🔒CT businesses are trying to get recommended by AI — and finding the rules are unwritten

A few years ago, a Connecticut consumer hunting for a Sunday brunch spot, reliable contractor, weekend hotel or new accountant would have typed the question into Google, scrolled past the ads and clicked on multiple links. Increasingly, that same consumer skips the links entirely, opens an AI chatbot and asks in plain English. The program responds with a short list that can feel like a recommendation from a knowledgeable friend. For decades, businesses tried to improve their standing in Google search results through search engine optimization, or SEO, which can involve tailoring website content to match what potential customers are searching for and earning links from other websites. The shift toward AI is now pulling Connecticut businesses into a new contest for consumers’ attention. Some marketers call it generative engine optimization, or GEO, with a goal that is less about driving website traffic and more about becoming the business an AI system recommends. The catch is that almost no one — including marketing consultants — can fully explain how that choice gets made. Google’s traditional search engine returns a ranked list and leaves the judgment to the user. A generative AI tool does more of that judgment itself, pulling from reviews, directories, news coverage and company websites to deliver what sounds like a conclusion. “SEO helps you get found. GEO helps you get chosen,” said Jill Adams, CEO of Avon-based marketing agency Adams & Knight. The shift is generating demand for advice, she said.

“More and more clients are asking how they can ensure their brands, their products, their services and their leaders are positively represented in results for AI-assisted searches,” Adams said.

Found, then chosen

Her colleague Kevin Renwick, the firm’s vice president of media, said clients often put it more bluntly: “We’re hearing questions like, ‘Why isn’t my business showing up in AI-generated answers?’ or ‘How do we become one of the brands AI recommends?’” What the systems reward, the firm argues, isn’t a website but a whole digital footprint that includes “media coverage, reviews, directories, executive visibility, third-party mentions,” said Michelle Bonner, Adams & Knight’s vice president of public relations. Businesses named consistently across credible sources are likelier to surface. And the payoff, Adams said, comes from consumer trust. If AI platforms consistently recommend a business, “that visibility can carry enormous influence because consumers increasingly perceive AI-generated answers as curated and trustworthy,” she said. But AI recommendations do not follow the same rules as traditional search rankings.

Sam Barrett, director of business development at Glastonbury-based marketing agency CashmanKatz, said GEO is not simply SEO by another name because businesses can’t influence AI recommendations the way they have long tried to influence Google search results. “A lot of what the individual AI companies — ChatGPT, Claude, whomever — deliver depends on the models the AI is trained on,” he said. “It’s not the new SEO, and I think people might conflate that.” His broader warning is that AI tools are opaque. “It’s still a black box in a lot of respects,” Barrett said. “You don’t have control over how the algorithms are built, what the models are trained on, or where the content is being pulled from. And AI still hallucinates.”

Five tools, five answers

To see how AI platforms handle the same recommendation request, the Hartford Business Journal recently ran an experiment. A single prompt — “Recommend a Connecticut employment attorney for a senior executive negotiating severance” — was entered into five platforms: OpenAI’s ChatGPT, Anthropic’s Claude, Google AI Mode, Microsoft Copilot and Perplexity. The answers were all over the place. Only two firms — Southport’s Carey & Associates and Madsen, Prestley & Parenteau, which has offices in Hartford and New London — appeared in four of the five responses. No firm consistently claimed the top spot. Four different firms were ranked first by at least one platform. New Haven-based boutique employment firm Garrison Law was the only firm named No. 1 by more than one tool — Claude and Google AI. The programs also made mistakes. Across three separate runs, Claude recommended Garrison Law, calling the firm “the most executive-focused option in the state” in one response while incorrectly placing the firm in Hartford. Perplexity made a different error, recommending a New York-based law firm, which does not have a Connecticut office. Still, the results weren’t entirely off base. While Claude’s description may overstate Garrison Law’s executive focus, the firm does counsel executives on severance matters. Partner Josh Goodbaum confirmed that and said the firm regularly publishes content on executive contracts, severance agreements and related employment law issues. He said Garrison Law doesn’t pay for Google keyword placement or specifically target AI tools, beyond ensuring its website is visible to search engines and their algorithms. Instead, Goodbaum said the same factors that have helped Garrison Law gain visibility in traditional search results, including client reviews, referrals and, as he put it, “consistent work to educate the public about their rights as employees,” may also be helping the firm appear in AI recommendations. “The best marketing strategy is to do great legal work and to provide useful resources to the public,” he said. When Garrison Law asks new clients how they found the firm, Goodbaum said, more are citing AI searches. “We’ve seen a distinct uptick in that category over the previous six months or so, and we expect that trend will continue,” Goodbaum said.

Adapting to AI

For businesses trying to improve their chances of appearing in AI recommendations, marketers tend to offer similar advice. That includes sharpening website copy, cleaning up online listings, building customer reviews and earning outside coverage, then waiting for AI systems to absorb that information. But businesses have limited control over how quickly or accurately that happens. Unlike with traditional search, there is no straightforward way to prompt an AI platform to quickly reflect a correction. A business that believes it has been mischaracterized must generally address the underlying information sources and wait, experts said. Asked how they advise clients in that situation, Adams & Knight said the approach is to identify inaccurate source material, correct it where possible and publish fresh, consistent information. Some businesses are tackling the issue directly. Mohegan Sun, the Uncasville casino and entertainment complex, has begun incorporating AI strategy into its marketing efforts.

AI platforms “are starting to take market share from traditional search,” said David Martinelli, chief marketing officer at Mohegan Sun. The resort expects the trend to continue and is adjusting its search and AI-related strategies accordingly, factoring in differences in how each major platform surfaces information. Mohegan also leans heavily on third-party coverage, Martinelli said, including news stories, reviews and travel-site rankings, and backs it with what he described as a steady volume of marketing communications.

Hype meets reality

However, while consumer use of AI is growing, that doesn’t necessarily mean businesses should rush to overhaul their marketing strategies, experts cautioned. A March survey by digital marketing platform BrightLocal found that 45% of consumers had used an AI tool for a local business recommendation, up from just 6% a year earlier. But broader consumer sentiment remains mixed. A June-July 2025 survey by Stamford-based research firm Gartner found that only about one-third of U.S. consumers believe AI chatbots rival traditional search engines for learning something new. More than two-thirds said they scroll past Google’s AI Overview. Kate Muhl, a Gartner vice president and analyst, said small businesses should be careful about making major investments aimed at influencing AI recommendations. “Proceed with caution,” she said. For businesses trying to reach Gen Z and Millennial consumers, Muhl said YouTube and social media may still be better places to focus. Barrett said CashmanKatz hasn’t built a GEO strategy yet because his clients’ customers still mainly find businesses through Google, while the consumers most comfortable with AI search are not the audiences those clients are trying to reach. There also needs to be a broader and better understanding of AI as a whole, he said. “Until we reach a tipping point where the return on the optimization effort is in our favor,” he said, “we’re going to continue optimizing mainly for SEO.”