Organic CTR in 2026: What Is a Good CTR on Google With and Without AI Overviews

What is a good CTR on Google: what the studies of CTR by position measure (SISTRIX, First Page Sage, Ahrefs, Seer and Amsive), how much AI Overviews cut the click on first place, why branded and non-branded searches do not compare and how to find your site’s expected CTR in Search Console.
Written with AI, reviewed by humans

In this article

See how your brand shows up in AI answers

Book a demo and find out what to fix first, with a GEO specialist.

Book a demo

Reading time: 11 minutes

There is no single ideal organic CTR: a good CTR depends on position, type of search and AI Overviews. In studies with a published method, first place gets 1.6% to 28.5% of clicks on average. So use the studies as a rough guide, and compare your site with itself.

In short

  • Search Console calculates CTR as clicks divided by impressions, and an impression counts even when the link is off screen.
  • The studies disagree on the level of first place, but agree on the direction: AI Overviews go with fewer clicks, even below first place.
  • Branded and non-branded searches do not compare, because branded ones rarely show an AI Overview and usually have a much higher CTR.
  • So the ideal CTR for your site comes from your Search Console. Remove the brand, group by position and divide total clicks by total impressions.

What is organic CTR, and how does Search Console calculate it?

Organic CTR is the share of times a regular search result was shown and got a click. According to the Performance report help, Search Console calculates CTR as the click count divided by the impression count. So the number depends on what Google counts as an impression, a click and a position.

The help page on impressions, position and clicks has three rules that change how you read the data:

  • Impression: counts when the link is on the current results page, even if you do not scroll to it. The exception is a link that only appears after a click.
  • Click: counts only if it leads to a page outside Google. Also, clicking the same link again after going back counts once.
  • Position: is the highest one your site holds in each search, and the report shows the average of those positions. So an AI Overview, which holds a single position, gives the same position to all its links.

AI answers change this math, so the guide on AI data in Search Console explains what goes into each report.

Why does a site’s average CTR mislead?

A site’s CTR mixes, mainly, branded searches, positions and devices. For example, in a hypothetical case of ours, query A has 1,000 impressions and 100 clicks, or 10%. Query B has 9,000 impressions and 90 clicks, or 1%. So the site’s CTR is 190 divided by 10,000, which is 1.9%, and not the simple average of 5.5%.

Organic CTR: illustration of a half-moon gauge next to a list of results with a cursor clicking the last one

What is the average organic CTR by position in Google?

There is no ideal CTR in a single table, because each study measures a different set of searches, device and period. Even so, the curves have the same shape: first place takes the biggest share, and tenth place falls below 3%. So the table shows the two studies that publish CTR by position without AI.

PositionSISTRIX (mobile)First Page Sage (no AI)
128.5%22.6%
215.7%10.2%
311%5.8%
102.5%0.5%

Who measures what in these studies?

SISTRIX, which sells SEO tools, analyzed more than 80 million keywords on mobile only. The study came out in 2020, before AI Overviews, but the page does not name the country.

First Page Sage, an SEO and GEO agency, used 1,184 client and panel domains. That is 3.67 billion impressions, from March to August 2026. The values blend desktop and mobile and follow US search volume. However, the sample is the agency’s own, and the page does not publish the data, so outsiders cannot check it.

Advanced Web Ranking, which sells rank tracking, publishes the top 20 curve every month. The data comes from the Search Console of millions of US searches, with filters for device and intent. The values appear only in charts, so we do not quote them here.

Only First Page Sage and Advanced Web Ranking name a market, the US; SISTRIX, Ahrefs, Seer and Amsive name no country. So if your audience is elsewhere, such as the UK or the EU, confirm the pattern in your own Search Console data. Also, page blocks change the curve: in SISTRIX, the featured snippet took 5.3 percentage points from first place. What changed in position zero is in the comparison between featured snippet and AI Overview.

How much does an AI Overview cut the organic CTR of the first position?

In studies that separate the searches, an AI Overview goes with an organic CTR for first place that is about half to more than 80% lower. The drop is 58% at Ahrefs and 84% at First Page Sage. Advanced Web Ranking also reports a cut on the order of half. The levels, however, are very different.

Study1st place without AI1st place with AI
Ahrefs (desktop, Dec 2025)3.9%1.6%
First Page Sage (Mar to Aug 2026)22.6%3.6%

The First Page Sage numbers are the most extreme of the group, come from the agency’s own sample and cover any AI answer, AI Overview or AI Mode.

Ahrefs, which sells SEO tools, picked 300,000 keywords: 150,000 with an AI Overview and 150,000 informational ones without. The study used desktop CTR from Search Console (sum of clicks over sum of impressions) and compared December 2023 with December 2025. Ahrefs then calculated the 58% against a projection of 3.7% without an AI Overview, not against the column beside it. The page does not name the country, and the study shows association, not cause.

In other positions, however, the drop is smaller. According to Ahrefs, it was 50.8% in second place, 46.4% in third and 19.4% in tenth. First Page Sage shows the same pattern: second place has 10.2% without AI against 1.4% with AI, and third place has 5.8% against 0.8%.

Why does first place go from 3.9% to 22.6% without an AI answer? Neither page explains the gap. In our reading, the set of keywords, the device, the period and the definition of position 1 all matter. So use these numbers for the direction and size of the drop, not as a good CTR target.

What changed from late 2025 to 2026?

Seer Interactive, a search marketing agency, tracked 53 brands from January 2025 to February 2026. That is 5.47 million queries and 2.43 billion organic impressions, with no split by position. The page does not name the country.

On searches with an AI Overview, organic CTR rose from 1.3% in December 2025 to 2.4% in February 2026. Without an AI Overview, however, it reached 3.8%. For informational searches, Seer estimates, per million impressions, about 33,500 clicks without an AI Overview, 20,700 with the brand cited and 9,400 without a citation. By our math, those are CTRs of 3.4%, 2.1% and 0.9%, rounded. Even so, Seer itself says it does not claim cause.

For the part that clicks do not reach, see how AI Mode and AI Overviews choose their sources. Likewise, the guide on SEO and AI covers the basics of both fronts.

Can you compare branded and non-branded organic CTR?

No, you cannot. Branded searches rarely show an AI Overview and usually have a much higher CTR, because people already want to reach the site. So the ideal CTR of one does not fit the other. Mixing them hides what happens on non-branded searches, where AI Overviews weigh most.

Amsive, a marketing agency, analyzed 700,000 keywords from 10 sites in 5 sectors. It focused on 10,000 that triggered an AI Overview and already ranked. The study came out in April 2025, but the page names no country or period. Only 4.79% of branded keywords triggered an AI Overview, and for them the average CTR rose 18.68%.

For non-branded keywords, on the other hand, CTR fell 19.98%. Across all keywords outside the top 3, the average drop was 27.04%. These numbers, however, are relative changes, not rates. In SISTRIX, searches with sitelinks, typical of people looking for a specific site, gave first place 46.9%, against 34.2% on purely organic results.

So split branded from non-branded before you compare any number. The branded queries filter in Search Console splits data by whether the term includes the brand name. However, it does not appear on sites with few impressions. In that case, regex in Search Console does the same filtering.

How do you find the good CTR for your own site?

The good CTR for your site, or rather the expected CTR, comes from your own Search Console. Remove branded queries, group by position and divide total clicks by total impressions in each group. This method is ours, not a Google standard; the studies serve only as a ruler.

  1. Export the data: in the Performance report, download queries with clicks, impressions and position, by device, for the last 3 months.
  2. Remove the brand: use the brand filter or regex.
  3. Cut the noise: drop queries with fewer than 100 impressions. The cutoff is ours, but it works as a starting point.
  4. Group by position: 1, 2, 3, 4 to 5 and 6 to 10, with the position rounded.
  5. Calculate the group’s CTR: add the clicks, add the impressions and divide one total by the other.
  6. Split out AI Overviews: tag the keywords that have an AI Overview and rebuild the groups. To do this at scale, see how to check whether a keyword has an AI Overview.

So your site’s CTR becomes a curve of its own. Then a query with a CTR well below the group’s curve is a candidate for a title review. A query in line with the curve but with few clicks may lose clicks because of position, search type or a block above it.

For example, an online store with an average CTR of 1.8% cannot tell from that number alone whether it is doing well. In Seer’s cohort, organic CTR was between 2.4% and 3.8% in February 2026, but those were 53 brands with no split by sector. So compare the 1.8% only after you remove the brand and group by position.

How can you raise CTR without changing position?

To reach a good CTR without changing position, adjust what shows in the result: title, snippet and the chance of being cited in an AI Overview. In our reading, these changes pay off more where the click still exists, that is, on searches without a ready answer. So measure by group of pages, before and after.

According to Google, the title link is often the main piece of information people use to decide which result to click. Generation is fully automated, and Google must crawl the page again to notice a change, which may take a few days or weeks. We recommend four rules:

  • Compare before you edit: review only the titles of queries well below your group’s curve.
  • Change one group at a time: that way, the changes do not mix.
  • Wait for the new crawl: measure the group’s CTR after a few weeks, in the same country and device.
  • Compete for the citation: at Seer, the cited brand got 120% more clicks per impression than the uncited one. Even so, it got 38% fewer than informational searches without a summary.

The full routine for reviewing titles and snippets is in the guide on many impressions and few clicks. Also, if clicks fell along with CTR, start with the diagnosis of an organic traffic drop. Finally, to answer the board, see the analysis of whether SEO is still worth it.

Want to know how your brand shows up in AI answers today?

Book a demo and see how your brand shows up in AI answers, with support from a GEO specialist.

Main sources

  • Search Console Help, Performance report (Search results): Overview and basic setup (2026).
  • Search Console Help, What are impressions, position, and clicks? (2026).
  • SISTRIX, Google CTR study with more than 80 million keywords (2020).
  • First Page Sage, Google CTR by position, with and without an AI answer (2026).
  • Advanced Web Ranking, monthly study of organic CTR by position (2026).
  • Ahrefs, update to the study on the effect of AI Overviews on CTR (2026).
  • Seer Interactive, AI Overview impact on CTR, April update (2026).
  • Amsive, CTR study with AI Overviews, branded and non-branded (2025).
About this articleWritten with AI and reviewed by humans, with verified sources. Found something out of date? Write to the NotHuman team.

Get new NotHuman articles by email

By subscribing, you agree to receive blog articles by email and to the Privacy Policy. Unsubscribe any time.

Keep reading

Demo

Want to know how your brand shows up in AI answers today?

Book a demo and see how your brand shows up in AI answers, with support from a GEO specialist.

Book a demo

Your privacy

We use necessary cookies to make the site work and Google Analytics and Microsoft Clarity statistics cookies to improve our content. You can decline the statistics ones; in the EU, UK and Switzerland they are only turned on after you accept. Cookie Policy.

Cookie preferences

Cookie Policy