AI market share and click-through rate - what the data shows

If you've read some of my previous posts on AI's impact on organic search traffic and how to optimise for LLM visibility, you'll know AI is answering more queries, fewer users are clicking through to websites, and visibility in AI results is better understood as a brand signal than a traffic channel.

In this post I'll share some of the latest data that backs that up.

 

Where people are actually searching

The starting point is understanding where AI-assisted search is actually happening - because the market share data most people look at only tells part of the story.

Data shared by StatCounter shows us that ChatGPT has 70% of the UK market share. Copilot and Perplexity sit well behind, with 10% and 8% share respectively. So within the AI tool category, it's a very concentrated market at the moment.

AI tool - UK market share

But this data has a significant blind spot, as it only standalone AI tools. It doesn't capture Google's AI Overviews or AI Mode, because those sit inside Google Search itself. So when someone gets an AI-generated answer on Google, it's recorded as a standard search - not as AI tool usage.

The scale involved makes this significant. In the UK, Google receives roughly twenty users for every one user to ChatGPT:

Google users vs. ChatGPT users

And with AI Overviews now appearing on around 55% of Google queries on average (but nearly all informational searches) it means most AI-assisted search is happening on Google, not ChatGPT - it's just invisible in the headline AI tool market share figures.

So the realistic priority order for visibility is Google's AI features first, ChatGPT second, then Perplexity and Copilot at some distance behind.

 

What click-through rates look like across each

Understanding where searches happen is one thing, but what happens next is where the picture becomes particularly clear.

We know that most AI searches happen on Google. But Google AI Overviews only have a CTR of around 1% according to Pew Research's data. So for the largest AI search platform, 99 out of 100 queries will be answered without clicking through to a website.

Google's AI Mode doesn't have widely published CTR data yet, but its design - being more conversational with fewer links surfaced - would indicate a lower click rate than AI Overviews. Estimates put the CTR at somewhere between 0.1% and 1%, but in most cases likely towards the lower end of that range.

ChatGPT has been measured more directly. Search Engine Land reported a 0.69% CTR across a large dataset recently.

Perplexity is designed to give citations more visibility than other AI engines, which is meant to encourage greater click behaviour. Even so, estimates still put it at somewhere between 0.6% and 1% CTR - broadly similar to ChatGPT despite the more prominent link design:

Perplexity example

Microsoft Copilot sits lower, with early case studies suggesting a 0.1% to 0.3% CTR for brand mentions.

 

What this confirms

Across every platform and feature, the pattern is the same: AI resolves queries. It doesn't route users onward.

As I've written before, this is how these products are designed — and it has a direct implication for how AI visibility should be measured and communicated internally. It's a brand and awareness function. Holding it to a traffic metric it was never going to meet sets the wrong expectation.

Next
Next

What real-world service can teach us about digital personalisation