Sistrix CTR study: how SERP layout changes position-1 clicks

Key takeaways

• Sistrix's study of more than 80 million keywords reports an average click-through rate of 28.5% at position 1, 15.7% at position 2 and 11% at position 3.
• Position 10 averages only 2.5% click-through rate in the Sistrix data, so position 1 earns more than ten times as many clicks.
• Position 1 click-through rate ranges from 46.9% with sitelinks to 13.7% when Google Shopping results appear on the page.
• Sistrix concludes that SERP layout determines how many potential clicks an organic result can get for a keyword.
• A single average CTR curve is a poor forecasting input, because the same position earns very different click shares under different layouts.

The evidence

Sample

80+ million keywords, billions of results, mobile

Hypothesis: Sistrix set out to better understand click rates in Google Search, and the study tests whether a single average click-through rate per position describes real behavior, or whether the layout of the results page changes how many clicks an organic result receives at the same position.

Method: Sistrix combined cleansed Search Console data with information on SERP layouts across more than 80 million keywords and billions of search results. The study covers mobile device data on the first results page. Sistrix says the Search Console data needed intensive cleanup before it was usable, and it compared click rates for pure organic results against results where a SERP feature such as sitelinks, a featured snippet, a knowledge panel, Google Shopping or ads was present. SEO Madman did not run this study; this piece summarizes the method as Sistrix published it.

Limitations: The study was first published on July 14, 2020, and Sistrix marks the page as modified in July 2025, so the absolute numbers predate AI Overviews and other current SERP changes. The page covers mobile data only, Sistrix itself notes the Search Console data required intensive cleanup, and the underlying keyword data is not published, so the figures cannot be independently reproduced. Treat the layout gaps as a directional finding, not a current benchmark.