How to Measure SEO ROI Without Lying to Yourself

To measure SEO ROI, total every cost of your search work, convert organic conversions into gross profit rather than revenue, then subtract cost from profit and divide by cost. The formula is simple. The difficulty is sourcing four honest inputs: close rate, gross margin, tracked conversions, and a fixed measurement window.

TL;DR

• The formula is gross profit from organic search, minus total SEO cost, divided by total SEO cost, times 100
• Total cost includes internal hours, tooling, and content production, not only the agency invoice
• Using revenue instead of gross profit turns a real 43.5 percent into a reported 219 percent on identical data
• Close rate and gross margin come from your CRM and your accounting. No agency can supply them, and borrowed figures make everything downstream decorative
• Exclude branded search before you calculate, or you are counting demand that already existed
• There is no trustworthy industry benchmark. The widely quoted 748 percent average comes from an SEO agency’s own client dataset

One-liner: Run the number on your own last four quarters before you judge anyone’s performance, including your own.

Key Numbers

What SEO ROI Actually Measures, and What It Does Not

In short: SEO ROI compares the profit organic search produced against everything SEO cost you. Profit, not revenue. Swap those two words and the number can inflate by a factor of three or more before you have made a single other mistake.

Return on investment is a ratio. You take what you got back, subtract what you put in, and divide by what you put in. Every ranking page on this topic agrees on that much. The disagreement starts one step earlier, at the question of what “what you got back” means.

It means gross margin, the money left after the cost of delivering the thing you sold. If organic search produced $50,000 in revenue and your gross margin is 30 percent, your return is $15,000. Plug in the $50,000 instead and a $10,000 spend reports 400 percent when the honest figure is 50 percent. Same quarter, same data, one substituted word.

The second confusion is traffic value. Semrush’s guide illustrates SEO’s worth by noting that Canva’s US organic traffic would cost $26.4 million a month to buy through ads [1]. That number is real and it is also not income. It is the price of clicks you did not pay for, which is a different claim from money that entered your account.

Third, ROI and payback period answer different questions. ROI is a ratio over a window you chose. Payback is a date, the point at which cumulative return crosses cumulative spend. A campaign can report a healthy ratio in month nine and still be underwater in cash terms. Owners get caught by this constantly. The report says the investment worked. The bank balance has not heard.

One ratio, three impostors. Get the definition right and the rest is arithmetic.

What You Need Before the Math Will Mean Anything

In short: four inputs decide whether your SEO ROI figure is real. Close rate, gross margin, tracked organic conversions, and a measurement window you set in advance. Miss two of them and you are not calculating anything, you are decorating an estimate.

The four numbers you cannot fake

A key event is GA4’s term for an action you have told it to count as a conversion, like a form submission or a purchase. Nothing is a key event by default. Someone has to configure it, assign it a value, and confirm it fires.

The window matters more than it looks. Google’s Analytics documentation sets the default lookback at 30 days for acquisition conversion events and 90 days for everything else [2]. That window is not a display setting. It is a definition of what counts as influence, and changing it does not reprocess history. Two quarters measured under different windows are not comparable, and nobody flags it for you.

What Google Search Console can and cannot tell you

Search Console is where most SEO reporting begins, and it holds none of what you need. It stores clicks, impressions, click-through rate, and average position on a rolling 16-month window. Once a date passes that edge, Google deletes it and there is no recovery [3].

It holds no revenue. No conversions. No customers. It cannot tell you whether a single one of those clicks was worth having.

That does not make it useless, it makes it upstream. Search Console answers whether people are finding you. Whether that mattered is a question for GA4 and your CRM together, which is also where what an SEO expert actually costs starts to become a number you can weigh against something.

So here is the honest exit. If you cannot state your close rate and your gross margin today, stop. A calculation built on two guesses returns a guess with decimal places on it, and decimal places are how a guess gets mistaken for a finding.

The Calculation, Step by Step

In short: total every cost, convert organic conversions into gross profit rather than revenue, then divide. Three steps produce two outputs, an ROI percentage and a payback date, and the second one is the one your cash flow cares about.

Step 1: Total your real cost

Your cost is not your invoice. Semrush’s guide gets this right and it is worth agreeing where the ranking pages are correct: the total includes agency or freelancer fees, content production, tooling subscriptions, link acquisition, developer time, and in-house salaries apportioned to the hours actually spent on search work [1].

The line almost everyone drops is their own time. If you or a member of your team spends eight hours a month briefing writers, approving copy, and sitting on calls, that is a cost. Price it at whatever you would pay to replace those hours and add it in. It never appears on an invoice, which is exactly why it goes missing, and on a small retainer it can be a fifth of the true spend.

For scale, one agency publishes a typical range of $2,500 to $7,500 a month, attributed to its own parent company [9]. Treat that as a vendor describing its own market rather than as a survey. Use it to check whether your spend is unusual, nothing more. Your figure is the one that goes in the denominator.

Total everything across the window you fixed in the previous section. That figure is your denominator, and it should feel higher than the number in your head.

Step 2: Isolate the revenue organic search actually caused

For ecommerce, this is a filtered report. Pull your GA4 monetization data, filter to organic search, and read the conversion value for the window.

For lead generation, you build the number. Take the value of a customer, multiply by gross margin, multiply by your close rate. That gives you what one tracked lead is worth.

Semrush teaches this as customer lifetime value multiplied by close rate [1]. That skips a step. Lifetime value is revenue, and revenue is not return, so the margin multiplier belongs in the middle. Leaving it out is the single most common way a lead-gen ROI figure gets inflated before anyone has made a mistake in the arithmetic.

Two cautions on the word “caused.” First, your attribution window decides which touchpoints are eligible for credit at all, and Google’s defaults sit at 30 days for acquisition events and 90 days for everything else [2]. A buyer who first found you 120 days before signing falls outside the second window entirely. The sale happened. The measurement stopped looking. Second, some of what organic search gets credited with would have arrived anyway, which is the next section’s problem.

Step 3: Run the formula, then find your payback date

The formula:

SEO ROI = (Gross profit from organic search – Total SEO cost) / Total SEO cost x 100

That gives you a ratio over a period. Now find the payback date, which the ratio hides. Lay cumulative spend and cumulative return side by side, month by month, and mark where the return line crosses the spend line. Early months usually return close to nothing while content indexes and rankings build, so the crossing lands later than an annual average implies.

The two are not the same event either. A campaign turns ROI-positive on paper when cumulative return passes cumulative spend for the window you are measuring. It breaks even in cash when the money has landed, which trails the report by however long your collection cycle runs.

What decides where that crossing falls is mostly not your industry. Four factors move it: average sale value, total management cost, the site’s existing search health, and, some way behind those, your sector [9]. A business with a high average sale and a lean cost base crosses early almost regardless of vertical. A low-ticket business carrying a heavy retainer can run a technically sound campaign for a year and still not have crossed. Check those four before you set an expectation with anyone.

The ratio tells you whether the investment worked. The date tells you when the money came back. Report both or you have answered half the question.

A worked example you can copy

Every figure below is illustrative, chosen to show the arithmetic rather than to represent any client result.

A lead-generation business, 12-month window.

Costs: retainer $2,500 a month, $30,000. Content production $700 a month, $8,400. Tooling $150 a month, $1,800. Internal time, 8 hours a month at $75, $7,200. Total: $47,400.

Return: 300 tracked organic leads over the year. Close rate 12 percent, so 36 customers. Average deal $4,200, giving $151,200 in revenue. Gross margin 45 percent. Gross profit: $68,040.

ROI: (68,040 – 47,400) / 47,400 x 100 = 43.5 percent.

Run the same numbers with revenue in place of gross profit and you get 219 percent. Both calculations use identical data. One of them is a return.

Payback: spend accrues at $3,950 a month from day one. If the first four months produce close to nothing and the remaining eight carry the full $68,040, cumulative return overtakes cumulative spend around month eight. A report covering the first two quarters would have shown this campaign underwater and been right to.

The Four Places the Number Gets Inflated

In short: an inflated SEO ROI figure rarely comes from bad arithmetic. It comes from four upstream choices about what counts, each defensible on its own, and each pushing the number the same direction.

Counting revenue instead of margin

The previous section showed the size of this one. The same twelve months reported 219 percent on revenue and 43.5 percent on gross profit. Nothing else changed. If a report hands you a percentage and does not state which of the two it used, that is the first question, and the answer is usually revenue.

The confusion runs deep enough that a page currently ranking in the top three for this query defines SEO ROI as revenue minus spend in its page description, and as gain from investment minus cost of investment in its key takeaways box, on the same page. Its worked example then describes a $50,000 figure as profit while the copy around it treats the same number as revenue [9]. If the pages teaching the calculation cannot hold that distinction for the length of one article, a quarterly report will not hold it by accident.

Using someone else’s close rate instead of your own

If you sell to leads rather than shoppers, close rate is the input that moves your result the most, and it is the one people are most tempted to borrow.

The borrowed figure is almost always the same one. Search the topic and you will find that leads from search close at 14.6 percent against 1.7 percent for outbound. Trace it and the ground gives way. Live pages attribute that number to HubSpot, to Search Engine Journal, to BrightEdge, to Ahrefs, and to Adcore, dated variously 2018, 2024, and 2025. One traceability audit reports that the widely quoted 14.6 percent close rate has no published methodology, and that HubSpot’s more transparent 2012 analysis of 150 businesses found 15 percent for SEO leads and 2 percent for outbound [4].

Five attributions, three dates, one number, no method. Your CRM knows your actual close rate and it took no survey to find out.

Letting branded search take credit

Branded search is someone typing your company name into Google. Non-branded search is someone typing the problem you solve. Both land in the same organic bucket, and only one of them represents demand that search work created.

A customer who already had your card, already heard you on a podcast, or already got a referral will still arrive through organic search. Counting that conversion as SEO’s return measures the channel’s plumbing, not its contribution. Economists call the distinction incrementality: what would have happened anyway, versus what happened because of the spend.

The fix takes ten minutes. Open Search Console, filter queries containing your brand name, and exclude them. Rerun the calculation on what remains. The number gets smaller and it gets true.

Trusting an attribution model that quietly switched

GA4’s default model distributes credit across touchpoints using machine learning. It requires conversion volume to do that, and a property below the threshold falls back to a rules-based model without telling anyone. The settings screen still says data-driven. The report is last-click.

Worth knowing before you repeat a threshold you read somewhere: the specific numbers in circulation, 400 conversions per conversion type over 28 days plus 10,000 paths, come from a Google help page labeled legacy Universal Analytics [5], describing a model that predates GA4. Google Ads publishes different figures again for its own product, 3,000 ad interactions and 300 conversions over 30 days [6]. Three products, three sets of thresholds, one number repeated as though it covered all of them.

Ask your provider which model actually ran during the reporting window, and how they confirmed it.

What a Good SEO ROI Looks Like, and Why the Benchmarks Mislead

In short: a good SEO ROI is one that beats your other acquisition channels and clears your cost of capital. There is no defensible industry average to measure against. The figures presented as averages come from sources with a stake in the answer.

Search the question and one number comes back everywhere: 748 percent. It appears as the median SEO ROI, and elsewhere on the same pages as the average, redated 2024, 2025, and 2026 depending on where you land.

It traces to First Page Sage, an SEO agency, drawn from its own client dataset of roughly 170 companies split 76 percent B2B and 24 percent B2C [7]. That is a vendor reporting returns on its own work. It may be accurate. It is still the least independent available source for the question of whether hiring an SEO vendor pays off, and no page that cites it says so.

The copying shows. Several pages state 748 percent, which is $7.48 returned per dollar, and 22:1, which is $22 returned per dollar, inside the same article, both labeled as the industry figure, neither reconciled against the other. Numbers assembled from other pages rather than measured tend to arrive in pairs like that.

The pattern is not confined to one company. Another ranking page opens with two figures, that 49 percent of marketers credit organic search with the highest returns and that organic search generates 40 percent of business revenue, and sources both to its own statistics page [9]. An agency citing itself as the authority on whether agencies deliver.

Three better reference points, all sitting in your own records:

• Your other channels. If organic cost per acquisition is $1,300 and paid is $2,900, organic is winning at any ROI percentage.
• Your cost of capital, meaning what the same money would have returned somewhere else, including doing nothing with it. Clear that and the investment is defensible.
• Your own trailing four quarters. Direction beats altitude. A number moving the right way under a fixed measurement window is worth more than a number that beat a stranger’s average.

One piece of arithmetic beats every benchmark on this page. If your average sale is $15 and you spend $1,500 a month, you need 100 sales a month before SEO has earned anything [9]. Run that division first. For some businesses it ends the conversation, and it is better to learn that in an afternoon than in year two.

None of these require a benchmark, and all three survive a conversation about what a retainer actually buys and how engagement models differ.

We are dismissing figures that flatter this industry, including this agency. That is the direction the error runs.

Validation Checks: Five Questions to Ask Your Report

In short: five questions separate an SEO ROI figure you can act on from one that only looks finished. They work on a vendor’s deck and on your own spreadsheet, and none of them require you to run the calculation yourself.

Question four is the one almost nobody asks, and it is the hardest to answer badly. Question five is the one that still helps when the other four come back unsatisfying, because a cost per acquisition you can compare across channels gives you a decision even when the ROI percentage is unreliable.

One note on how to read the answers. A provider who says “I do not know, let me check the settings” on question four is showing you more than one who answers in half a second. These are configuration questions, and configurations get changed by people who have since moved on. Certainty here is not always a good sign.

Run these on your own reporting first. Most of the time the gap turns up on your side of the table, in a close rate nobody has updated since 2023 or a margin figure that predates your last price change.

Why the 2026 Numbers Are Harder to Read Than 2022’s

In short: organic traffic and organic revenue have come apart. A search result can hold its position, keep its impressions, and send fewer people to the site, which means traffic trends are now weak evidence in either direction and conversions have to carry the argument.

AI Overviews are the AI-written summaries that appear above search results. A zero-click search is one that ends without the person visiting any site, because the answer arrived on the results page.

The scale of the shift has been measured. The Pew Research Center analyzed the browsing activity of 900 US adults across nearly 69,000 Google searches in March 2025, and found that when an AI Overview appeared, 8 percent of searches produced a click on a standard search result, compared with 15 percent when no summary was present [8]. Clicks on the links inside the summaries themselves occurred in around 1 percent of visits [8].

That is a research institution measuring real browsing behavior, which makes it a different class of evidence from the agency benchmarks in the previous section.

Here is what it does to your calculation. Impressions can hold steady or climb while clicks fall, because the ranking did not change, the behavior did. If your ROI narrative rests on session counts, the number now moves for reasons that have nothing to do with the work being done. Traffic decline is no longer proof that SEO failed, and traffic growth is no longer proof that it worked.

Two adjustments follow. Judge organic performance on conversions rather than sessions. Session counts and rankings are vanity metrics when read on their own, and they have become less informative than they were, while a conversion still requires someone to arrive. And segment Search Console by query intent before drawing conclusions, because informational and transactional queries are moving on different trajectories now.

Nobody has a settled baseline for any of this yet. Anyone quoting you a precise industry-wide figure for what AI search cost your vertical is estimating, and you should treat that estimate the way you would treat any other convenient number.

Frequently Asked Questions

What is the formula for SEO ROI?

SEO ROI equals gross profit from organic search minus total SEO cost, divided by total SEO cost, multiplied by 100. The inputs matter more than the arithmetic. Total cost includes internal hours and tooling, not only the agency invoice, and the return figure should be gross profit rather than revenue.

Should I use revenue or profit in the calculation?

Gross profit. Revenue is what the customer paid you, and profit is what you kept after delivering the thing they bought. The same twelve months of data can report 219 percent on revenue and 43.5 percent on gross profit, so a figure quoted without stating which one it used is not yet an answer.

What is a good SEO ROI?

One that beats your other acquisition channels and clears what the same money would have earned elsewhere. There is no reliable industry average to measure against, because the figures published as averages come from SEO vendors reporting on their own client work. Your own trailing four quarters is a better reference point than any benchmark.

Does SEO ROI vary by industry?

Enormously, which is the problem with industry averages. The most widely cited dataset reports 317 percent for ecommerce and 1,389 percent for real estate from the same source. A range that wide means the aggregate figure describes no individual business closely enough to plan against.

How long does SEO take to pay off?

Longer than one quarter and usually longer than two. Early months carry full cost while content indexes and rankings build, so cumulative return tends to cross cumulative spend somewhere in the second half of the first year. Calculate your own payback date rather than accepting a general timeline.

Can Google Search Console show me SEO revenue?

No. Search Console holds clicks, impressions, click-through rate, and average position on a rolling 16-month window, and it stores no revenue, conversion, or customer data at all. It answers whether people are finding you. Whether that was worth anything is a question for your analytics and your CRM together.

How do I track organic conversions in GA4?

Configure the actions that matter as key events, assign each one a monetary value, then filter your reporting to organic search. Nothing is tracked as a conversion by default. For lead generation, the value of one lead is customer lifetime value multiplied by gross margin multiplied by your close rate.

How do I separate branded from non-branded organic?

Open the Search Console performance report, filter queries containing your company name, and exclude them. What remains is closer to demand that search work created rather than demand that already existed. Run your ROI calculation both ways and report both numbers.

Why did my organic traffic drop but leads stay flat?

Informational and transactional queries are moving on different trajectories. Pages answering general questions lose clicks to AI summaries faster than pages answering buying questions, so a site can shed sessions while holding the sessions that convert. Segment Search Console by query intent before treating a traffic decline as a failure.

What should I do if my SEO ROI comes back negative?

Check the measurement before you change the vendor. A negative figure built on a borrowed close rate, a guessed margin, or a window that still includes branded search is not yet evidence about anyone’s performance. Once the inputs are yours, compare cost per acquisition against your other channels, then look at scope and average sale value before you look at contract terms.

Run Your Own Number First

The calculation is arithmetic. The honesty is the part that takes work, and most of that work happens before you divide anything, in deciding what counts as a return and what counts as a cost.

So run it on your own last four quarters before you change anything else. Total the real spend, including the hours nobody invoices for. Convert conversions to gross profit using your close rate and your margin. Exclude branded search. Fix the window and write down the date you fixed it.

If the number comes back negative, you have learned something worth knowing. If it comes back unknowable, that is the more common result, and it points at a measurement problem rather than a vendor problem. Close rate and margin are not SEO deliverables. They come from your CRM and your accounting, and no agency can supply them for you.

Once you can answer the five questions in this article about your own reporting, you are in a position to ask them of anyone else. If you want that conversation, we are here.

Definition Bank

Entity Cards

Google Search Console

Google Analytics 4

The 748 Percent SEO ROI Benchmark

Sources

1. Semrush, “The ROI of SEO: How to Measure SEO ROI (with Formulas),” updated May 15, 2024.
2. Google, Analytics Admin API, AttributionSettings reference documentation, accessed 2026.
3. Google, Search Console Help, performance data retention and the rolling 16-month window, accessed 2026.
4. ProfileTree, inbound marketing statistics traceability audit, 2026, citing HubSpot’s 2012 State of Inbound analysis of 150 businesses.
5. Google, Analytics Help, “[UA] MCF Data-Driven Attribution model application and limits [Legacy].”
6. Google Ads Help, “About Switch to DDA,” data-driven attribution eligibility thresholds.
7. First Page Sage, SEO ROI benchmarks report, dataset self-described as 170+ client companies, 76 percent B2B and 24 percent B2C.
8. Pew Research Center, study of Google AI Overviews and browsing behavior, 900 US adults, March 2025 data, published July 2025.
9. SEO.com, “What Is Your SEO’s ROI? Learn How to Calculate Yours Now,” updated August 12, 2026.