SEO is worth it when the margin from organic traffic exceeds the cost of the work before the budget runs out. It is a division, not a verdict on the channel: monthly fee divided by the value of a click. The result tells you how many clicks a month you need to break even, and from there you quickly see whether search demand in your sector contains those clicks or not.
We looked at the first page of Google UK for "is seo worth it" on 26 September 2026. It opens with an AI Overview that answers yes and lists generic benefits, followed by Reddit and Quora threads, a video carousel and five articles from agencies and tool vendors. Only one of them, from The SEO Works, models the return with real campaign figures, and it works on payback periods rather than on the monthly sum. The calculation you need before signing a contract is not the first thing any of these pages gives you.
The value of a click is where everything starts
An organic click is worth the average margin on a customer multiplied by the probability that the click becomes a customer.
Value of a click = margin per customer × conversion rate
With a margin of £500 and a 2% conversion rate, a click is worth £10. A £1,000 monthly retainer pays for itself with 100 organic clicks a month. Not "more traffic": one hundred clicks, counted.
Both figures should come from your own data. The conversion rate from Google Analytics, because the difference between 1% and 3% triples the traffic you need and moves break-even by months. The margin after costs, not the list price. They are the same two figures you will later need to calculate SEO ROI after the fact, instead of arguing about it on gut feeling.
How many clicks you need to cover the fee
The fees in the table sit within the bands the UK market publishes: from about £300 a month at the bottom to over £20,000 at the top, with most ongoing offers between £500 and £5,000. We lined up the full bands, and who wrote them, in the article on how much SEO costs.
| Margin per customer | Value of a click | Clicks/month for a £500 fee | £1,000 fee | £2,500 fee | Impressions/month equivalent to the £1,000 fee |
|---|---|---|---|---|---|
| £200 | £4 | 125 | 250 | 625 | 5,882 |
| £500 | £10 | 50 | 100 | 250 | 2,353 |
| £2,000 | £40 | 13 | 25 | 63 | 588 |
| £5,000 | £100 | 5 | 10 | 25 | 235 |
The last column is the one that decides. With a margin of £500 per customer, if the entire search demand relevant to you is worth fewer than 2,400 impressions a month, a £1,000 retainer will not pay back even at full speed. It is not a question of how good the agency is: the clicks simply are not there.
What a position is really worth: 1,000 queries measured
Most CTR-by-position curves you find online are taken from large aggregate studies, on samples of sites that do not look like yours. This one comes from Search Console: an Italian outdoor equipment ecommerce site we manage, web search, Italy only, from 1 July to 25 August 2026. One thousand queries, 21,701 impressions, 922 clicks, overall CTR 4.25%.
| Average position band | Queries | Impressions | Clicks | Weighted CTR |
|---|---|---|---|---|
| 1.0 - 1.9 | 118 | 3,547 | 696 | 19.62% |
| 2.0 - 2.9 | 77 | 3,934 | 91 | 2.31% |
| 3.0 - 3.9 | 85 | 2,311 | 46 | 1.99% |
| 4.0 - 5.9 | 131 | 2,408 | 50 | 2.08% |
| 6.0 - 10.0 | 206 | 4,561 | 23 | 0.50% |
| 10.1 - 20.0 | 275 | 4,537 | 15 | 0.33% |
| over 20 | 108 | 403 | 1 | 0.25% |
| Total | 1,000 | 21,701 | 922 | 4.25% |
The jump that matters is between the first and second band: 19.62% against 2.31%. For the same impressions, dropping one average position costs eight times the clicks. From page two down traffic becomes a rounding error: 4,537 impressions in the 10-20 band produced fifteen clicks in eight weeks.
Two caveats on the data, so that it is useful. Search Console's average position is an average across different impressions, so a query "in band 2" was never steadily in second place: aggregation compresses the CTR of the top bands. And the export stops at a thousand rows, the API limit, so the long tail beyond the thousandth query is not included and the "over 20" band is underestimated. The ratio between the top bands still holds; the absolute value of the tail does not.
How much AI has taken out of the sum
The Pew Research Center tracked the browsing of 900 US adults and analysed 68,879 Google searches made in March 2025, 12,593 of which had an AI-generated summary at the top. When the summary was there, users clicked a traditional result in 8% of visits; when it was not, in 15%. Links inside the summary were clicked in 1% of cases. It is a US sample, and should be read as such.
Rand Fishkin of SparkToro, using the Similarweb panel, measured that in the first four months of 2026, 68.01% of Google searches in the United States ended without a click, against 60.45% in 2024. Again: desktop and mobile, United States.
The number that is usually missing we have on a real cluster. On that of Macropix, an Italian manufacturer of LED walls and LED screens, 32.7% of the tracked keywords show an AI Overview, and those keywords account for 24.7% of the cluster's search volume. When the AI Overview appears, 43% of searches end without a click. Multiply the two shares and just over 10% of that cluster's volume is traffic that will never reach any site. For the break-even sum, that means raising the impressions you need by 10%, not throwing out the model.
The share varies a great deal by sector and query type. Before applying a percentage you read somewhere, measure how many of your own keywords actually trigger an AI Overview, and for which intents: on transactional searches zero-click searches are still marginal; informational searches are where it hurts.
How much demand there really is in the UK
The break-even sum assumes that someone is searching. According to Ofcom's Adults' Media Use and Attitudes report, published on 2 April 2026 (7,533 adults, fieldwork September to November 2025), only 6% of UK adults have no home internet access, and 83% of that group are aged 65 or over. One in five online adults (20%) goes online only through a smartphone. Demand is not concentrated among twenty-somethings: if you sell boilers or hearing aids, your customer is online.
On the other side, the one that was supposed to turn everything upside down, the UK figures move faster than in much of Europe but are shallower than the story suggests (we go into more detail in the guide to marketing for small businesses). In its Business Insights and Conditions Survey the ONS measures that in June 2026 around 35% of UK businesses with ten or more employees used at least one AI technology, up from around 12% in September 2023. Among businesses with 250 or more employees the figure is 49%. But adopters use on average 1.6 AI technologies, and only 10% say they use AI extensively. The curve is rising fast, but adoption is still thin: anyone who puts SEO off today because "in two years everyone will search with AI" is betting on a replacement that UK data do not yet show.
The opposite also holds, and it is why we have not written that nothing changes: the two curves, falling clicks and rising adoption, are moving in the same direction. Anyone setting up organic search work today should do so knowing that part of the return will come as citations inside generated answers, that is GEO, and not only as sessions counted in Analytics.
Three cases where the sum does not add up
Demand is too small
This happens with new products, with categories people do not know they should search for and with very vertical B2B niches. If the volumes of all relevant queries added together do not reach the impressions in the table, the channel cannot pay back the fee, however well it is worked. In these cases demand has to be created elsewhere and captured later, and SEO becomes a second-year investment.
The margin is too low
With £200 of margin per customer and a 2% conversion rate, a £1,000 retainer needs 250 clicks a month, which means almost 5,900 impressions. An ecommerce site with a low average basket only gets there with a wide catalogue: that is why ecommerce SEO is won on category pages and not on a handful of head keywords. With thin margins and a narrow catalogue, the sum does not close.
The time horizon is shorter than the technical lead time
In Ahrefs' 2025 study only 1.74% of published pages reach the top 10 within a year, and of those that do, 40.82% get there within the first month: either a page moves straight away, or it needs much more than thirty days. In Morningscore's February 2026 survey of 75 professionals, 67.2% see the first result between the second and fourth month. We collected the studies and our own numbers in the article on how long SEO takes to show results. If you need revenue within the quarter, that budget works harder on Google Ads, and SEO can start when cash flow allows.
What actually broke even, on three projects
Macropix sells LED walls to large companies and public bodies in Italy: high margin, modest search volumes, exactly the top-margin row of the table. The main keyword in the cluster moved from position 88 to 2, and in July 2026 the domain held 25.55% share of voice on the LED wall cluster, ahead of Amazon. On the paid side, €805 invested produced conversions at €44.93 each, with an average CTR of 3.59% and an average CPC of €0.81; compared with the previous twelve months, conversions grew by 84.2%.
The outdoor ecommerce site from the table sits at the opposite end: low margin, high volumes. In twelve months it recorded 583,661 impressions and 13,704 clicks from Italy, with an average position of 6.6 and a CTR of 2.35%. The sum closes on the number of clicks, not on any single ranking.
The third is a services site that started from nothing: average position from 47.5 to 16.9 in eight months, a 64% improvement, with daily clicks growing from 1.31 to 1.78. In absolute terms these are crumbs, and by month eight the fee had not yet been recovered. This is the phase in which almost everyone turns off the tap, and in which you need leading indicators and goals written with numbers and dates to know whether what you are paying for is working.
What the three have in common is the business model, not the sector. A B2B company with deals worth tens of thousands of pounds breaks even on ten clicks a month; a neighbourhood business breaks even on even lower volumes but on local queries, where the competition is a dozen shopfronts, not the whole national market.
The cost of stopping
The sum always includes the cost of starting, almost never the cost of stopping. A paid campaign switches off and the next day traffic is zero, but no capital has been burnt: it was rent. Organic rankings work the other way round. They build up slowly and are lost slowly, which is reassuring until you realise that a page's age is itself a competitive advantage: in Ahrefs' 2025 study 72.9% of pages in the top 10 are more than three years old, against 59% in 2017.
If you pause for six months and restart, you do not pick up where you left off. You restart further back, because in the meantime the pages that overtook you have built up months of age that you will have to buy back. That is why a small, continuous retainer almost always beats a large one-off project, and it is also the most useful question to ask a supplier before signing: not "how much traffic will you bring me", but "what happens to my site if I stop in a year". If they cannot answer, the quote is a list of activities, not a plan. The twelve changes that move the needle most and the projects we applied them to show where we start; if you want the sum done on your own numbers, the starting point is our SEO service.
Frequently asked questions
