AI is rewriting how customers find you

For two decades, the way a company got found was settled. A customer typed a question into a search box, scanned a page of ten blue links, and clicked. An entire discipline grew up around winning a spot on that page. We called it search engine optimization. Whole businesses lived and died on where they ranked. It was a stable game with known rules, and most small companies eventually learned to play some version of it.

That page is now disappearing. Increasingly the customer types the question and gets an answer. It is conversational, sourced from across the web, and written by a model instead of handed back as a list. Google places AI Overviews above the links for a growing share of queries. Hundreds of millions of people now ask ChatGPT, Perplexity, and Gemini the questions they used to type into a search bar. The ten blue links are not gone. They are sliding below the fold, below the answer, and out of the habit.

There is already a name for the new game. It is called Generative Engine Optimization, or GEO, and the term barely existed before a 2023 research paper gave it one. The shift it describes is not cosmetic. SEO was about being the best ranked page. GEO is about being the source an AI trusts enough to cite in its answer. Those are different problems. A page can be perfectly optimized for keywords and still never get quoted by a model that decides, in a sentence or two, which businesses are worth mentioning at all.

The stakes are higher than they look, because of a quieter trend underneath all of this. Even before AI, most Google searches already ended without a click. The answer showed up on the results page and the user never visited a site. The AI answer takes that to its logical conclusion. When the model resolves the question completely, the click may never happen. For a small business, the old goal was to get the visit and win on your own site. The new goal is harder. You have to get into the answer itself, because for many customers the answer is now the whole interaction.

Search rewarded the best optimized page. AI rewards the most trustworthy source.

I have watched a platform shift like this before, from the inside. I spent years leading digital strategy for some of the biggest names in automotive, including Nissan, Mazda, Scion, and Toyota. I lived through the move from desktop to mobile and the rise of search as the front door to every purchase. The lesson from those transitions was always the same. The companies that suffered were not the ones with the worst product. They were the ones who treated the old channel as permanent and woke up to find the customer had moved. The channel changes faster than the business does, and the gap between the two is where companies quietly lose ground.

This matters more for small companies than for the giants, and I find that genuinely interesting rather than just threatening. A large incumbent had an accumulated SEO moat. It was built from years of backlinks, domain authority, and content libraries a newcomer could not replicate. AI answers reset a meaningful part of that. A model deciding which regional logistics provider or care operator or specialty manufacturer to name is weighing relevance, clarity, and trust signals. It is not simply rewarding whoever spent fifteen years gaming the algorithm. That reset is an opening for the company with real substance and no legacy moat, which describes a great many of the small businesses we look at.

The opening only rewards a particular kind of company, and this is the part that connects to everything else we believe. You cannot keyword stuff your way into an AI's trust the way you once could into a rankings page. The models, imperfectly but more and more, reward what is actually true. They reward clear information about what you do, consistent signals across the places customers and machines look, evidence that real people vouch for you, and substance a model can verify instead of slogans it has to discount. The businesses that win the answer tend to be the businesses that deserved to. That is a strange and welcome thing to be able to say about a distribution channel.

So when SwellPoint diligences a company today, how its customers will find it in five years is no longer a side question. It is part of underwriting durability. A business with great operations and no plan for the shift from search to AI is exposed in a way that will not show up in this year's numbers. It shows up in the slow erosion of the years after. This is also an area where our experience is directly useful. We can help a portfolio company become the trustworthy, legible, well attested source that the next generation of discovery rewards. The front door to every small business is being rebuilt. The companies that notice early, and build real substance behind the door instead of tricks in front of it, are the ones who will still be easy to find on the other side.