How does generative A.I. shape user habits? 🧠
✨ Multimodal: the real revolution in Search?
Searching by text, image or video — generative A.I. promises to change everything. But beyond the buzz around Google and OpenAI, let’s discover how multimodality will actually transform the way we find relevant information.
While many discuss the market share of Google and OpenAI, one topic often remains in the background. Yet it is critical for the future of Search: multimodality and the way users will leverage all the technologies brought by generative A.I. to search for relevant information.
Is Voice Search finally viable?
While Voice Search long remained a private joke among SEO consultants due to its limited application fields and the tools’ inability to properly analyze queries, the latest developments in generative A.I. now make many things possible. ChatGPT’s advanced voice mode now combines powerful capabilities with the ability to use Search to answer questions more accurately. Google, for its part, also offers multimodal tools.
For users and advertisers, the shift lies mainly in how we search. Historically, the user was “forced” to format their query to get a suitable result (i.e., turning natural language into a set of keywords). Voice-based modes completely eliminate this need, and asking a question naturally has never been easier.
This could lead to a complete keyword fragmentation. While it was still possible to group sector-related queries into broad categories (e.g., thousands of different keywords to search for an iPhone 16, which is already impressive), each voice query now has the potential to be unique.
The underlying need behind a query won’t change. However, each request may be more precise and contain more detail, giving advertisers the opportunity to create even more tailored content. The challenge will, as always, be collecting this data anonymously and continuously improving their knowledge of the audience. These increasingly specific queries have been a trend for years, but it could now accelerate. With longer queries, generative AI results will become more accurate — and therefore more relevant — making the technology even more attractive to users!
Video stream for even more interactivity
Whether it’s Project Astra, presented by Google at a Google I/O, or the video stream in ChatGPT’s advanced mode, generative A.I. keeps unlocking increasingly advanced possibilities.
Once again, the potential changes are significant, and the trend is toward ever more customization. Each user will increasingly be able to search for something unique. Of course, the needs and pain points behind each query won’t fundamentally change, but the way queries are formulated will be drastically different.
For advertisers, each medium will gain even more importance. While image search has long benefited e-commerce and specific topics, it’s now easy to imagine that many other sectors could gain visibility by embracing multimodality.
What’s next?
Beyond these solutions, new possibilities continue to emerge. OpenAI has launched its “deep-search mode”, capable of pulling information from many sources and analyzing it in detail to extract the most relevant insights.
Google, on the other hand, introduced Project Mariner, where the A.I. can take control of a user’s computer to perform specific tasks on their behalf (ordering an item, booking a spot, etc.), resembling the rapidly growing AI agents in both popularity and power.
It’s no longer the user who must conform to a platform and its search methods — it’s the artificial intelligence that gives users the power to search in any way possible to find the best answer to their needs.
In such a world, SEO likely finds its place where it always belonged: analyzing users’ needs and pain points, and helping marketing teams build websites that answer those with maximum relevant content about a product, brand, or service. A crossroads between business intelligence, persona development, copywriting, and website optimization.
How to prepare?
As with the other parts of this article series, it’s hard to fully prepare for something that’s constantly evolving and for which we don’t yet have all the information. However, there do seem to be a few constants:
Keep a clear view on user needs
Currently, a good SEO consultant will use search volume data from user queries to understand pain points and needs, and suggest relevant content optimization (for example, adding category pages that match generic product searches like “men’s razor,” “hairdryer,” etc.). In the future, depending on what these tools will offer, the good SEO consultant will either leverage available data or return to more “hands-on” but valuable methods such as interviews. Without this work, it’s impossible to properly optimize a site.
Continue optimizing for conversational content
With Voice Search finally having the potential to become a widely used tool, it’s worth optimizing relevant pages even more in a Q&A format — like in a real conversation. Of course, UX should remain a key factor, but it’s often possible to combine both.
Think beyond just text
Even if this idea isn’t completely new, it’s a good reminder that answers don’t always have to be in text form. Diagrams, plans, videos, and podcasts can sometimes provide just as much value as a long blog post. Depending on the topic, it’s critical to not forget what your users are actually expecting. Intent is everything — and in that regard, nothing new under the sun.
Test, test, and test again
With more tools and possibilities than ever, it’s important not to lose sight of the real user’s daily reality — and to run your own query tests. Putting yourself in the shoes of a real user with a real problem allows you to explore concrete cases and understand what a specific search delivers, and how a website or content should be structured to perfectly meet that demand.
