The rise of AI search experiences: What do you need to know?

By Andy Vale
4 Jun 2025

7 min read

The research, consideration, and conversion phases of a buyer journey have never been static, and marketers have to react to these evolutions of consumer behaviour, especially as we’re now seeing a significant rise in AI search experiences.

A few months back, we noticed certain clients were now getting hundreds of clicks every month from AI chatbots. Whether they were looking for advice, sources, insight, or just holiday planning, they’d been speaking with various LLM providers and following their curiosity onto our clients' websites.

That initial trickle has now turned into thousands of clicks originating from answers in AI chatbots every month for a rising number of clients. In fact, our data from thousands of client accounts shows that traffic from AI chatbots has leapt by a staggering 1037% in the space of a year. So, we took a deep dive into our data to understand the stories behind the numbers, as well as what you can do about it.

What is the biggest source of AI chatbot traffic?

Currently, OpenAI’s ChatGPT is the leading source of chatbot traffic, generating 80% of clicks (up from 76% six months ago). The biggest runners-up are Perplexity and Gemini, Google's offering. An unsurprising winner because of its strong market presence among both tech-savvy early adopters and the layman.

What industries are getting the most traffic from AI search experiences?

So far, businesses in the B2B space have seen the most traffic from AI chatbots, with four out of our top five clients being SaaS businesses, marketing agencies, or research firms. This is likely because many early adopters work in tech (or tech-adjacent) roles and spaces.

But on a broader level, that picture looks set to change. In the B2C sector, the highest-ranking businesses we analysed in the following verticals have seen leaps over the last six months. This is what that looks like for those clients:

Travel - up 175%
Property - up 583%
Healthcare - up 211%
Education - up 113%
Charity - up 232%

The jump in traffic in these sectors (while B2B growth has been much steadier) suggests a growing consumer confidence in using these channels for in-depth discovery and research.

Has ChatGPT traffic caught up with Google?

As a channel for traffic, it’s still early days, and this traffic represents a small percentage of total web visits. Recent data from Similarweb shows there are still 31.8x more visitors to Google every month than ChatGPT, but the latter has already overtaken Bing. Our data tells a similar story, as Google is still a much larger driver of traffic to our clients than all chatbots combined.

Google drew over 30 times more visitors than the next biggest GenAI search destination*

Here's the breakdown of US mobile and desktop web visits by platform:

Note: *gemini.google.com; **chatgpt.com, previously chat.openai.com; ***google.com

* Source

However, in the last six months, we’ve seen clicks from AI chatbots jump up by 41%, showing a steady rise even if the growth is less intense than the initial burst in activity. We also see it being a source for the vast majority (over 90%) of our clients, so there's a good chance that people are already seeking answers related to your space. So a key question is – are you answering them?

Can you track calls that come from AI search experiences?

Yes! In fact, you can track calls that came via clicks from all AI chatbots. These calls are now a daily occurrence for many clients. Because of its distinct nature from other web sources, it’s worth understanding this source as its own unique channel.

You can find guidance on setting up a channel within Infinity here. Before you do this, it’s probably a good idea to round up a list of the URLs for the most widely used chatbots so you can set them as referring domains.

Can you track calls from within AI search experiences?

Alternatively, if you want to track calls directly from chatbots or AI summaries where the caller never visited your site, assign channel-specific fixed numbers to appear in AI-visible listings.

Put that fixed number into an llm.txt file on your site and list it as your contact number; this means chatbots will be more likely to surface that number as the contact number for your business. Then make sure you’re labelling that fixed number with the right source, channel, and campaign within your call tracking platform so your reporting is clear.

For more information on this, make sure you catch our recent webinar with I-COM on how to nail attribution in a zero-click era.

Free webinar

How to track when no one clicks - lead attribution in the AI era

What else can I do with traffic from AI search experiences?

You may not have the same ability to target certain audiences and funnel them in the same way as you do with PPC or your traditional SEO strategy. So, it’s crucial that your value proposition is still clear across your website and sets consistent expectations of your brand.

For example, if you’re a travel company focused on tailored luxury trips, you may receive traffic from people who asked an LLM for advice on a route through a certain country and things to do. But if they’re looking for more of a backpacking, couch-surfing experience (and price point) then they might not be the ideal lead for your sales team to handle as they’d be unlikely to convert.

It might also be a case of you looking at how to make sure each page converts visitors. PPC or SEO landing pages are often optimised to capture leads. But, the way customers discover your digital real estate has just grown massively, so conversion rate optimisation needs to be front of mind on every page that potential customers might find in an AI search experience.

Cover Image: Calls to conversions for marketers

Free resource

Turning calls into conversions

Five steps marketers need to take

Get your eBook

How else can I use AI to optimise my leads?

We’ve all heard how AI models are only as good as the data you feed them. So if you’re using tools like Google’s Performance Max to improve your targeting, you need to consistently feed it high-quality lead and conversion data to guide the algorithms and get more of your ads in front of more high-intent buyers.

If you sell over the phone, you’ll know that that’s where a lot of your biggest leads come from. Often, they tend to have queries they want to clarify before committing to a purchase, especially if it’s a big-ticket item like a holiday or new policy. Smart Outcomes automatically identifies these leads by both their intent and their outcome. This can then be fed back into bidding engines to target similar high-intent, high-relevance individuals.