Tag: AI traffic

  • How to Track AI Traffic in GA4: A 15-Minute Check (and What It Can’t See)

    How to Track AI Traffic in GA4: A 15-Minute Check (and What It Can’t See)

    If you have no idea whether AI tools are sending people to your website, start with a 15-minute check in Google Analytics 4, then fill the gaps it leaves with three sources GA4 can’t reach.

    Updated 11 November 2026.

    Clients have been asking us some version of this question for about a year, and the wording has moved from “is AI reading my site?” to “how do I know, and how do I track it?” That is a healthy change, because it means AI search has become a channel owners want to see in the monthly report, next to organic, paid and referral.

    This guide starts where most small B2B businesses already have data: GA4. It walks through the built-in AI Assistant channel, the reasons it can undercount, a custom channel group that recovers your history, and the things GA4 cannot see at all. It is the practical companion to our plain-English guide to AEO and our piece on how AI crawlers read your website. After reading, try blocking out a quarter of an hour before your next marketing meeting, opening GA4 and working through the four steps, so you walk in knowing whether AI is a channel worth watching for your business.

    How much AI traffic should a small B2B site expect?

    For almost every website we look at, sessions from AI assistants are still a small share of the whole. If you run ads, or have strong organic or social traffic, visits that begin in ChatGPT or Perplexity will sit low in your channel report. The owners who care most are looking for a new channel, or have watched organic search soften and want to know where the readers went. For them, a small channel that is growing is exactly the pattern worth tracking early.

    What AI traffic looks like on our own site. In the twelve months to September 2026, about one in twelve sessions on contentifai.agency arrived from an AI assistant. ChatGPT sent the most, then Perplexity and Gemini, with a handful from Claude, and around half were engaged sessions. We don’t forecast from a channel this size. We watch its share each month.

    The word “sessions” is doing careful work in that box, because most of what AI companies do to a website involves no person at all. Cloudflare’s analysis of AI bot activity across its network found that training accounted for nearly 80% of AI crawling by July 2025, with search indexing at about 17% and user-triggered requests at roughly 3% (Cloudflare, 2025). The traffic you can see in analytics is that last, thin category: a person asked a question, the tool fetched or cited your page, and the person clicked. Everything else is machines reading, and that belongs to the crawler conversation rather than this one.

    The 15-minute AI traffic check in GA4

    Back in 2025 our answer to “can this be tracked?” was yes, with some work, and a prediction that the big platforms would bring it in-house once they had settled the data streams and the privacy questions. Google has now done that in GA4, which makes it the right place to start. You need Viewer access to your GA4 property for steps 1, 2 and 4, and Editor access for step 3. The first two steps take five minutes, the channel group takes five and the exploration takes the last five.

    Step 1: Open the AI Assistant channel (2 minutes)

    Since 13 May 2026, GA4 has sorted visits from recognised AI tools into their own channel. Google’s release notes describe a dedicated way to measure traffic from popular AI assistants, and its default channel group definitions describe AI Assistant as visits from sources like ChatGPT, Gemini, DeepSeek, Copilot or Grok, with the medium set to “ai-assistant” (Google, 2026). Go to Reports, then Acquisition, then Traffic acquisition, set the date range to the last 90 days, and look for the AI Assistant row beside Organic Search and Referral.

    Step 2: Note the share, the engagement and the tools (3 minutes)

    Write down two numbers: AI Assistant sessions as a share of all sessions, and the channel’s engagement rate against Organic Search. Then add Session source as a secondary dimension, so each tool shows by name. If you want to track Grok, ChatGPT or Perplexity traffic separately, this is the view that splits them.

    Why GA4’s built-in AI channel can undercount

    Before you treat that number as the full picture, check it against four gaps we see on client sites and our own.

    • It starts in May 2026. On our own site, AI visits only began appearing under AI Assistant in June 2026, and earlier ones sit under Referral.
    • It depends on a referrer. Some AI apps open links without passing one, so those visits land in Direct, and no GA4 setting can recover them.
    • Google’s list can lag new tools. A newer assistant may not be on the list yet, and its visits stay in Referral until it is added.
    • It leaves out Google’s own AI answers. Google’s channel definitions exclude AI Overviews and AI Mode, so clicks from them count as Organic Search.

    Steps 3 and 4 close the first and third gaps. The other two need the sources further down.

    Step 3: Build a custom AI channel group with a regex (5 minutes)

    A custom channel group lets you define your own AI channel and, according to Google’s custom channel groups guidance, apply it to your reports retroactively (Google, 2026). That gives you the history the built-in channel lacks.

    1. Go to Admin, then Data display, then Channel groups, and select Create new channel group. GA4 copies the default group for you to edit.
    2. Name it “AI traffic check”, add a new channel called “AI tools”, and set the condition to Session source, “matches regex”, with the expression below.
    3. Drag “AI tools” above Referral. GA4 places each visit in the first channel it matches, so order matters.
    4. Save, return to Traffic acquisition and switch the report to your new channel group.
    .*(chatgpt|openai|perplexity|gemini\.google|bard\.google|copilot|claude\.ai|grok|deepseek|meta\.ai|mistral|you\.com|poe\.com).*

    Two details save a lot of head-scratching. Google’s regex guidance says GA4 matches the full value and is case sensitive by default, which is why the expression starts and ends with “.*” and stays in lower case (Google, 2026). Standard GA4 properties allow two custom channel groups, so name this one clearly. Check the source list in step 2 once a quarter and add any new tool to the expression.

    Step 4: Save an exploration for page-level detail (5 minutes)

    Go to Explore and start a Free form exploration. Add Session source and Landing page as dimensions, Sessions and Engaged sessions as metrics, and a filter on Session source that matches the same regex. The result shows which pages AI tools send people to, and once you save it, next month’s check takes a minute.

    What GA4 can’t see

    GA4 counts people who clicked through to your site with a referrer intact. Several kinds of AI activity never reach it, and knowing which ones keeps your numbers in proportion.

    • Clicks from AI Overviews and AI Mode, which arrive looking like ordinary Google searches.
    • Visits with no referrer, which land in Direct.
    • AI crawlers, which fetch pages without running your analytics tag, so they appear in server logs and never in GA4.

    The three questions no tool answers yet

    With GA4 and the sources below running, you will know that AI tools read your site and roughly how many people they send. Three things will still be missing, and no tool on the market reliably fills the gap.

    What the person asked. Search Console’s AI report holds back the queries, and GA4 never had them. You can see that someone came from ChatGPT; you cannot see whether they asked “best content agencies in the North East” or “is schema markup worth it”, which are very different visits.

    What the tool said about you. An impression or a referral tells you a link was shown. It does not tell you whether the answer was accurate, whether you were the recommendation or the fourth name in a list of six, or whether the tool used your page to back up a fact in an answer about something unrelated to your services. Being a credible reference is good for you, but it is a different thing from being chosen.

    What happened next. The jump from “read an AI answer” to “became a client” is the one every marketer wants to measure, and it is still invisible. Appearing in AI search, being recommended by it and being bought from are three separate steps.

    This is why we do not hand clients an ROI formula for AI search: a reliable one does not exist yet. What we can say is simpler. If your buyers are asking AI tools questions in your category, you will want to be in the answers, and the sensible measure for now is whether your share of that channel grows month on month.

    Beyond GA4: three sources that fill the gaps

    GA4 tells you who arrived. These three sources tell you where Google’s AI showed your pages, which bots came to read them and what your buyers say themselves.

    Search Console: the generative AI performance report

    Clicks from AI Overviews and AI Mode count as Google search visits in GA4, so analytics cannot tell you how often Google’s own AI used your pages. Search Console now can, up to a point. The generative AI performance report shows how many times links to your site appeared in AI Overviews and AI Mode, by page, country, device and date, and Google rolled it out to every site worldwide on 31 August 2026 (Google, 2026). You will find it under Performance. It gives impressions only, with no queries, clicks or position, so treat it as a presence measure for seeing which pages Google’s AI draws on.

    Server or CDN logs: which bots came, and why

    Analytics counts people, whereas logs count requests, and the logs are where you see which AI companies are reading your site and why. The major providers publish separate user agents for training crawlers, search-index crawlers and fetches made for a live user, and a CDN such as Cloudflare groups them for you. A quarterly look is enough. Our AI crawlers post covers the named bots.

    Your forms: ask people where they found you

    Every method above depends on a referrer surviving the journey, and many will not. The one source you control is the person. Add “an AI assistant (ChatGPT, Gemini, Perplexity or similar)” as an option to every “how did you hear about us?” field you have: the contact form, the discovery call notes, the onboarding questionnaire. Ask at the first opportunity, because the enquiry that never converts still tells you the channel is working.

    Want AI traffic measured for you each month?

    This is a Contentifai service. Measure is our monthly AI visibility and web performance report: we review your Google Analytics, Search Console and presence in AI answers, explain what changed in plain English, and give you three prioritised recommendations. From £200 a month on a 28-day rolling agreement.

    Four mistakes that skew AI traffic numbers

    Plenty of businesses will want to run this in-house, and that is a good choice when someone has the time to look at the numbers each month. The GA4 check takes a quarter of an hour, and the other three sources an afternoon at most. The harder part is reading the results without fooling yourself, and over the past year we have watched the same four mistakes make AI traffic look bigger, or smaller, than it is.

    • Counting crawler hits as visitors. Server logs record every request a bot makes, so a single training crawl can look like hundreds of “AI visits” if it is added to the analytics total.
    • Reading Search Console AI impressions as clicks. The report contains no clicks, so an impression count is a presence figure and nothing more.
    • Waiting until the channel is big enough to matter before tracking it. Leave it until then and you will have no baseline to compare against when it does matter.
    • Relying on the built-in AI Assistant channel alone. It misses history, stripped referrers and newer tools, and a form field would have caught what analytics dropped.

    Avoid those four and the numbers you collect will be modest but trustworthy, which is the right footing for a channel still finding its size.

    Five lines for your monthly AI traffic report

    Once the tracking is in place, keep the monthly view to five lines, whether you write the report for yourself or ask an agency to produce it for you.

    1. Share of sessions from AI tools, using your custom channel group, and whether it is rising or falling.
    2. Engaged-session rate for that channel against organic, because a small channel that engages well is worth more than its size suggests.
    3. AI impressions by page in Search Console, to see which content Google’s AI is drawing on.
    4. The count of “AI assistant” answers on your forms, your one source that does not depend on a referrer.
    5. The pages AI tools send people to, from your saved exploration, which tells you what to write more of.

    That last line is where measurement connects back to the work. Tracking is the easy part now that Google has built most of it. What earns the citations is still the content: specific, useful, written so a machine can quote it, on a site the crawlers can read. Measurement tells you whether that work is paying off, and it cannot do the work for you.


    Tracking AI traffic: what we covered

    This guide covered a 15-minute check in GA4, the four reasons its AI Assistant channel can undercount, a custom channel group and exploration that recover your history, and the three sources that fill the gaps GA4 leaves.

    Google has changed its AI reporting twice in 2026 already, and the menu paths above may move again, so treat the method as the durable part and the clicks as the part to check.

    If you would rather have this measured and explained each month, our AI visibility and web performance report does it from £200 a month, and Monthly Blog Content writes the pages AI tools cite. Contact us to get started.

    Frequently asked questions about tracking AI traffic

    The questions owners and marketing managers ask us most about measuring AI search.

    How do I track AI traffic in Google Analytics 4?

    Open Reports, then Acquisition, then Traffic acquisition, and look for the AI Assistant channel, which GA4 has assigned automatically since 13 May 2026. For history before then, and for newer tools, create a custom channel group with a regex on Session source and apply it to your reports.

    Why does GA4 show so little AI traffic?

    Partly because the channel is still small for most sites, and partly because GA4 undercounts it: the AI Assistant channel starts in May 2026, some apps strip the referrer so visits land in Direct, and Google’s list can lag new tools.

    Can I track Grok, ChatGPT and Perplexity traffic separately?

    Yes. In the Traffic acquisition report, add Session source as a secondary dimension and each AI tool shows by name, for example chatgpt.com, perplexity.ai or grok.com. A saved exploration filtered with a regex keeps the view for next month.

    Does GA4 show traffic from Google AI Overviews and AI Mode?

    No. Google excludes AI Overviews and AI Mode from the AI Assistant channel, so those clicks appear as Organic Search. Search Console’s generative AI performance report shows how often your pages appeared in them, as impressions only.

    We have no idea if AI is sending us traffic. Where do we start?

    Start with GA4’s AI Assistant channel and note its share of sessions over the last 90 days. Add a custom channel group to recover earlier visits, then add an AI option to your “how did you hear about us?” form fields.

    What is the difference between an AI crawler visit and AI traffic?

    A crawler visit is a machine fetching your pages to train a model or build an index, and it shows in server logs, never in analytics. AI traffic is a person who asked an AI tool a question and clicked through, and it shows in GA4 as a session from an AI source.

GDPR Cookie Consent with Real Cookie Banner