For years, "how is our website doing" had a fairly clean answer: open Google Analytics, look at organic search, paid search, referral, direct, social, and you had the picture. That picture is now incomplete, and the gap between what your dashboard shows and what is actually happening is one of the most important measurement problems in marketing right now.
Google just admitted the problem exists
On May 13, 2026, Google added a dedicated "AI Assistant" channel to GA4’s Default Channel Group reports, automatically classifying recognized traffic from tools like ChatGPT, Gemini, and Claude under its own medium and channel, instead of burying it inside generic Referral data. Search Engine Land covered the rollout as a sign that "AI-driven discovery is becoming important enough for Google to treat as its own traffic category," and Search Engine Journal noted that Google described the goal as helping marketers "monitor how generative AI impacts your business by tracking user clicks, trending AI sources, and how this traffic compares to traditional channels like organic search."
That is a genuinely useful update, and if you have not checked it yet, it is worth a look. But it is also a good illustration of just how hard this measurement problem really is, because of what it does not capture:
Google’s own AI Overviews and AI Mode clicks are still counted as Organic Search, not as AI traffic at all. Perplexity, one of the highest intent AI referral sources, is not consistently included in the new channel. And a large share of AI driven visits arrive with no referrer at all, landing silently in Direct, where they are indistinguishable from someone typing your URL directly into a browser.
In short: GA4’s update is a real step forward, but it is a floor, not a ceiling. The traffic you can now see is only part of the traffic AI is actually sending you, and an even smaller part of the influence AI is having on whether people choose you at all.
The traffic picture: AI is growing fast, but search still dominates
Here is why this matters so much right now. Both halves of this picture are true at the same time, and a good content strategy has to hold both in view.
On one hand, AI driven sessions are growing at a pace traditional channels never matched. One industry report tracked AI sourced web sessions growing 527% year over year between early 2024 and early 2025, and Google’s AI Overviews now reach more than 2 billion people every month, expanding from roughly 6.5% to over 20% of Google searches during 2025 alone.
On the other hand, Google still processes an estimated 13.7 billion searches per day and holds about 90% of global search market share, and even ChatGPT’s search like query volume is estimated at only around 12% of Google’s. Ahrefs found that AI Overviews have reduced click through rates for top ranking Google content by 58%, up from 34.5% the year before, which tells you the real shift is not "people stopped searching." It is that fewer searches now end in a click, on Google or anywhere else.
The practical takeaway: you cannot treat AI visibility as a side project, but you also cannot abandon the search fundamentals that still drive the overwhelming majority of discovery. The two are converging, not swapping places.
This is why AEO matters, and why you need to monitor your key themes
This convergence is exactly what Answer Engine Optimization (AEO), sometimes called generative engine optimization or GEO, is responding to. As one industry FAQ put it, SEO is about ranking pages for clicks, while GEO is about being selected as a source in synthesized answers. Those are related goals, but they are not the same goal, and they require different monitoring.
In practice, this means going beyond "what keywords do we rank for" and asking "when someone asks an AI assistant about our category, are we part of the answer." That requires regularly checking how your brand, your competitors, and your core topics show up across ChatGPT, Gemini, Perplexity, and Google’s AI Overviews, not just once, but as an ongoing watch on your key search themes as AI answers shift over time.
How are people actually asking? Think in long form, conversational queries
One of the most useful shifts in mindset is recognizing that the way people phrase questions to an AI assistant is often very different from the two or three word queries they type into a search box, but the research tools for understanding this have been sitting in plain sight for years.
Google’s own keyword intent categories (informational, navigational, commercial, and transactional) and the "People Also Ask" boxes that appear on search results pages are some of the best available windows into how people naturally phrase longer, conversational questions, the exact style of question people now ask AI assistants directly. As one guide to building topic clusters puts it, People Also Ask boxes are "gold for cluster page ideas because they’re literally what searchers want to know", and that is just as true when the "searcher" is a person typing a multi sentence prompt into ChatGPT as when they are typing into Google.
If you are not already mining People Also Ask results, related search suggestions, and long tail question keywords for your core topics, that is one of the highest leverage, lowest cost places to start understanding how your audience is framing questions to agentic engines.
Beyond your own site: social, Reddit, Medium, and G2 matter more than ever
Perhaps the biggest shift AEO requires is recognizing that AI assistants do not only cite your website. They cite the conversation happening around your brand and your category, wherever it lives.
CMSWire reported on Reddit’s rise as a major AI citation source, noting that "Answer Engine Optimization now demands that marketers track and participate in community conversations, not just optimize owned domain content for traditional keyword rankings." Different platforms lean on different sources too: Jasper’s guide to GEO notes that ChatGPT frequently cites Reddit and Wikipedia, while Perplexity emphasizes sources like Reddit, LinkedIn, and G2.
For SaaS companies in particular, this makes review platforms like G2 genuinely strategic, not just a box to check for sales enablement. When someone asks an AI assistant to compare tools in your category, the answer is often shaped by what is publicly said about you on G2, Reddit threads, Medium posts, and similar third party platforms, not just your own marketing pages.
This is also where comparison content and listicles earn their place in a content strategy. "X vs Y" comparisons, "best tools for [use case]" roundups, and structured pros and cons content give AI engines clear, structured signals about where you fit relative to alternatives, the same kind of signal a human researcher would look for when deciding between options.
The constant through all of this: build genuinely useful content clusters
With all of that said, here is the part that has not changed, and will not change. Every credible source on AEO and AI search circles back to the same point: AI systems favor coherent, well organized, genuinely helpful content because that is also what helps real people. As one guide on content clusters and AI search put it, models favour coherent, well sourced material when forming answers, and "disjointed pages might still rank, but they rarely earn consistent citations. In the age of AI search, coherence is currency".
Practically, that means building out content clusters: a strong pillar page on each core topic in your business, supported by focused pieces that each answer one real question your audience has, comparisons, explainers, pricing guides, and use case breakdowns, all linked together so both readers and AI systems can see the full picture of what you know and how you can help.
Measuring AI’s impact on your site will keep getting easier, GA4’s update is proof of that. But the answer to "how do we show up well" has not actually changed. It is still about giving the people (and increasingly, the AI agents acting for them) the most useful, complete, well organized information to make a confident decision about your business.