Your Customers Are Searching. AI Is Answering. Your Business Isn't in It.

When the automobile arrived in the early 1900s, roads were still designed for horses - narrow, unpaved, built around the cadence of animal movement. Cars could technically navigate them, but the infrastructure was wrong. The businesses positioned along those roads, the supply chains built around them, the entire economic geography of commerce had been optimized for a different mode of transit than what was to come. The businesses built around the old infrastructure either adapted or were absorbed by the ones that did.

We are at an equivalent inflection point, and most leadership teams have not yet registered the shifting roads beneath their feet.

The internet, as it was conceived and built, assumes a human-centered design. Every architectural decision - how content is indexed, how pages are structured, how search results are ranked, how digital channels present themselves - was engineered around human eyes consuming information, human intent driving queries, with human attention as the monetizable asset. That assumption has shaped trillions of dollars of capital allocation toward digital presence, search share, and content strategy. It shaped how industrial businesses, manufacturers, and B2B organizations decided what their digital footprint needed to look like. It shaped what they thought was worth building. Now, those assumptions are, by the weight of the available data, empirically wrong.

In June 2026, Cloudflare CEO Matthew Prince confirmed what his network's data had been signaling for months: bots now generate 57.5% of all HTML web traffic, with humans accounting for the remaining 42.5%. This is the first time in the internet's history that automated requests have held the majority. Non-human traffic has been climbing for years - API calls, crawlers, automated processes - but the arrival of AI agents as a new and dominant category accelerated the crossing of that threshold by nearly two years, and changed what that threshold means. Previous bot traffic didn't make purchasing decisions. These do. Prince wrote simply: "Welp, that happened faster than I predicted." The change is not incremental - AI-powered bots grew 187% from January to December of 2025, while human traffic grew 3.1%.

GPTBot, OpenAI's web crawler, grew from 4.7% of verified bot traffic in July 2024 to 11.7% a year later. Anthropic's ClaudeBot doubled in the first half of 2025 alone. These are not background processes running maintenance tasks - they are the front-end of AI systems that hundreds of millions of people are actively using to research, evaluate, and make purchasing decisions. Perplexity's monthly query volume grew from roughly 250 million to 780 million queries within a year. ChatGPT scaled from 300 million to 900 million weekly active users between late 2024 and early 2026. Gartner, in a February 2024 analysis that now reads as conservative, predicted that traditional search engine volume would decline 25% by 2026 as AI systems absorbed the early stages of discovery and research.

The mechanism that matters here is different from search as leadership teams have typically understood it. When a human searches Google for an industrial supplier or a capital equipment category, they receive a page of links, they click, they read, they form an impression, and they may return. The entire architecture of digital presence investment - search share, content strategy, web channel build-out - was designed to win that moment of attention.

When an AI agent researches the same question on behalf of a human, it does not navigate a list of links and form impressions. It crawls, synthesizes, and produces a conclusion if it didn't already have the answer baked into its model. If a business's content is structured in a way that human readers find intuitive but that AI systems cannot parse and weight efficiently, that business effectively does not exist in the answer. The businesses that built their customer acquisition around search visibility are simply absent from those conclusions.

The scale of this shift is already visible in what is happening to the open web. Zero-click searches - queries where the user receives a synthesized answer without clicking through to any source - increased from 56% to 69% of all searches between May 2024 and May 2025. Google's AI Overviews have expanded from appearing in roughly 6.5% of searches in January 2025 to 48–60% of all searches by mid-2026, depending on the tracker - BrightEdge puts the figure at 48%, Xponent21 at 60% - with coverage reaching as high as 95% for comparison and informational queries. Google expanded AI Overviews to more than 200 countries and 1.5 billion monthly users as of May 2025, and Google itself reported that AI Overviews drove a 10%+ increase in overall usage for the query types where they appear. That trajectory - from 6.5% to nearly half of all searches in under a year - is the number that matters. When AI Overviews appear above organic results, the click-through rate for the top-ranked organic listing drops by approximately 79%. Organic traffic to major news and media properties fell from 2.3 billion monthly visits in mid-2024 to under 1.7 billion by May 2025, a contraction of more than 600 million visits per month in less than a year, and the Washington Post, NBC News, and Mail Online each lost between 40% and 44% of their organic traffic in that window. These are not niche publications with fragile digital strategies - they are among the most sophisticated content organizations in the world.

The counterargument that nearly every executive team will raise internally is a reasonable one: our traffic is still predominantly human, our customers still find us through traditional channels, our analytics do not suggest a crisis. For most businesses operating today, that observation is probably accurate. It is also, however, precisely the observation that the owners of horse-dependent businesses were making in 1908. Only this time the transition is not gradual. What looks like a manageable transition at 5% AI-mediated discovery becomes a structural disadvantage at 30%, and a free cash flow problem at 60%.

These trends drove conversations I've had with leadership teams across industrial and manufacturing businesses I’ve worked with - specifically about how their brands get face time in model outputs, how to streamline agentic access to product information, and how to ensure their businesses show up legibly in AI-mediated discovery. When those conversations happened, the gaps they encountered were real. At the inflection point of horses and automobiles, there was nothing wrong with the organizations that wanted to continue servicing the horse and carriage industry. Yet, the rapidly shrinking total addressable market left little room for many of them to survive. The organizations I am working with are making the decision to reposition their digital presence for AI-mediated discovery - not abandoning what worked before, but not betting the next decade on it either.

AI agents navigate differently than humans. They evaluate differently, synthesize differently, and are beginning to act differently - not just informing decisions but, increasingly, making procurement recommendations, shortlisting vendors, and initiating contact sequences on behalf of the humans who deployed them. The businesses that show up legibly and authoritatively in that layer of decision-making will compound their share of wallet and organic growth over the next several years. The ones that don't will continue to optimize beautifully for a visitor base that is shrinking as a percentage of the traffic that actually matters.

The businesses winning in AI-mediated discovery have already shifted the boardroom conversation from impressions and click-through rates to model outputs, agentic access, and brand legibility to AI systems. That is the competitive lens that matters now, and it belongs in a capital allocation conversation, not a marketing subcommittee.

People still ride horses today. They also still visit websites manually and use traditional search. However, the majority travel by car, and the majority of the discovery and evaluation layer is already moving through AI systems, with the trajectory pointing in one direction. The question worth your time right now: is your business already aligning to cars or still optimizing for horses?

───────────────────────────────────