Position 1 on Google earned a 5.96% click-through rate across 53 UK sites in 2026, not the 25-30% that industry studies still quote. Position 4 beat it outright at 8.61%, and 74% of all impressions sat at position 11 or worse, where CTR collapsed to under 0.1%. Those are the headline findings from 2,615 query rows of Google Search Console data covering the 90 days to 7 July 2026, pulled from a portfolio of small UK sites ranking for non-brand, long-tail queries. The textbook CTR curve, a smooth slide from 30% at the top to 2% at the bottom of page 1, simply does not describe what small sites experience. This study shows what does.
I run SEO for a portfolio of UK sites and consult on AI search and organic strategy, so this data comes from properties I manage directly, not from a scraped panel. The Cite this study section below explains how to reuse the numbers.
The Full CTR Table: 53 Sites, 2,615 Queries, 90 Days
The table below is the complete dataset: every query row with at least 50 impressions across 53 sites, bucketed by rounded average position. Total volume was 382,941 impressions and 1,667 clicks, a blended CTR of 0.44%.
| Position | CTR | Clicks | Impressions | Query rows |
|---|---|---|---|---|
| 1 | 5.96% | 43 | 721 | 5 |
| 2 | 0.22% | 7 | 3,199 | 6 |
| 3 | 1.58% | 44 | 2,784 | 14 |
| 4 | 8.61% | 377 | 4,381 | 24 |
| 5 | 2.96% | 150 | 5,069 | 37 |
| 6 | 1.59% | 407 | 25,649 | 71 |
| 7 | 1.02% | 262 | 25,582 | 81 |
| 8 | 0.27% | 40 | 14,787 | 60 |
| 9 | 0.75% | 73 | 9,783 | 76 |
| 10 | 0.53% | 45 | 8,514 | 67 |
| 11-20 | 0.25% | 119 | 48,279 | 324 |
| 21+ | 0.04% | 100 | 234,193 | 1,850 |
The curve is not smooth, not monotonic, and nowhere near where the big published studies put it.
Finding 1: Position 1 Earned 5.96%, Not 27%
Position 1 delivered 43 clicks from 721 impressions, a 5.96% CTR, roughly a fifth of the figure most SEO decks still present as normal. Backlinko and Semrush's widely cited historical analysis put position 1 at 27.6%, and Advanced Web Ranking's monthly curves have long shown similar or higher figures for many segments. This dataset says a small site ranking first for non-brand, long-tail queries should expect nothing like that.
The gap is not a measurement error, it is the modern SERP. Before a searcher reaches the first organic listing they pass AI Overviews, featured snippets, local packs, shopping units, and People Also Ask boxes, each of which either answers the query outright or diverts the click. The growth of zero-click behaviour is well documented, and the trend lines in my AI search statistics roundup show the layers above the organic results getting thicker every quarter. Ranking first now means ranking first among the blue links, which is a smaller prize than it was.
One honesty note: only 5 query rows sat at position 1 in this dataset, on 721 impressions. That is a small sample, and I flag it plainly in the methodology below rather than pretending the top of this curve is as robust as the bottom.
Finding 2: Position 4 Beat Position 1
Position 4 earned 8.61% CTR against position 1's 5.96%, a 44% advantage for a ranking three spots lower. Position 2, meanwhile, earned just 0.22%, worse than position 10. The tidy descending staircase in the textbook charts is nowhere in this data.
The explanation is that query intent mix dominates position. The 24 query rows at position 4 happened to be queries where searchers actively wanted a website, so 377 clicks came from 4,381 impressions. The 6 query rows at position 2 included high-impression informational queries that resolved on the SERP itself, so 3,199 impressions produced 7 clicks. Same page 1, wildly different outcomes, and position explains almost none of the difference.
The practical lesson: two rankings at the same position are not worth the same. A position 4 ranking on a query with click intent beat every other bucket in this study, while a position 2 ranking on a zero-click query was close to worthless. When you audit a site, segment queries by whether the SERP leaves any clicks to win before you celebrate or panic about a position.
Finding 3: 74% of Impressions Sat Beyond Page 1 and Earned Almost Nothing
Positions 11 and worse accounted for 282,472 of the 382,941 total impressions, 74% of everything Google showed these sites, and returned 219 clicks, a combined 0.08% CTR. The position 21+ bucket alone held 234,193 impressions, 61% of the total, at 0.04%. Page 3 and beyond is functionally invisible: 1,850 of the 2,615 query rows, 71% of the dataset, lived there and generated 100 clicks between them.
Flip that around and the concentration is stark. Page 1 (positions 1-10) held 26% of impressions but produced 1,448 clicks, 87% of all clicks in the study. Visibility in Google is not a gradient, it is a cliff, and the cliff edge is the bottom of page 1.
This is why "average position" is a trap as a KPI. A site with three page 1 rankings and two hundred page 3 rankings will report a terrible average position while earning all of its traffic from those three queries. The average is dominated by impressions you were never going to convert into clicks. Report clicks and impressions at the query level instead, the same way I break down performance in the UK SEO statistics data and on my own results page.
Finding 4: Positions 6-10 Are Where the Money Is
Positions 6-10 held 84,315 impressions, 22% of the dataset, at a blended CTR of roughly 1%. That is the striking-distance zone: enough existing visibility to prove demand, and a CTR gap wide enough that movement pays off disproportionately. Position 6 alone produced 407 clicks, more than positions 1, 2, and 3 combined (94).
The upside of movement inside this dataset is dramatic. A query moving from position 8 (0.27%) to position 4 (8.61%) would multiply its CTR by more than 30 times. From position 10 (0.53%) to position 4, 16 times. No title tag rewrite, schema tweak, or thumbnail test moves CTR by anything close to that, which is why ranking improvement on queries already at 6-10 should sit above CTR optimisation in your priority list.
If you want to find your own striking-distance queries, filter Search Console for positions 6-10 with meaningful impressions, or run a page through my free website grader to see where the quick wins sit.
Why This Data Disagrees With the Big CTR Studies
Both curves are real, they just measure different internets. The large published studies aggregate millions of queries across sites of every size, and that aggregate is dominated by big brands and brand-name searches. When someone searches "asos" or "hmrc login", position 1 gets clicked at enormous rates because the searcher already chose the destination before typing. Those brand impressions pour into the position 1 bucket and drag the published average up to 27% and beyond.
Strip brand queries out and the picture changes. This dataset is 53 small UK sites ranking almost entirely for non-brand, long-tail, niche queries: comparisons, questions, product research, local intent. Nobody typed these queries looking for one of my sites specifically. Every click had to be won against the SERP furniture and against nine other listings, which is exactly the situation most small-site owners are in.
So when you benchmark your own Search Console data, choose the right yardstick. If your traffic is mostly non-brand, the 5.96% position 1 and 8.61% position 4 figures here will predict your reality far better than a 27.6% headline from a study whose sample is dominated by brands you are not.
Methodology
Source. Google Search Console API, query-level data across 53 UK-managed sites in my portfolio.
Period. 90 days ending 7 July 2026.
Filtering. Only query rows with at least 50 impressions were included, leaving 2,615 query rows. This removes one-impression noise but biases the sample toward queries Google already surfaces regularly.
Bucketing. Each query row was assigned to a position bucket by its rounded average position for the period. Positions 11-20 and 21+ were aggregated because per-position samples get thin beyond page 1.
Anonymisation. Domains and query strings are not published. Aggregates per bucket are published in full in the table above and as JSON in the site repository.
Limitations, stated plainly. The samples above position 5 are small: 5 query rows at position 1, 6 at position 2, 14 at position 3. Treat the exact top-of-curve percentages as indicative, not definitive. The niche mix (affiliate, local service, and informational content) shapes the intent profile, and the data is UK-weighted, so SERP feature density will differ from US-centric studies. Average position is itself a lossy metric, since a query shown at positions 2 and 6 averages to 4. None of these caveats change the direction of the findings, but they should stop anyone quoting "8.61% at position 4" as a universal constant.
Cite This Study
You are welcome to reuse any figure, the full table, or the charts you build from it, commercially or otherwise, with a link to this page as the source: https://sunnypatel.co.uk/blog/google-ctr-study/. Suggested citation: "Google CTR by Position in 2026, Sunny Patel, study of 53 UK sites and 2,615 queries via the Google Search Console API, 90 days ending 7 July 2026." No permission request needed, an attribution link is the only condition. If you spot an error or want the bucketed JSON, get in touch and I will share it.
Frequently Asked Questions
What is a good CTR for position 1 on Google?
For a small site targeting non-brand, long-tail queries, anything above 5% at position 1 is realistic and anything above 10% is strong. In this study, position 1 earned 5.96% across 53 UK sites. The 25-30% figures quoted by large industry studies are inflated by brand queries, where the searcher already intends to click the top result. Judge your CTR against the intent mix of your own queries, not against a global average.
Why is my CTR so low at position 1?
Because position 1 in the organic results is rarely the first thing a searcher sees. AI Overviews, featured snippets, local packs, shopping units, and People Also Ask boxes sit above the first blue link and absorb clicks before your listing gets a chance. Informational queries also increasingly resolve on the results page itself, so a high position earns an impression but no click. Check what the SERP for your query actually looks like before blaming your title tag.
Is average position a useful metric?
Mostly no, and it can actively mislead. In this dataset, 74% of all impressions sat at position 11 or worse and earned under 0.1% CTR, so a handful of high-impression page 3 queries can drag your average position down while your traffic comes entirely from a few page 1 rankings. Track clicks and impressions for individual query and page pairs instead, and treat average position as a rough directional signal at best.
How do I improve CTR from Google search results?
First, move rankings from positions 6-10 into the top 4: in this study that jump multiplied CTR by up to 30 times, which no title rewrite can match. Second, rewrite titles and meta descriptions on pages already in striking distance so they front-load the query and a specific benefit. Third, add structured data where it earns a richer result. Prioritise the queries where you already have impressions, because that is demand you are currently wasting.
