How to Track LLM Traffic and Measure AI-Driven Customer Acquisition

According to a 2025 Semrush study based on more than 80 million clickstream records, the number of unique domains receiving traffic from ChatGPT grew by 300% between July and December 2024. As AI assistants become an increasingly important discovery channel, marketers are looking for reliable ways to measure their impact on website traffic and customer acquisition 

For marketers and SEO professionals, this raises an obvious question: How much website traffic actually comes from AI?

The answer is more complicated than opening Google Analytics and looking at referral sources. While modern analytics platforms can identify some visits from AI assistants, they can’t capture every customer journey that starts with an LLM.

Someone might discover your business in ChatGPT, remember your brand name, and search for it on Google five minutes later. Another user may copy your website URL from an AI response and type it directly into their browser. Both visitors were influenced by AI, but neither will necessarily appear as LLM traffic in your reports.

That makes LLM traffic tracking different from traditional search attribution.

In this guide, you’ll learn how to track LLM traffic, understand the limitations of current analytics tools, and build a more accurate picture of AI-driven customer acquisition.

Why Is Tracking LLM Traffic Different from Traditional Search?

LLM traffic behaves differently because AI assistants change how people discover and visit websites. Unlike Google Search, where users typically click a search result immediately, AI-powered journeys often include several additional steps before someone reaches your site.

How AI Assistants Influence Customer Journeys

A typical Google Search journey looks like this:

Google Search → Click search result → Website

An AI-powered journey is often much less direct:

ChatGPT → Read recommendation → Remember brand → Search in Google → Visit website

Or:

Gemini → Copy website URL → Open browser → Visit website

In both examples, the customer originally discovered the business through an AI assistant. However, Google Analytics may never attribute that visit to ChatGPT or Gemini.

This changing behavior is why businesses are paying much closer attention to LLM traffic analytics than ever before. That attention is well founded: according to Adobe, web traffic from AI-driven referrals in the United States increased more than tenfold between July 2024 and February 2025, and AI referrals are rapidly closing the gap with traditional channels in conversion rates and revenue per visit. 

This changing behavior is why businesses are paying much closer attention to LLM traffic analytics than ever before.

Why Attribution Becomes More Difficult

Analytics platforms only know what happens immediately before someone lands on your website.

If a visitor clicks a link directly from ChatGPT, the visit may include referral information that identifies the AI platform as the traffic source.

However, many users don’t behave this way. They might:

• remember the company name,
• search for it later,
• type the URL manually,
• return the next day,
• or visit from another device.

Once those extra steps happen, the original AI interaction often disappears from the attribution chain.

This means AI can influence customer acquisition without receiving credit inside your analytics platform.

What Counts as LLM Traffic?

LLM traffic refers to website visits that originate from AI-powered assistants or are influenced by interactions with them. The challenge is that not every AI-assisted visit is recorded the same way.

For practical purposes, marketers should think about three different categories.

Only the first category is easy to identify.

The second and third categories are still AI-influenced customer journeys, but they become nearly impossible to attribute with complete certainty using standard analytics tools.

That’s why LLM traffic tracking should always be viewed as one piece of a much larger measurement strategy.

How to Track Direct LLM Traffic in Google Analytics 4

Google Analytics 4 can identify some direct visits from AI platforms, making it a useful starting point for LLM traffic analytics. While it won’t capture every AI-assisted visit, it helps establish a baseline that you can monitor over time.

1. Open The User Acquisition Report

Navigate to:

Reports → Acquisition → User Acquisition

Select a consistent date range, such as the last 28 days, and compare it with the previous period. Looking at trends instead of isolated numbers makes it easier to spot changes in AI-driven visits.

If your website receives relatively little AI traffic, consider comparing the last three months instead of only four weeks. Larger datasets often reveal patterns that shorter periods may hide.

2. Filter Traffic Sources Associated with AI Platforms

Create a traffic source filter that includes popular AI assistants.

Common platforms include:

• ChatGPT
• Google Gemini
• Microsoft Copilot
• Perplexity
• Claude
• Other AI-powered assistants that pass referral information

This allows GA4 to display sessions that originate directly from these sources.

As new AI products enter the market, it’s worth reviewing your filter regularly to ensure it reflects the platforms your audience actually uses. That filter list changes fast: according to Digiday, ChatGPT referrals grew 52% year over year between September and November 2025, while Gemini referral traffic jumped 388% over the same period, driven partly by a surge in Gemini’s own user base. 

3. Compare Changes Over Time

Looking at a single month’s traffic rarely tells the whole story.

Instead, monitor:

• total AI referral sessions,
• month-over-month growth,
• engagement metrics,
• conversions,
• assisted conversions from AI referrals.

Even if the numbers are still relatively small, upward trends often indicate growing AI visibility.

For many businesses, LLM traffic is still an emerging acquisition channel. Tracking changes over several months provides much more meaningful insights than evaluating one reporting period in isolation.

What Can GA4 Actually Measure?

Google Analytics works well for identifying direct referral traffic from AI assistants.

It can answer questions like:

• How many visitors clicked directly from ChatGPT?
• Is AI referral traffic increasing?
• Which AI platforms send the most visitors?
• How do AI referrals convert compared to other channels?

These insights are valuable, but they represent only one part of the overall picture.

These insights are valuable, but they represent only one part of the overall picture. Peer-reviewed research published in Marketing Science found that, roughly a year after ChatGPT’s referral traffic began, it still accounted for less than 0.2% of total site traffic in the study’s dataset, about 200 times smaller than Google’s organic search, which is a reminder that referral counts alone will understate how much AI is actually shaping discovery. 

Why Doesn’t Google Analytics Show the Full Picture?

Google Analytics can only attribute visits based on the information it receives at the moment someone lands on your website. If the connection between an AI assistant and the visit is broken, GA4 has no reliable way to know that AI influenced the user’s decision.

This is one of the biggest challenges in LLM traffic analytics today.

AI-assisted Journeys Often Continue Outside The AI Platform

Many marketers assume every AI-driven visit will appear as referral traffic. In reality, that’s rarely how people behave.

Consider these common scenarios:

In every example except the first, the user originally discovered the brand through an AI assistant. However, Google Analytics no longer has enough information to connect that visit to the original AI interaction.

This is why how to track LLM traffic is a more complex question than simply filtering referral sources.

Direct AI Referrals Represent Only Part of The Story

Businesses should treat AI referrals as a useful signal rather than a complete measurement of AI-driven acquisition.

In our experience, direct referrals may represent only a small percentage, sometimes around 5%, of the total traffic influenced by LLMs. The remaining visits are often attributed to Direct or Organic Search because users continue their journey outside the AI platform.

That doesn’t mean AI failed to generate those visitors. It simply reflects the limitations of today’s attribution models.

If AI introduces someone to your brand, but they later search for you on Google, AI still influenced the conversion, even if analytics doesn’t show it.

What Are the Limitations of LLM Traffic Tracking?

No analytics platform can currently capture every AI-assisted customer journey. Understanding these limitations helps you interpret your reports more accurately and avoid drawing the wrong conclusions.

Attribution Is Limited To The Last Identifiable Source

Google Analytics only records the traffic source that can be identified when a visitor reaches your site.

If someone:

• copies your URL,
• types your domain manually,
• switches devices,
• shares the recommendation with a colleague,
• or comes back several days later,

the original AI interaction is usually lost.

This isn’t unique to AI. Similar attribution challenges have existed for email, podcasts, offline advertising, and word-of-mouth marketing for years.

AI Adoption Is Changing User Behavior

Large language models encourage users to research differently.

Instead of clicking multiple search results, people often:

• compare brands inside ChatGPT,
• ask follow-up questions,
• narrow their options,
• then visit only one or two websites.

This means AI increasingly influences purchasing decisions long before someone appears in your analytics platform.

As AI adoption grows, marketers should expect attribution gaps to become more common rather than less.

How Can You Measure AI Visibility Beyond Traffic?

Traffic is only one indicator of AI performance. To understand how visible your brand is across LLMs, you need to monitor additional signals alongside referral visits.

Looking at multiple metrics provides a much more realistic picture of AI-driven customer acquisition.

Monitor AI Visibility

Ask the major AI assistants questions your customers are likely to ask.

For example:

• Best CRM software for startups
• Top project management tools
• Best SEO agency for SaaS companies

Track whether your brand:

• appears in recommendations,
• is mentioned by name,
• receives citations,
• or isn’t included at all.

This type of monitoring helps identify visibility opportunities before they translate into measurable traffic.

Track Branded Search Growth

Branded searches often increase as AI becomes a stronger discovery channel.

Someone may first encounter your company in ChatGPT, then search your business name on Google. Although Google Analytics records this as Organic Search, the original discovery happened inside an AI assistant.

Monitoring branded search volume alongside LLM traffic helps identify this pattern.

Watch Direct Traffic Trends

A steady increase in Direct traffic can also indicate growing AI influence, especially if other acquisition channels remain relatively stable.

While Direct traffic includes many different visitor types, it may also contain users who:

• copied your website URL,
• bookmarked your site after seeing it in AI,
• or returned after an earlier AI recommendation.

Direct traffic alone doesn’t prove AI influence, but it becomes a valuable supporting metric when analyzed alongside AI referrals and branded search.

Combine Multiple Signals Instead of Relying on One Report

No single metric tells the entire story. The most reliable approach combines several indicators.

Looking at these metrics together creates a much more complete understanding of AI-driven customer acquisition than referral traffic alone.

What Are the Best Practices for LLM Traffic Analytics?

Successful LLM traffic tracking focuses on trends, not perfect attribution. Since no analytics platform can capture every AI-assisted interaction, the goal is to combine different data sources and monitor changes over time.

Here are a few practical recommendations:

1. Track AI referral traffic consistently. Use Google Analytics 4 to monitor visits coming directly from AI assistants and compare performance month over month.
2. Measure more than sessions. Pay attention to engagement, conversions, and assisted conversions to understand whether AI traffic delivers business value.
3. Monitor branded search volume. Rising branded searches may indicate that more users are discovering your business through AI before searching for you directly.
4. Review Direct traffic trends. Unexpected growth in Direct traffic may support other indicators that AI visibility is improving.
5. Regularly evaluate your AI visibility. Check whether ChatGPT, Gemini, Copilot, Perplexity, and other assistants recommend your brand for relevant queries.
6. Avoid relying on a single KPI. AI-driven customer acquisition is still evolving, and measuring it requires multiple complementary signals rather than one report.

LLM Traffic Is Measurable, But It Is Not Fully Attributable

Google Analytics 4 can show visitors who click directly from AI platforms like ChatGPT, Gemini, or Perplexity. However, those referrals represent only part of the customer journey. Many users discover your brand through AI, then search for your business later, type your URL manually, or return another day. In these cases, the visit is usually attributed to Organic Search or Direct traffic instead of the AI platform that influenced the decision.

The most effective way to measure AI-driven customer acquisition is to combine multiple signals. AI referral traffic, branded search growth, Direct traffic trends, conversions, and AI visibility together provide a much more accurate view of AI’s impact than any single report.

As AI assistants continue to reshape how people discover information online, marketers who understand these attribution gaps will be better equipped to evaluate performance, optimize content, and make smarter SEO decisions.

Ready to Improve Your AI Visibility?

If you want to understand how your business appears in AI search experiences and identify new opportunities for growth, our team can help.

Book a free AI Visibility Audit and discover how your brand performs across ChatGPT, Gemini, Perplexity, and other leading AI platforms.

FAQ