures most often repeated online come from First Page Sage, an SEO agency, which reports three-year averages from its own clients’ campaigns run between early 2021 and late 2025, ranging from 317% for ecommerce to 1,389% for real estate.⁵ It describes content-led campaigns with a typical fee of about $120,000 a year and doesn’t publish how many campaigns it averaged or how it decided which revenue SEO produced, so we’d read them as one agency’s report on its own clients rather than a target.⁵ We sell SEO ourselves, so weigh our view with that in mind.

Two comparisons tell you more than a benchmark:

• Against your ceiling, the most you can afford to spend to win a customer, which what is a good marketing ROI applies to every channel.
• Against paid search on the same searches, cost per customer from your own records on both sides, since in accounts we manage paid search shows which keywords produce revenue and SEO earns its keep by winning customers on those terms without paying for every click.

When Does an SEO ROI Report Mislead?

An SEO ROI report misleads when it values traffic at what the clicks would have cost in ads, credits SEO with customers who searched for your name, counts a recovery as growth, or is judged before the pages have had time to work, and each of these makes the return look surer than it is.

• Traffic value presented as return. Ahrefs defines organic traffic value as what the traffic from a site’s rankings would cost “if that traffic was paid via PPC instead,” calculated from its own estimates of each keyword’s traffic and cost per click.⁶ The figure can’t show how many of those visitors became customers, and you might never have paid for those clicks at all.
• Branded searches counted as SEO. A buyer who searched for your name after a referral, an event or a sales call would have found you with almost any website, which is why the brand bidding page treats branded paid clicks with the same suspicion.
• A recovery counted as growth. We’ve seen redesigns that didn’t map old pages to new addresses lose search traffic, and when an agency restores it, the baseline is the level before the drop, not the low point after it.
• Buyers who never clicked. A buyer who read your answer in an AI Overview and phoned the next day shows up as a branded search or a direct call, which what is zero-click search covers, and the intake question is the record most likely to catch it.
• A verdict after one quarter. Rankings, impressions and non-branded clicks are the signs to track while you wait, as how to choose an SEO agency describes, but the return is measured in customers.