What We Have Learned From LLM Referral Traffic

The chat is out of the bag! People love sophantic chatbots that give them compliments, instant answers and offer a compassionate shoulder to cry on, and that's the world we live in.

If you thought we had a problem touching grass at the start of the 2020s, the next 5 years pushing towards 2030 are going to see grassy plains without a single human footstep coming to an area near you.

While AI is overhyped and starting to form a significant investment bubble, none of us know when or if it will pop and while the industry remains cash rich, the subsidised usage trend we see today looks to persist.

VCs know that nothing will kill AI interest faster than charging a market rate for token usage and adding a little bit of margin for profit. If that were to be the case, the vast majority of users would be priced out, so for now, we're getting the benefit of cheap subscriptions and generous free packages too.

As AI chatbots like ChatGPT, Claude, and Gemini continue their meteoric rise in user adoption, marketers are rushing to understand how these platforms might reshape web traffic patterns. While there's justified excitement around being featured in LLM responses, the data tells a more nuanced story about what these mentions actually mean for your business.

The Traffic Reality Check

Despite the explosive growth of AI chat platforms, their referral traffic remains surprisingly modest and by modest, I mean it's a joke. Even the deadest of dead email databases could squeeze out more monthly traffic than the LLM for most sites right now.

While ChatGPT, Claude, and Gemini have built substantial user bases—with ChatGPT alone serving hundreds of millions of users—these platforms still generate a fraction of the referral traffic that Google delivers to websites.

The fundamental issue is user behaviour: people tend to be remarkably sticky within these conversational interfaces. When users get comprehensive answers to their queries, they simply don't feel compelled to click through to external sources.

This pattern has held even as platforms have introduced more prominent citations, improved link visibility, and enhanced attribution features. The improvements in citation design haven't translated into proportional increases in click-through behaviour.

Why Users Stay in the Chat

If you thought social media apps closed the floodgates from their newsfeeds and Google was turning the taps tighter with all their rich snippets and referencing Google products, LLMs take platform stickiness to a whole new level.

The stickiness makes intuitive sense.

When someone asks an LLM a question and receives a complete, satisfying answer, the user's immediate need is fulfilled. Unlike traditional search, where users expect to click multiple links to piece together information, LLMs provide synthesised responses that feel like endpoints rather than starting points.

This creates a critical insight for marketers: not all LLM mentions are created equal.

Not All Mentions Drive Clicks

Simply being cited or mentioned in an LLM response doesn't guarantee traffic. In fact, if your content is used to help the LLM fully answer the user's question, you may receive attribution without clicks. The user got what they needed without leaving the conversation. This is particularly true for informational queries.

If someone asks

"What are the health benefits of meditation?"

And receives a thorough answer with your article cited, they have little reason to visit your site—the value has already been extracted and delivered to them.

What Actually Drives Click-Throughs

However, certain types of links perform significantly better within LLM environments:

Custom Tools and Interactive Resources

When an LLM mentions a specialised calculator, design tool, or interactive resource, users are far more likely to click through. These tools require user action and provide value that cannot be replicated within the chat interface.

If you've built a mortgage calculator, recipe generator, or design configurator, mentions in LLM responses can drive meaningful traffic because users need to actually use the tool.

E-Commerce and Product Pages

Product recommendations and shopping-related queries also generate higher click-through rates. When users are in buying mode and an LLM suggests specific products or links to purchase pages, the intent to take action is already present. The LLM serves as a research phase, but the transaction needs to happen elsewhere, creating a natural click-through point.

Service Provider Discovery

Perhaps most interestingly, LLMs appear to confer a credibility premium on the mentioned service providers. When someone asks for recommendations—whether for a plumber, web designer, marketing agency, or consultant—being mentioned by the AI carries implicit endorsement value.

If you thought Google rankings instilled a sense of quality about your business, if the all-knowing, all-wise large language model gods recommend you, the average consumer sees it as the word of Lisan Al-Gaib himself.

Users appear to treat LLM recommendations with heightened trust, viewing them as filtered, vetted suggestions rather than raw search results. This credibility boost can significantly improve conversion rates for service-based businesses. The user arrives at your site with elevated confidence and is more likely to convert compared to traditional search traffic.

LLM Mentions and citations - SEMRUSH

Strategic Implications for Marketers

Understanding these dynamics reshapes how we should think about LLM optimisation:

1. Don't Chase Informational Mentions: If your content serves purely educational purposes, LLM citations may boost your brand visibility but won't necessarily drive traffic. Focus on the brand equity rather than expecting click-throughs.
2. Build Interactive Value: Invest in tools, calculators, and resources that require hands-on interaction. These are natural click magnets within AI conversations because they can't be replicated in chat.
3. Optimise for Consideration Phases: If you sell products or services, focus on being mentioned when users are in discovery or comparison mode. The credibility premium is real and valuable. As long as you are mentioned for those end-of-funnel keywords or rather queries, you're pretty much 97.5% of the way to fully optimised for LLMs. Don't let anyone sell you on a solution that is any different.
4. Think Beyond Traffic: A mention without a click still has value. Brand awareness, implied endorsement, and credibility building happen even when users don't leave the chat. These softer benefits are harder to measure but shouldn't be dismissed.
5. Let your competitors overinvest: There are going to be sites in your niche that think there's an untapped market for users in AI, they'll pay for AI marketing services and tools, and eventually learn that they overpaid for that traffic.

The Slow Road Ahead

As LLMs continue evolving, we'll be sure to see all sorts of weird and wonderful changes, but changes that give us extra traffic will not be one of them; in fact, LLMs are going to push to become even stickier.

Some platforms are experimenting with shopping features, deeper integrations, like flight bookings and even checkout options and new ways to connect users with external resources without ever leaving the chat.

Does this mean even less traffic? Chat, are we cooked? Probably!

So while we wait for the open internet to become closed to more sites, the fundamental tension between comprehensive in-chat answers and click-through incentives persists.

For now, marketers should approach LLM optimisation with realistic expectations. These platforms are growing and influential, but they're not replacing Google's referral traffic anytime soon. Success lies in understanding the specific contexts where LLM mentions drive value—whether that's direct clicks, credibility building, or positioning yourself as the go-to option when users are ready to take action.

As for site owners and business owners, my only warning to you is to tread carefully, there will be a lot of snake oil salesmen and women promising you a rabid horde of customers are hiding behind those chat logs, and to unlock it all, you have to do is pay for their GEO/AEO course, hire them or subscribe to their SAAS.

It's lies. ALL LIES!

And if you disagree, we can fight it out in the comments or take it outside, I'm good either way!

Contact us

If you would like us to improve the performance of your paid search or want to know more about digital marketing for your business, then don’t be shy. We’re happy to assist. Simply contact us

Are you looking to promote your business?

Business owners can create their free business listings on nichemarket. The more information you provide about your business, the easier it will be for your customers to find you online. Registering with nichemarket is easy; all you will need to do is head over to our sign-up form and follow the instructions.

If you require a more detailed guide on how to create your profile or your listing, then we highly recommend you check out the following articles. 

• How to create a nichemarketer profile 
• How to create a nichemarket business listing 

Recommended reading

If you enjoyed this post and have a little extra time to dive deeper down the rabbit hole, why not check out the following posts on SEO and search updates.

• Google Launches Neural Matching Algorithm
• Is Pushing Live Too Many Pages Risky For SEO
• Why Ranking #1 Does Not Always Equal Improved Leads or Sales
• 10 Reasons Why Your Site Doesn't Rank For *Insert Keyword*
• Google Trial More Related Searches On Mobile
• Google Launches Expandable Featured Snippets

Tags: LLM, AI, Chatbot