Less visits, more time for permanence, more rebound rate: this is traffic from the Ia ecosystem
We have been talking about the impact that the conversational tools ecosystem such as Chatgpt, Perplexity or Gemini have on websites. A crucial issue for any company, and of which we have analyzed different studies, with different conclusions. Even referents of the digital sector as Sam Altman or Pichai SUCK They have repeatedly talked about it.
But the field of artificial intelligence evolves (very) quickly, and every day new studies appear that allow us to understand how the ecosystem is changing. The last one is the work of Ahrefs, One of the world references in SEO tools, which has shed light on an emerging phenomenon: The traffic generated by AI platforms.
And his conclusions reveal a surprisingly different user behavior pattern from traditional search traffic.
Traffic from IAS is different from the “traditional”
The study, which analyzed the traffic of approximately 82,000 websites between May and June 2025, compared key behavioral metrics between search traffic from Google, Bing, etc. (including search functions with Ia like Google Ai Overviews), traffic of pure ia (the one from Chatgpt, for example) and general traffic. The results are revealing:
Less pages per visit: Users who arrive from AI platforms visit, on average, 4 pages. This involves 1.2 pages less than visitors from search engines and 1.5 pages less than the general average.
Lower navigation depth: These users also visit less pages per session, which indicates that they do not “curly” so much for the site. The users of AI visit 2.27 pages per session, compared to 2.79 of users from search engines and 2.99 of the general average.
More time on the site, but concentrated: Despite navigating less pages, IA visitors spend more time on site (about eight more seconds). That is, its time is concentrated in a few specific pieces of content.
Ahrefs's interpretation is that IA users have a clearer intention. These users have carried out their previous research through the same AI and arrive at the site with very defined objectives and an accurate idea of what they are looking for, eliminating the need to explore more.
The bouncing paradox and conversions
One of the most striking data in the study is the rebound rate. IA visitors have A major rebound rate (67.8%) that search users (63.7%), and that the general average (62.4%). This data could be interpreted as a sign of greater demand; If the content does not meet your expectations immediately, they leave.
However, here lies the paradox and the most relevant point for business: in a previous study, conducted from data from its own website, Ahrefs observed that, despite the greatest rebound, AI visitors convert a significantly higher rate. In their particular case, conversions on their own website from AI traffic were 23 times greater than those of traditional search visitors.
This is a striking fact and that You have to take with tweezerssince while the rest of the conclusions mentioned refer to the analysis of 82,000 websites, the conversions data proceeds exclusively from the particular case of Ahrefs.com. In any case, a high rebound combined with high conversions could indicate a “once” behavior: the Ahrefs user arrives, quickly finds what he needs (or validates information), performs the desired action (conversion) and leaves.
Towards a new type of user
Ahrefs findings emphasize that AI traffic is not intrinsically “better” or “worse” than search traffic, but simply different. Its behavior points to users with a very specific intention and that possibly They are in a more advanced stage of the conversion funnel.
For content creators and marketing professionals, this implies the need to adapt strategies:
Ultra-Relevant and Concise content: The content must be direct, precise and quickly satisfy the need of the AI user to avoid the rebound.
Optimization for conversion: Given its high conversion potential, it is crucial to optimize the experience on the page to facilitate the desired action (purchase, registration, download, etc.).
Redefine the “quality” of traffic: Traditional engagement metrics may not be the only one or the best indicator of traffic quality from AI. True value could reside in the effectiveness of conversions, even with less duration or depth interactions.
Image: Chatgpt
