2025 is the year of AI. Google has launched its AI overviews, which provide a synopsis of data and content found on related websites. Like the Knowledge Panel, People Also Ask, and Featured Snippet sections, Google’s AI overview allows users to click on the websites from which Google sourced its answers from. This has led many webmasters and SEO professionals to ask, “Where can I see my AI traffic in Google Search Console or Google Analytics?” As of June 2025, the answer to this question is that Google is not separating AI Overview traffic from traditional search traffic. But that doesn’t mean there aren’t some workaround options.
How To Use Tools Like SEMRush To Track SERP Features
There are a variety of SEO tools on the market, including SEMRush, that provide a holistic view of how a domain is ranking on Google. This overview includes providing estimations for how many searches your website received rich results for. Rich results are any of the SERP features that differ from true organic search listings. For example, featured snippets, which sometimes appear at the top of a SERP, are displayed when there is a relatively one-to-one answer to a question. Knowledge panels are usually sidebars in Google that provide an itemized list of attributes for a specific topic (for example, if you asked Google “How many people live in Boston,” you might get a Knowledge Panel for the city that includes population, year established, etc).
AI overviews are considered a SERP feature just like knowledge panels and featured snippets. In tools like SEMRush. You can filter your organic keywords to show just the instances where your website appeared in an AI Overview.
Using A Filtered AI Search Audience in Google Analytics 4
Many SEO professionals are looking to track not just AI overviews but third-party AI sites like ChatGPT. Both ChatGPT and AI Overview use search-based methods. Originally, many LLMs like Chat GPT had a training method, meaning their sites were fed a dataset that they then used to answer user questions. This led to potential concerns that data would be incomplete or out of date. To fix this, Chat GPT (like Google) now uses a search-based method. Basically, when you ask ChatGPT a question, it makes searches on search engines, reads through the top-ranking results, and compiles its answer. Because of this method, both Google’s Gemini and ChatGPT can frequently provide links to their users to the sites they used to source their responses.
Unfortunately, there are currently no features in Google Analytics 4 to look at your AI search traffic separately from overall organic traffic. Worse still, technically speaking, these traffic sources are already in separate camps. Users who get to your site from AI chatbots like ChatGPT would be marked as referral traffic, while users who get to your site from Google’s AI Overviews would be marked as Organic traffic. That said, there is a workaround. In Google Analytics 4, you can filter the traffic in a report by either using prebuilt audiences (all users, organic traffic, etc.) or by building your own filter.
You can build a filter that uses a regular expression like this:
Session source, matches regex, ^.*ai|.*\.openai.*|.*copilot.*|.*chatgpt.*|.*gemini.*|.*gpt.*
This would force the audience to be limited to users who came from those sites. Of course, this is a very manual method, so it is essential that you track all of the AI sites you want to look for.
How Does AI-Overview Traffic Differ From Traditional Search
AI-search traffic is relatively new compared to traditional search traffic. That said, people are noticing some unique differences. AI search traffic tends to spend longer on websites. The leading theory here is that the people who are searching on AI chat sites or via AI overview are most likely in the research phase of their journey. And because the AI chatbots and AI search are supposed to answer their question, the only reason they’d click on a supporting link is if they essentially wanted to learn more or get more context for their question. In those cases, AI search might give an end user a brief synopsis, but then they would likely want to read further from a human-generated source. The type of user who is looking for a quick answer would likely not click on the supporting links. This leads to an ecosystem where AI-search provides a low number of users, specifically a user type that is looking for explanations, more details, and potentially an interaction with a human creator.
Optimize Your Online Presence For Traditional Search and AI
Boston Web Marketing provides expert SEO services for local businesses that help them show up in traditional search, AI overviews, and more. To get started with our team, please contact us for a free website audit.