How this Search Console CTR analysis works
Click-through rate depends heavily on position: a result at position 2 is clicked several times more often than one at position 8. A CTR that looks poor in isolation may be perfectly normal for its rank, so this analyzer does not just sort by lowest CTR. For every row with enough impressions it looks up the CTR expected at its average position and flags the rows that earn less than half of it, plus any that got impressions but no clicks at all.
The result is the set of pages where your snippet, not your ranking, is the problem. For each one it estimates the clicks missed — impressions multiplied by the gap between expected and actual CTR — so you can fix the biggest gaps first. Rows beyond position 10 are left out by default, because a low CTR on page two is expected.
Why a page can rank well and still get few clicks
- The title does not match what the searcher typed, or is truncated so the useful part is cut off.
- The meta description gives no reason to click, or Google has rewritten it into something unhelpful.
- A featured snippet, AI overview, ad block, video or People Also Ask box sits above the result and answers the question.
- The query is informational and the answer is visible in the results without clicking.
- A competing result has stronger signals such as ratings, price or a recognisable brand.
How to get your Search Console export
Open the Performance report for your property, pick the date range you want, then use Export at the top right and choose Download CSV. Search Console gives you a ZIP that holds one CSV per table — Queries, Pages, Countries, Devices and so on — and you can drop the ZIP straight onto the tool; it finds the tables it needs and ignores the rest.
Exports follow the language of your Search Console interface. The tool recognises the column names in English, German, Spanish, French, Italian, Portuguese, Dutch and Polish exports, and tells you if it cannot find the columns it needs.
What the expected-CTR figures mean
The expected click-through rate comes from a generic curve in line with published industry studies, where position 1 earns roughly a quarter of clicks and the rate falls steeply down page one. Your own site, query types and search features will differ, so the analyzer uses the curve to find outliers and rank them, not to judge any single number.
How to use it
- Export the Pages table (or the Queries table) from the Performance report.
- Drop the CSV or ZIP onto the box, or try the sample data.
- Set the minimum impressions and how far below expectation a row must be to be flagged.
- Rewrite the titles and descriptions of the biggest gaps first, and export the list to CSV.
Frequently asked questions
What is a good CTR in Search Console?
It depends almost entirely on position and query type. Position 1 typically earns a quarter or more of clicks, position 5 around six percent, and position 10 a couple of percent, with brand queries far higher. Compare against your position, not a single number.
How do I improve CTR?
Make the title match the query and lead with the benefit, keep it short enough not to be truncated, write a description that gives a reason to click, and use structured data where it earns rich results such as ratings.
Why does a page have impressions but no clicks?
Usually the snippet is not persuading, a SERP feature is answering the question above it, or its average position is worse than it looks because a few good rankings are averaged with many poor ones.
Does CTR affect rankings?
Google says it does not use Search Console click data as a direct ranking signal, but a better snippet gets more clicks from the position you already have, which is worth having regardless.
Is my Search Console data uploaded anywhere?
No. The file is read and analysed entirely in your browser and is never sent to a server. There is no login, and the only thing the site records is an anonymous count that a file was analysed.
Expected CTRs come from a generic curve; use the list to find outliers, then check each result in the live search page.