How to get found in 2026: search, AI answers, and what actually works

Getting found in 2026 means being visible in more than one place at once: the traditional search results, the answers that AI tools like ChatGPT, Gemini, and Perplexity now generate, and the physical world your customers move through. The good news is that the work which earns one of these increasingly earns the others, so a small business does not need a different plan for each.

This is the starting point for Found & Ranked, our series on how businesses get discovered and chosen. Search has changed more in the last two years than in the previous ten, and a lot of what is now sold as the answer is either old advice in new packaging or simply wrong. Here we map the whole landscape plainly. As the series grows, each part gets a closer look of its own.

What actually changed

For twenty years, getting found online mostly meant ranking on Google and earning the click. That is no longer the whole game. People increasingly ask an AI assistant a question and read a synthesised answer without clicking any link at all. By early 2026 roughly six in ten Google searches ended without a click to any website, according to SparkToro's analysis, and on the searches where Google shows an AI Overview the click-through rate to the top result falls sharply. The job has widened: not just to rank, but to be the source an answer is built from.

This matters less for some businesses than the headlines suggest, which we will come to. But the direction is not in doubt, and pretending otherwise does nobody any favours.

Why the fundamentals did not die

The reassuring part: most of what has always worked still works, and now does double duty. A site that loads fast, can be read by a machine, answers real questions clearly, and states who and where the business is consistently across the web is exactly what both a search engine and an AI assistant need. Google's own AI features draw on the same index as ordinary search, and Anthropic's Claude leans heavily on a web search index too, so good search foundations are the price of entry to being cited by AI at all.

In other words, there is no separate "AI strategy" that bypasses the basics. The basics are the strategy. What changes is what you build on top of them.

What is genuinely new: being cited, not just ranked

The new skill is getting your content quoted inside an AI answer. Research from Princeton, presented at the KDD conference in 2024, tested this directly and found that the things which help are structural and editorial, not keyword tricks: a clear, direct answer near the top, specific facts and figures, named sources, and clean, fluent writing. Stuffing keywords did nothing. Encouragingly for smaller businesses, the same research found that a well-structured page from a modest site can be cited ahead of a vague one from a giant, because the AI is judging the passage in front of it, not just the size of the brand behind it.

This is the part of the field often called GEO (generative engine optimisation) or AEO (answer engine optimisation). The names are new and still settling. The underlying idea is old: say something true and useful, clearly, and make it easy to find.

The local picture is better than the headlines

If your customers are mostly local, the zero-click story is far less frightening. The decline in clicks is concentrated on informational searches, the "what is" and "how do I" questions an AI can answer in a sentence. Searches with intent to buy or hire, and local searches for a real business nearby, still send people to websites and still convert. The clicks that do survive an AI summary tend to convert better, because the person has already read the overview and arrived knowing what they want. For a plumber, a solicitor, a bistro, or a charity, the people most likely to become customers are the most protected group of all.

Some of the best ways to get found are not digital

It is a strange blind spot in most marketing advice, but worth saying plainly: the highest-return marketing for a small business is often not on a screen at all. A clear sign on the van or the shop, a well-targeted leaflet, a conversation at a local networking morning, a phone answered like a human being, these still win real work, and frequently for less money than another month of content.

They are also not separate from your online visibility. A sign people remember sends them to search your name later, which builds the branded searches that tell Google you matter. A job done well earns the review that local AI search leans on. Turning up in your community earns the genuine mentions and links that build your authority. Offline work feeds online visibility. Any honest account of getting found has to include the doormat and the phone, not just the algorithm.

A word on the advice market

Because the field is new and a little frightening, it has attracted a great deal of confident nonsense. Some of it is harmless, some of it costs real money for no return. A clear example: through 2025 and 2026 a file called llms.txt was widely promoted as essential for AI visibility, yet Google has said plainly it does not use it, no major AI provider treats it as a ranking signal, and a large study of tens of thousands of sites found no measurable effect. Throughout this series we will say what does not work as readily as what does. That is the only kind of advice worth having.

What this means for your website

Getting found in 2026 is not one trick, and it is not a panic. It is a fast, clear, honest website that answers the questions your customers ask, described consistently wherever your business appears, supported by the offline presence that earns trust in the first place. Do that and you are findable by search engines, citable by AI, and memorable to the people who walk past your door. That is the approach we take to the work that follows once your site is live, and the thinking behind the wider growth work we do for businesses that are already up and running.

The rest of the Found & Ranked series digs into each of these ideas properly, from the new mechanics of AI citation to the local and offline work that quietly does the heavy lifting.