These days there is a gesture that is repeated over and over again: open a chatbot or a generative search mode, write a question and wait for a direct, orderly and apparently definitive answer. There is no list of links and no need to compare ten pages to decide which one to trust. The promise of comfort is evident, but behind that everyday gesture a much deeper crack is opening up. For years, internet search has been one of the tech industry’s big money-making machines. If AI begins to answer everything for us, the question is no longer technical, but economic: who pays for that answer and who is left out.
The first clear sign that something is moving came at a very specific time in the trading calendar. During the last Black Friday, the big language models started sending real traffic to top-tier online stores. According to Semrush data cited by The Wall Street Journal, twenty large retailers received an average of 183,000 daily visits from AI tools, a figure still small compared to Google, but almost eight times higher than the previous year. The volume is still marginal, but the trend no longer goes unnoticed by those who make a living by attracting and converting users.
When the response replaces the click. Traditional search worked as a referral system: the better positioned a page was, the more traffic it received. The emergence of AI alters this scheme by offering closed answers that, in many cases, reduce or eliminate the intermediate step. This change does not guarantee greater quality or reliability; the models can make errors, mix sources or generate incorrect information. But it does transform the distribution of attention. If the user stops visiting thousands of sites and the interaction, in many cases, is concentrated on the platform that responds, the economic model that has sustained the web for years comes into tension.
This shift in attention has triggered an immediate reaction on the business side. As AI-generated responses begin to influence which brands appear and which disappear from the user’s radar, a new concern arises: how to “be” within those responses. Hence the idea of optimizing for search with AI, a still diffuse terrain in which traditional agencies, newly created startups, such as Evertune or Profound, coexist, and platforms that try to offer metrics, tools and promises of visibility in systems that, by definition, function as black boxes.
The emergence of AI search has not generated consensus, but rather a clash of interpretations. Part of the sector believes that change is incremental and that good practices are still relevant, even if they are now expressed in a different way. In front of them are those who openly speak of a change of era and defend that visibility in generated responses requires a new discipline. Companies, brands and investors move between both extremes, with millions of dollars at stake.
The signs that resist change. In a field that is not very standardized, many of the tactics that best fit generative search are not radically new. Authority, context and editorial clarity remain relevant factors, as does offering useful and verifiable information. Some companies, explains Semrush, are fine-tuning formats, summaries or structures to make it easier for models to read, but without breaking with their previous practices.
When social context enters the equation. Compared to classic SEO, AI seems to rely more on signals external to the website. According to data analyzed by Profound, recency weighs especially heavily in this type of response. And, according to Semrush, user-generated content is also gaining relevance, from forums to comments on social platforms, which models use as raw material to understand products and brands. That introduces a variable that is difficult for brands to control: the real conversation. It is no longer just about optimizing pages, but about understanding that the collective story also influences what the AI returns.
For years an entire industry has been built around a very specific premise: appearing on Google to influence a purchasing decision. SEO specialists, digital marketing agencies, advertising tools and platforms have made a living by optimizing visibility, information and messages that took the user to a store. This system worked because the search acted as an intermediary and referred the potential buyer. If the AI starts responding, recommending and prioritizing or suggesting which link to show to Buy, the entire gear is reconfigured. The question is no longer just how to attract visitors, but how to make money when the intermediation changes hands.
Images | Google | Austin Distel | 1981 Digital
In WorldOfSoftware | The risk that OpenAI goes bankrupt goes far beyond its future as a company: the entire sector depends on it
