The use of artificial intelligence (AI) tools such as ChatGPT, Gemini, or Copilot has created a new source of traffic to websites: visits that arrive via links suggested by these models. To identify and measure this type of traffic in Google Analytics 4 (GA4), you can set up filters (with regular expressions), segments, audiences, or even a custom channel group that groups these sessions under a name like “AI Organic Search” (or whatever you prefer). This will allow you to analyze their impact separately and make better decisions.
You’ve surely already seen that AI chats like Perplexity or ChatGPT itself often cite their information sources, or else recommend links directly. Those links can include campaign parameters like the famous UTM_Source, which are used to identify the source, the medium, or even the keyword that triggered the click. If you’re not yet familiar with using these tags, I recommend reading the article “How to Use Google Analytics UTMs?” where I explain their usefulness and how to apply them step by step.
But what happens if those links don’t include UTMs? That’s where source/medium analysis or capturing the referring domain comes into play. For that, you’ll need to make a few simple tweaks in your GA4 configuration. Nothing complicated, and best of all: it doesn’t break anything. You’ll only be adding visibility into how users arrive from these “smart chats.”
This kind of traffic has similarities with traditional organic traffic, or even with what some are already starting to call Generative Search Traffic. In that sense, if you’re interested in optimizing your content so it’s chosen by language models like ChatGPT, I also invite you to explore the emerging approach to SEO for LLMs, also known as GSO (Generative Search Optimization), which we cover in another article.
And yes, this kind of traffic is growing, so it’s a good time to start tracking it properly.
- You can measure traffic from ChatGPT and other models.
- It’s advisable to group it to better understand its behavior.
¿Ya sabes lo que son las REGEX o Expresiones regulares en Marketing Digital?
Guia de expresiones regulares para Google Analytics
Potencia tu Análisis de datos con Expresiones Regulares: Automatización Eficiente de Análisis de Datos
Importance for Retail and Brands in the Era of Generative AI
AI assistants (ChatGPT, Bing Chat, etc.) are becoming a kind of “new Google” for many users. According to Gartner, by 2026 there could be a 25% decrease in the volume of traditional searches, with more than half of organic traffic shifting toward AI-powered searches (known as Generative Engine Optimization, GEO). It is also estimated that 79% of consumers will use AI search engines next year, and 70% already trust their results.
What does this imply? That more and more customers may ask for product recommendations directly from a conversational AI. If your brand doesn’t show up at that moment, it falls off the radar at a key decision point. Ignoring this trend means putting your brand’s visibility at risk in the new digital environments.
Curso Gratis de ChatGPT
Este curso diseñado en Múltiples artículos por
Martín Garay es un recorrido completo y accesible que te lleva de ser un principiante curioso a un usuario avanzado y consciente de la inteligencia artificial.
6 capítulos, con fundamentos de cómo usar ChatGPT hasta técnicas especializadas para perfeccionar cada interacción. Pasamos por la introducción básica, la optimización de respuestas, las consideraciones de seguridad y precisión, los usos en programación, la creatividad para evitar repeticiones, y finalmente la construcción del prompt perfecto.
What are traffic sources in Google Analytics?
Traffic sources represent the origin through which people access your site. Every time someone enters (from a computer, a mobile phone, or another device), that session is tagged with a specific source. Understanding them helps you spot which channels or campaigns generate more visits and conversions, so we can plan more profitable digital strategies.
Several types of sources usually appear in Google Analytics:
Direct traffic: Refers to visits where the person types your URL directly into the browser or clicks a bookmark (with no referral data).
- Referral traffic: Happens when the user arrives from another website that links to yours (for example, a banner, a blog, or an email).
- Search engines: Includes people who click organic results from Google, Bing, Yahoo, etc.
- Other: Encompasses sources that don’t fit into the previous categories, such as custom campaigns or situations where a clear “source” is not identified.
Nuevas fuentes de trafico: los LLM (Modelos de Lenguaje)
Actualmente surge otro tipo: el tráfico que proviene de modelos de IA (o Large Language Models, LLMs). Estas herramientas, como ChatGPT, Perplexity, chat.mistral.ai, aistudio.google.com, v0.dev o groq.com, pueden mencionar tu sitio en sus respuestas y redirigir usuarios a través de sus enlaces. Antes no se distinguía este canal, así que etiquetarlo de manera específica permite medir su volumen y calidad.
How to view traffic sources in Google Analytics?
In GA4, under Acquisition → Traffic acquisition, you can filter sessions coming from these platforms:
- Click the funnel icon (Build filter).
- Select the “Session source” dimension.
- Choose the “Matches regex” operator.
- Enter the REGEX.
- Apply the change.
REGEX:
.*(gpt|gemini|claude|copilot|llama|mistral|groq|perplexity).*
This way you’ll see only AI traffic and you’ll be able to examine number of visits, bounce rate, and other indicators that matter for your project.
To deepen the analysis, we could (1) add the Secondary Dimension “Landing Page,” so we can see exactly which pieces of content were cited by the AI.
Create an Audience in Google Analytics to measure “AI Source”
Las audiencias te sirven para reunir usuarios que cumplan ciertas condiciones y luego analizar su comportamiento o hacer remarketing.
- Andá a Configure → Audiences.
- Hacé clic en New Audience → Create a custom audience.
- En “Include users when”, indicá:
- Session source (o source)
- matches regex
- Nombrá la audiencia, por ejemplo: “IA Organic Search”.
regex:
.*(gpt|gemini|claude|copilot|llama|mistral|groq|perplexity).*
Crear un Segmento “IA Organic Search”
El funcionamiento de los segmentos es similar, con la diferencia de que podés crear “User segment” o “Session segment” y aplicarlos en Explorations.
- En Explorations, dentro de Segments, hacé clic en “+” → Create segment.
- Elegí “User segment” o “Session segment”.
- Sumá la misma condición (regex) mencionada antes.
- Poné un nombre, por ejemplo: “Segmento – IA Organic Search”.
Custom Channel Group para IA
Los Channel Groups en GA4 permiten categorizar el tráfico en grupos personalizados. Si creás uno para IA, vas a ver este tipo de visitas separado de los demás.
- En Admin → Data Settings → Channel Groups, creá o editá un grupo de canales.
- Añadí un canal llamado “IA Organic Search” y aplicale la regex usada antes.
- Ajustá la prioridad para que se aplique antes que otras reglas.
Recomendaciones
- Actualizar la expresión regular (regex): Si aparecen nuevas plataformas o cambian, modificar al patrón.
las IA suelen cambiar sus entornos y nombres como Google ‘Bard’, a ‘Gemini’ esto demanda que estes alerta de las actualizaciones y mejores tus filtros. - Evaluar la calidad de las visitas: Observá métricas de interacción y conversión. Ademas trata de intensificar para que tipo de respuestas tus contenidos están siendo mas relevantes
- Combinar con otras funciones: Compará el trafico de redes sociales o búsquedas orgánicas en Explorations.
- Ordenar los reportes: Crear un canal específico para IA te permite identificar de forma clara su aporte a la estrategia.
- Aprende a utilizar Expresiones regulares !
En definitiva, con estos pasos podés detectar, agrupar y analizar el tráfico que viene de herramientas de IA, incorporándolo a tu estrategia de Marketing Digital.