AI is rewriting commercial visibility from the search results up, and brands are already behind

The logic of commercial visibility is changing faster than most marketing organizations have registered. David Conforti puts the structural shift plainly: where a search engine once returned ten links, AI systems return one to three results. If a brand is not in that range, Conforti argues, it might as well be ranked hundredth. That is not a description of a competitive disadvantage. It is a description of commercial nonexistence.

The supply side of advertising is changing at a matching pace. Cliff Weitzman reports that Speechify already tests close to 1,000 AI-generated ads per day, alongside roughly 8,000 human-made organic creatives produced each month. The sheer volume of AI-assisted creative iteration Speechify is running today would have required a production infrastructure that no single company could have sustained a few years ago. That volume is now a baseline, not a ceiling.

The production capability is spreading, but unevenly. Rory O’Driscoll estimates that roughly 10% of CMOs can currently run AI-driven campaigns effectively. His projection is that competitive pressure will push that number to 20% and then to 40% over time, not because the holdouts will be persuaded but because they will be forced out by those who adapted earlier. That is a Darwinian framing, and the evidence supports it: the cost and speed advantages of AI-native creative production are large enough to compound into market share differences that are difficult to recover from.

We test almost a thousand day generated ads a day right now on top of the roughly 8,000 organic creatives we make with humans every single month. Cliff Weitzman

The discovery layer is where the shift becomes most visible to consumers. Martin Casado points to image-based product queries as a concrete example: a user photographing a coat and asking what it is and where to buy it would not have produced a reliable answer five or ten years ago. It works now. That single capability, extended across product categories, represents a meaningful change in how purchase intent gets resolved. The query does not go to a list of links. It goes to an answer.

Andy Fang’s data from DoorDash illustrates what AI-curated discovery produces at scale. Fifty percent of the restaurant order trajectories initiated through Ask DoorDash end with users ordering from restaurants they have never ordered from before. Grocery orders through the same system show basket sizes roughly 40% larger than non-AI-assisted orders. Both figures suggest that AI recommendation is not simply replicating existing preferences at higher speed. It is expanding the consideration set in ways that translate directly into revenue.

Looking forward, Liva Ralaivola places real-time generative ad creative within a two-to-three year horizon. The further projection, within five years, is a model where users actively ask AI assistants to surface a curated selection of options, whether shoes or travel, and consent explicitly to receiving advertising as part of that process. That is a different relationship between user, AI, and brand than the one the display advertising era established. The advertising is not interrupting the discovery process. It is the discovery process.

The brands that treat this as a future problem are already behind. The creative production advantage is operational today, as Speechify’s ad volume demonstrates. The visibility stakes are real today, as Conforti’s one-to-three-result dynamic shows. What Ralaivola and Fang describe is the fuller architecture of what is already partially built. The question for any brand with meaningful exposure to consumer discovery is not whether to prepare for this shift. It is whether the organizational and creative infrastructure being built now will be sufficient to compete when the shift finishes arriving.