Distribution of unique keywords in the sample by length (words)

[sharminakter]

Computer data

Exploratory analysis
We didn’t use any additional filters for our keywords, and we assume that our sample is representative of the overall distribution of keywords. Let’s take a look at some of the key findings from this sample:

Distribution of unique keywords in the sample by length (words) / Semrush Zero-Click Study
Here we see the distribution of search frequency for keywords of certain lengths, with 3-4 word searches leading the way. 1-2 word searches follow closely behind, showing that the vast majority of searches are relatively short. This distribution seems quite normal for the overall distribution of keyword length; in fact, we looked at similar metrics in our State of Search report earlier this year.

In the graph below, you can see the breakdown of how long it takes a user to make a decision about their next step after the first search. As stated in the methodology, this time is limited to two minutes to avoid suspicious data coming from bots, or extended pauses that could actually come from another session.

Distribution of the number of seconds it takes users to make a decision after the first search
Distribution of the number of seconds it takes for users to israel telegram data make a decision after the first search / Semrush study on zero-clicks
If we look at how long it takes people to interpret and act on their search results, we see that their decisions are generally very quick. In the vast majority of cases, it takes less than 15 seconds to decide what to click on.

This data highlights the importance of the visual structure of SERPs. Internet users do not spend much time analyzing the results, but make quick decisions to get what they want.

Logically, additions like image thumbnails or expandable tabs, like those that appear with an FAQ rich result, influence how people interact with the results. And that’s before you even consider highly visual SERP features like image packs.