Translate a ranking position into an estimate of monthly clicks, and see what moving up would actually be worth. Three curve models are included, because in 2026 a single CTR curve is no longer honest - the presence of an AI Overview changes the answer by roughly half.
CTR & Traffic Forecast
LiveThe full curve
Why there are three curves now
Until about 2023, one CTR curve was enough - rank higher, get proportionally more clicks, and the numbers held steady across studies. That has broken, and the reason is what Google now puts above the results.
AI Overviews appear on roughly 13-14% of all searches in the US, and closer to a third of purely informational queries. Where one appears, the first organic result is pushed below the fold. Semrush's analysis of 17.8 billion queries put the loss at position 1 at around 55%; Ahrefs' study of 300,000 keywords reported a 58% reduction for top-ranking pages, with position 2 down about 51% and position 3 about 46%. Seer Interactive measured overall organic CTR on those queries falling from 1.76% to 0.61%.
There is a more hopeful counterpoint worth knowing: Seer's April 2026 follow-up found those figures stabilising, and in some segments recovering substantially over two months. The picture is still moving, which is exactly why forecasting from a single fixed curve is now misleading.
Search your target keyword and look at the actual results page. AI Overview at the top? Use the AI Overview curve. Clean list of ten blue links? Use the clean curve. Forecasting across a portfolio of mixed keywords? The blended average is the safer planning number. Guessing wrong in the optimistic direction is how traffic forecasts end up 40% too high.
What the numbers actually say
- The gap between position 1 and 2 is the largest on the page - roughly double, in every study. Nothing else on page one comes close to that jump.
- Positions 8-10 are close to worthless in traffic terms. Moving from 9 to 8 changes almost nothing; moving from 4 to 3 changes a great deal. Prioritise accordingly.
- Roughly 60% of Google searches now end without any click to an external site. Impressions rising while clicks stay flat is not a reporting error - it is the current SERP.
- Local pack CTR is a different curve entirely, and a much flatter one - going from first to third in the map pack costs a couple of percentage points, against nearly thirty in organic. If you're doing local SEO, don't use these numbers.
Common mistakes
Frequently asked questions
Why do the studies disagree so much on position 1?
Reported figures range from about 19% to 39.8%, and the spread is methodological rather than contradictory. Studies differ on whether they include branded queries, whether the dataset is mobile-heavy, and crucially whether AI Overviews were present. First Page Sage's higher numbers describe clean results pages; GrowthSRC's lower ones reflect SERPs where an AI Overview sits above the results. Both are correct for what they measured.
Where do I get search volume to enter?
Google Keyword Planner gives free ranges to anyone with an Ads account, and the paid tools give more precise figures. If you already rank for the term, Search Console impressions are better than any of them - impressions are actual measured demand for your specific SERP rather than a modelled estimate.
Is CTR a ranking factor?
Google has consistently said it is not used as a direct ranking signal, and attempts to game it don't produce durable results. That doesn't make CTR unimportant - it's the difference between a position being worth traffic and being worth nothing. Optimise it because clicks are the point, not because you expect a rank change.
How do I improve CTR without changing position?
Rewrite the title to match the query's actual intent rather than your internal page name, write a description that promises something specific, and add structured data where a rich result is available - a star rating or FAQ expansion makes a listing physically larger. Pull your worst CTR-to-position outliers from Search Console and start there; those are where the gap between expected and actual is largest.
Does this work for keywords I don't rank for yet?
Yes - set the current position to 20 or leave it blank in effect, and the target to wherever you realistically expect to land. Be conservative: reaching position 3 for a competitive term is a multi-month project, and a forecast built on reaching position 1 is a forecast built on the least likely outcome.
What value per click should I use?
For an ad-supported site, divide your RPM by 1,000 to get revenue per pageview. For a business site, multiply conversion rate by average order value or lead value. Leaving it at zero is fine - the click figures stand on their own, and a made-up value only makes the forecast look more precise than it is.
Related calculators
See the full list of Business calculators, or try: