Free SEO ROI CalculatorProfit, payback month and three scenarios in one forecast.

Return on investment has a simple standard formula: gain minus cost, divided by cost. The formula is rarely the problem. The problem is what people put into the "gain" slot. Most SEO ROI estimates use revenue, and revenue still contains the cost of goods, delivery, payment fees and support.

Here is how much that choice changes the answer. An online store earns $60,000 in extra organic revenue in a year and spends $24,000 on SEO. Revenue ROI is ($60,000 − $24,000) ÷ $24,000 = 150%. The store's gross margin is 40%, so the real gain is $24,000 of gross profit. Profit ROI is 0%. The program is the same in both cases, yet one number says "excellent" and the other says "break-even."

Two more problems distort SEO ROI. First, ROI is not time-adjusted, as the Wikipedia entry notes, so a 6-month view and a 24-month view of the same program can tell opposite stories. Second, Wikipedia also points out that a marketing ROI figure is only credible when the program's effect is isolated from other influences. Branded searches grow with TV ads, PR and word of mouth, and SEO should not take credit for that growth.

The definition we recommend fixes all three issues in one sentence:

Formula: SEO ROI = (Gross profit from incremental non-branded organic traffic − Total SEO cost) ÷ Total SEO cost × 100, over a stated period.

This works because each part answers a question your CFO will ask. Gross profit is the figure finance compares with every other investment. "Incremental non-branded" isolates the effect of SEO from brand demand. A stated period makes the result comparable with other channels and with your next forecast.

Most SEO forecasts start at the site level: "we expect organic traffic to grow 30% next year." The number sounds clear, but it mixes several things that SEO does not control. Brand demand, seasonality and pages that nobody worked on all move the site total.

A software company sponsors a popular podcast. Branded searches double, total organic traffic rises 25%, and the SEO team receives the credit. The next quarter the sponsorship ends, organic traffic falls, and the same team now has to explain a "decline" it did not cause. A site-level forecast can also hide the fact that most of the growth came from one blog post that never produces a lead.

The solution is a smaller unit of measurement: one landing page and one group of non-branded keywords that the page should rank for. Search engines rank pages for queries, not websites for percentages. A page-level forecast lets you check each line against reality later.

SparkSEO, a performance-oriented SEO agency based in Barcelona, builds its pricing on this unit. Under its performance-based SEO model, each landing page is tied to one approved group of non-branded keywords. The monthly fee has a base part and a performance part that rises when those rankings improve and falls when they drop. A fee that follows rankings only works when both sides agree, before work starts, on what each ranking gain is worth, so every engagement starts with a forecast per landing page. The calculator on this page uses the same structure. For calibration, the agency publishes its own results: across 14 clients with at least 12 months of data, the median growth in non-branded organic clicks was 71% after 12 months.

In practice, a page-level forecast is a table with four columns: the landing page, its keyword group (a main keyword plus close variants), the current position and a realistic target position. A realistic target is one that the current top-ranking pages suggest you can reach in 6 to 12 months with better content and technical fixes. This format works because every line is auditable. Six months later, you can open Search Console, filter by the keyword group and see whether the forecast held.

Keyword groups and click-through rate

Click-through rate (CTR) is the share of searchers who click your result after they see it. A common shortcut multiplies search volume by the CTR at the target position and calls the result "new traffic." This overstates the gain, because the page usually gets some clicks already.

Imagine a page in position 9 for a keyword with 1,900 monthly searches. At a 2.6% CTR, it already earns about 49 clicks a month. If you forecast position 2 at 16% CTR and count all 304 clicks as new, you double-count the 49 clicks you already have. The opposite error is also common: forecasts ignore the long tail. A page that ranks for its main keyword usually also ranks for dozens of related searches, and those clicks often add up to more than the main keyword brings.

The formula that avoids both errors:

Formula: Incremental visits per month = Monthly searches × (CTR at target − CTR now) × Long-tail factor × AI Overview factor

Here is the formula applied to three keyword groups for a B2B HR software company. The long-tail factor is 1.8, and the AI Overview factor is 0.812 (explained in the next section).

For the first row, the math is 2,800 × (11% − 1.2%) = 274 clicks, × 1.8 = 494, × 0.812 = 401 visits.

Two inputs deserve your own data. The default CTR curve in the calculator is a starting point. Your Search Console CTR by position for non-branded queries is better, because CTR changes with ads, shopping results, maps and brand familiarity in your market. For the long-tail factor, a page that already ranks well is the best reference. Its total clicks divided by the clicks from its main keyword give you the ratio, and that ratio reflects how your pages actually perform.

Adjusting for AI Overviews

AI Overviews are the AI-written summaries Google shows above the classic results for some searches. They answer part of the question on the results page, so fewer people click through. Pew found that users clicked a traditional result on 8% of visits with a summary and on 15% without one. That is a drop of about 47%. Only 1% of visits included a click on a link inside the summary itself.

The share of searches that show a summary varies a lot. In Pew's March 2025 data, 18% of all searches produced an AI summary, and 60% of searches phrased as questions did. Commercial software searches fall between these two figures, so the B2B example assumes 40%. The AI Overview factor is then 1 − (0.40 × 0.47) = 0.812. In plain terms, the keyword groups deliver about 19% fewer clicks than an old CTR curve would suggest.

The practical method is to open the live results for each keyword group, note whether an AI Overview appears, and use that share in the calculator. This keeps the forecast conservative and honest. It also gives you a reason to earn citations inside AI answers. Those citations bring brand exposure that the ROI model does not count, so any value they add sits on top of your estimate.