What AI Search Changes for Your Business

If you have sat through a vendor pitch in the last year, you have heard that AI search "changes everything" and that your business needs a brand new discipline, a new budget line, and probably their software, to survive it. Some of that is real. Most of it is noise. This briefing separates the two so you can decide what actually deserves your attention.

What is actually changing

For twenty years, a search engine answered a question by handing you ten blue links and letting you choose. Tools like ChatGPT, Perplexity, Google's AI Overviews, and Gemini increasingly do something different: they read across many pages, write a direct answer, and cite a handful of sources underneath it. The user often gets what they need without clicking anything.

This is not a forecast. The Pew Research Center, tracking the real browsing activity of 900 U.S. adults, found that when an AI summary appeared, users clicked a link on just 8% of those searches, versus 15% when no summary was present. Only 1% of visits to a page with an AI summary resulted in a click on one of the cited sources.

The practical takeaway is simple. The new prize is not being the eighth link on the page. It is being one of the few sources the AI quotes and names. That is the new visibility, and for many informational queries it is the only visibility that remains.

The honest reality check

A whole consulting category, often branded "GEO" (generative engine optimization), has appeared to sell you a way to win this game. Read the fine print and much of the advice is repackaged search engine optimization with a new label, padded out with tactics that the data does not support.

Here is what the evidence actually shows. When Ahrefs studied which pages get cited in Google's AI Overviews, they found the citations come overwhelmingly from pages that already rank in Google. In their analysis, the large majority of cited pages were already sitting somewhere in the top 100 search results. Separately, Ahrefs reported that the strongest single factor correlating with AI visibility was simply how often a brand is mentioned across the web. In other words, the things that earn AI citations are the same things that have always earned trust: ranking well, being mentioned by others, and being a recognized name.

"To be eligible to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet... There are no additional technical requirements. Structured data isn't required... and there's no special schema.org markup you need to add."

Google Search Central, AI Features and Your Website documentation. developers.google.com/search/docs/appearance/ai-features

That quote, from Google's own documentation, deflates a lot of the sales pitch. There is no secret markup that unlocks AI citations. The eligibility bar is the same one that has always applied to ordinary search.

One more myth deserves to be retired. A popular strain of advice tells you to optimize for long, elaborate "prompt templates," as if customers interrogate AI in carefully worded paragraphs. They do not. Semrush, analyzing real clickstream data, found that when people use ChatGPT in its search mode, their prompts average between roughly 4 and 9 words, which is essentially the same length as a normal Google query. People type short, keyword-like questions. Building a strategy around imagined essay-length prompts is solving a problem your customers do not have.

The one technical thing that matters

If an AI crawler cannot read your page, you cannot be cited. Full stop. Google's documentation is explicit that its AI features draw on "publicly accessible, crawlable content." This sounds obvious, yet it is where real money is lost, because modern websites quietly break it in ways that are invisible to a human visitor.

A page that loads its content through heavy JavaScript, a robots file that accidentally blocks AI crawlers, a slow or error-prone server, content trapped inside images with no text equivalent: a person sees a normal page, while the machine sees nothing it can quote. Machine-readability, meaning clean HTML, sensible structure, sound schema where it helps, and open crawl access, is not an edge. It is table stakes. It is the price of being eligible to be cited at all.

What you should actually do

Do not panic, and do not chase hacks

There is no clever trick that vaults an unknown, hard-to-read site into AI answers. The levers are the durable ones, and they compound slowly.

Make sure the foundation is readable

This is the highest-return, most overlooked action. Confirm that AI crawlers can reach and parse your most important pages. It is a finite, fixable engineering task, not an open-ended marketing program.

Keep being genuinely authoritative

Be the clearest, most trustworthy answer to the questions your customers ask, and earn mentions from credible places. Since AI citations track closely with ranking and brand strength, the work that builds those also builds your AI visibility. There is no shortcut that substitutes for it.

Measure the new traffic as it grows

Referral traffic from ChatGPT, Perplexity, and similar engines is still small for most businesses, but it is rising and it is measurable. Start tracking it now so you can see the trend with your own data rather than reacting to someone else's headline.