The New SEO KPIs in the AI Era

For years, SEO has been measured through established metrics: rankings, clicks, organic traffic, CTR, and conversions. Today, with artificial intelligence integrated into search engines and conversational tools, this model is no longer enough. 

When Google displays an AI Overview or a user asks ChatGPT, Perplexity, or Gemini for advice, a brand can be visible even without generating an immediate website visit. That is why talking about SEO KPIs and artificial intelligence means changing perspective: not only measuring who arrives on your site, but also who sees you, cites you, and considers you a source. 

Why organic traffic is no longer the only metric that matters 

The phenomenon is already clear: impressions are growing, while clicks are decreasing. This is the so-called Great Decoupling—the separation between visibility and traffic. 

A page may appear more often in search results, be included in an AI Overview, or contribute to a generative answer while receiving fewer visits. This does not necessarily mean that SEO is performing worse. It means that part of the answer is being consumed before the click. 

This is a crucial point to explain to clients and management. A drop in CTR should not always be interpreted as a loss of relevance. In many cases, it indicates that the user is already finding an answer in the SERP or in an AI environment. Google confirms that site performance in AI features such as AI Overviews and AI Mode is included in the Search Console Performance report under the “Web” search type. 

The issue is that today Search Console does not always provide a separate, immediate view distinguishing traditional organic results from presence within AI experiences. This is why new qualitative and quantitative metrics need to be integrated. 

The new KPIs to monitor in 2025–2026 

Traditional SEO KPIs remain important, but they need to be complemented by indicators designed to measure SEO in the AI era. 

The main KPIs are: 

• AI citation rate: how often the brand or website is cited in answers generated by ChatGPT, Perplexity, Gemini, or AI Overviews.  
• AI brand mentions: how often the brand is mentioned, even without a link.  
• Generative share of voice: how often the brand appears compared with competitors on strategic prompts.
• AI referral traffic: visits coming from AI tools, tracked in GA4.  
• Zero-click queries: queries with high impressions, low CTR, and possible presence of AI-generated answers.  

The point is not to replace rankings, organic sessions, and conversions. Its to read these metrics together. A piece of content may generate less traffic while increasing the brand’s presence in AI-generated answers. And that presence can influence decisions that will convert later, perhaps through another channel. 

How to track AI traffic with GA4 

Tracking generative AI traffic in GA4 is one of the most interesting and still underdeveloped areas. Not everything can be measured perfectly, but a lot can already be organized. 

The first step is to monitor referrals. In GA4, go to Reports > Acquisition > Traffic acquisition and analyze the Session source / medium dimension. Here you may find sources such as chatgpt.com / referral, perplexity.ai / referral, copilot.microsoft.com / referral, or similar domains. Google Analytics classifies traffic into channels through rules, and referral traffic is included among the default channels available in reports. 

The second step is to create a dedicated segment or exploration using a regex that captures the main AI sources: chatgpt|openai|perplexity|copilot|gemini|claude. 

Finally, it is useful to create a custom channel group called “AI Referral,” separating these visits from generic referral traffic. Default channel groups cannot be edited, but GA4 allows users to create custom channel groups. 

Tools for monitoring visibility in AI systems 

Alongside GA4 and Search Console, dedicated tools are needed to monitor AI visibility. 

Semrush has introduced an AI Visibility Toolkit to monitor brand presence, competitors, and relevant prompts in AI search. Ahrefs, with Brand Radar, makes it possible to observe how brands appear in AI answers, as well as across sources such as YouTube and Reddit. There are also specialized platforms like Profound, designed to monitor visibility in generative answer engines. 

In Search Console, on the other hand, the analysis remains more indirect: it is necessary to observe queries with high impressions, declining CTR, stable position, and the possible presence of AI Overviews. This helps distinguish an SEO issue from a change in SERP behavior. 

How to build an SEO report updated for the AI era 

A 2025–2026 SEO report should integrate three levels: organic performance, generative visibility, and business impact. 

The most effective structure is simple: 

• Classic metrics: rankings, clicks, impressions, CTR, organic conversions.  
• AI metrics: citations, AI brand mentions, generative share of voice, AI referral traffic.  
• Strategic interpretation: which content pieces are losing clicks but gaining visibility, which prompts cite competitors, and which sources influence the answers.  

For non-technical stakeholders, the message must be clear: we are not measuring less—we are measuring better. Organic traffic remains fundamental, but it no longer tells the full story of visibility. 

At Mashfrog, this approach combines SEO, GA4, data analysis, and AI visibility expertise to help companies understand not only how much traffic is arriving, but how present the brand is in new decision-making environments. Because in the AI era, the question is no longer only “How many users visited the website?” but “How many times were we part of the answer?”. 

These measurements, however, only truly work when they rest on a solid technical foundation. That is why the topic of new KPIs is closely connected to Technical SEO in the AI era: schema markup, structured data, information accessibility, and readability for generative systems are the starting point for measuring citations, mentions, and visibility in AI engines. 

A holistic vision of artificial intelligence 

In this direction, Mashfrog is increasingly refining its expertise toward an agentic AI as a service business model: the ability to design digital ecosystems that are readable, measurable, and actionable by artificial intelligence systems. Through multidisciplinary teams—including SEO specialists, data analysts, UX experts, AI engineers, and digital governance professionals—we support companies in building environments where data, content, services, and platforms can interact with machine-readable systems. 

The value, therefore, no longer lies only in maintaining a presence on Google or improving a single indicator. It lies in holistically orchestrating skills and technologies to understand how AI systems can read and support a brand. Those who govern this shift do not just measure visibility: they begin to guide it. 

FAQ – User-friendly questions and answers 

How do you track traffic from ChatGPT in GA4? 
You can monitor it in GA4 by analyzing referral sources, such as chatgpt.com / referral, and creating dedicated segments or explorations for AI sources. When the link is controllable, it is useful to add UTMs such as utm_source=chatgpt. 

What replaces CTR as an SEO KPI? 
CTR is not replaced, but complemented by new KPIs such as AI citations, generative brand mentions, share of voice in AI engines, and referral traffic from platforms such as ChatGPT or Perplexity. 

Are AI Overviews visible in Search Console? 
Performance related to Google’s AI features is included in the overall data in the Performance report, but it is not always separately distinguishable from traditional organic results. 

Which KPIs are needed to measure SEO in the AI era? 
In addition to rankings, clicks, and conversions, you need AI citation rate, AI brand mentions, generative share of voice, traffic from AI platforms, and zero-click query analysis. 

Is organic traffic still important? 
Yes, it remains a fundamental metric. However, on its own, it is no longer enough to describe brand visibility when answers are generated directly by artificial intelligence.