Google is testing a new display of search results with AI

Google once again upsets its search engine with web guide. Therefore, a new presentation that organizes and prioritizes results in different categories using AI.

The advent of generative models such as Chatgpt. Similarly, Copilot, Mistral or Gemini has profoundly transformed our way of looking for information. Therefore, Instead of entering the keywords into a search engine. Nevertheless. Consequently, then consulting the sites most likely to contain the answer, more and more people will ask their favorite chatbot directly. Similarly, More and more Internet users are therefore turning back from anything, search engines.

To counter this user “leak”, Google has decided to integrate artificial intelligence, the magic solution, in Search. However, Last May, the company launched Its new IA mode. Therefore, capable of responding to complex requests in natural language and synthesizing results google testing new display search new (see our article). However. Since July 24, it has google testing new display search been testing another research function, called Web Guide, which organizes links through categories generated by AI.

Web Guide: better hierarchical results

With web guide. Therefore. Google displays at the top of the page the two links he considers to be the most relevant, then a summary generated by AI. Next comes a list of links (educational resources, YouTube videos, Quora questions, Reddit discussions, etc.) grouped into different categories. We therefore move away from the traditional linear classification for a more hierarchical approach to the results. And. unlike AI Overview. which directly displays an answer generated by AI – with the risks of errors that go with -, the links remain central.

If we take the example given by the google testing new display search new firm. a research “How to travel alone in Japan“will bring out links classified under the labels”Complete guide for a google testing new display search solo trip to Japan“. “Solo travel experience story in Japan“or again”Tourist recommendations“. It is then up to the Internet user to choose the category of results which seems to him most relevant. This is a research method called fan-outor “request in fan” in the language of Molière. AI – here a personalized version of Gemini – generates several searches similar to that carried out – as. “Best destinations in Japan to travel alone”. “Security advice for solo travelers in Japan”, “cost of a solo trip to Japan”etc. Additionally, – Before classifying the results in the suitable category.

For the moment. web guide is only available for a limited number google testing new display search new of Internet users. namely those enrolled in the Google Search Labs program, via the web tab in Google Search. The fact remains that these new research methods shake a well-established google testing new display search ecosystem. which includes the media, brand or individual sites, e-commerce actors, sponsorships, etc. So far. each site editor has been trying to optimize its visibility on Google to climb in the search results. capture attention and attract traffic. Today. the web giant seems to question this patient referencing work, at the risk of dropping the attendance of many pages. It will therefore be necessary that all these actors understand how Gemini algorithms work. adapt to different new research modes. And it may not be cake …