How Google’s AI Overviews are breaking the traditional startup growth formula

For years, the formula was simple. A startup built a website, ranked on Google, and converted that ranking into customers walking through a digital front door. That formula is breaking, and the break is happening inside Google itself.

The company has spent the past two years folding AI Overviews and its generative search experience directly into the results page so that a growing share of queries never require a click. Google answers the question on the spot, and the website that once earned that ranking gets nothing for it.

The scale of that shift is no longer speculative. Nosike Nwigene, a communications advisor who has tracked the shift closely, points to Similarweb’s tracking of click-through rates, which shows a measurable decline in queries where AI Overviews now appear, alongside separate analyses from Ahrefs and Seer Interactive that have found the same pattern even on pages that rank at the very top of search results.

He also cites Gartner’s forecast that traditional search traffic could fall sharply as more users turn to generative tools for answers instead of links, and a related point from Cloudflare’s chief executive, Matthew Prince, who has argued that AI is altering the basic economics of the web by cutting the traffic that once flowed to the publishers and businesses whose content trained these systems in the first place.

The numbers behind that shift have grown harder to ignore. Zero-click searches in the United States, queries that end without any click at all, rose from roughly 60% in 2024 to about 68% in the first four months of 2026, according to research from SparkToro built on Similarweb clickstream data.

Ahrefs’ own analysis found that AI Overviews cut click-through rates by roughly a third for informational keywords, a finding its content marketing director called out publicly as a direct contradiction of Google’s claim that AI Overviews benefit publishers.

Seer Interactive’s September 2025 study put the loss even higher, finding that affected sites saw somewhere between 58 and 61% of their clicks disappear. Gartner’s own projection that traditional search engine volume could fall by as much as a quarter within the next two years as AI chatbots absorb query share was the most aggressive forecast among them when it was published, and the trend lines since have not proven it wrong.

For Nigerian startups, the practical effect is already visible, according to Babatunde Sarumi, a Digital Marketing Specialist who has watched the shift unfold across several sectors.

Read also: Meet AI Mode, Google’s new tool to enhance users’ search experience

Sarumi describes two distinct kinds of loss.

• The first is passive, and it affects startups with weak technical foundations, sites that block search crawlers, neglect sitemaps, or otherwise stay invisible to the systems now scraping the web to build AI answers.
• The second is active, and it comes from the zero-click nature of AI Overviews itself. When Google can synthesise an answer directly on the results page, informational traffic that once went to a startup’s blog post now goes nowhere, or worse, it goes to Reddit, Nairaland, or a handful of high-authority international sites that AI systems trust as sources.

“If a business isn’t actively optimised to be cited as a source inside that AI window, they are losing clicks to Reddit, Nairaland, or high-authority international sites. Relying solely on WhatsApp or Instagram makes this difficult, effectively locking a business out of the modern discovery funnel.” – Babatunde Sarumi

A quiet pivot toward being cited, not ranked

The response Sarunmi describes among his clients is not abandonment of search but a redirection of budget within it.

Founders in fintech, food, and telemedicine, sectors where search intent runs high, are shifting spend away from keyword-driven SEO and toward what practitioners call Generative Engine Optimisation, or GEO. The audience for this content is no longer a search algorithm but a language model deciding which sources are trustworthy enough to cite.

Sarumi says the pivot begins with a technical audit, but the real work is building content that ChatGPT, Gemini, Claude, and Grok can ingest and trust. He points to a shift already showing up in how customers describe finding a business, telling founders directly that they asked ChatGPT for a recommendation and were pointed to that company by name.

That anecdote captures something the industry has been slower to name. AI tools are becoming a default reference point for a growing number of users making everyday decisions, and a business absent from that reference point may find that ranking well on Google no longer guarantees it is found at all.

What startups are doing differently

Sarumi’s advice to clients rests on two pillars: technical trust and AI-first content architecture. The technical baseline has not changed much. HTTPS security, mobile optimisation, and a properly maintained Google Business Profile remain the floor. What has changed is what sits on top of that foundation.

He tells clients to place direct, concise answers within the first hundred words of an article so that AI systems can extract them cleanly. He pushes them to adopt schema markup, the structured data that helps search systems understand precisely what a business is, who founded it, and what it offers.

And he steers them away from generic blog content that AI can easily reproduce without attribution toward original case studies, proprietary data, and expert commentary that a model has to cite because it cannot simply invent it.

None of this guarantees survival for startups slow to adapt. It does suggest that the businesses treating this moment as a technical audit rather than a marketing footnote are the ones already showing up inside the answers their customers are asking for. The rest are still optimising for a search engine that is quietly handing fewer and fewer people back.