5 AI Engines. 5 Different Answers. One Profit Gap.
We asked ChatGPT, Gemini, Perplexity, Google AI Overviews, and Claude the same question: who should I call for commercial HVAC service in Orange County? We got five different answers. The same company appeared in exactly one of them. Most of the businesses we tested did not appear in any.
This is not a technology story. It is a revenue story. And if you own an established service business, it is probably your revenue story right now.
The Test That Changed How We Think About Visibility
The experiment was straightforward. We picked five service categories common to regional trades businesses: commercial HVAC, fire protection inspection, residential plumbing, office moving, and commercial pest control. For each category, we submitted the same prompt to five AI answer engines simultaneously: a natural-language question a real buyer would type. Something like, "Who are the most trusted commercial HVAC contractors in Orange County?"
The results were not just different. They were structurally different. One engine pulled from review aggregators and weighted recent volume. Another weighted how frequently a company was mentioned in third-party editorial content. A third appeared to weight the freshness of a business's own published content. A fourth drew heavily from structured local data. The fifth seemed to weight something closer to topical authority, how many times a company had been cited in the context of answering a specific type of question.
Five AI engines asked the same question and returned five different winners. The business that appeared in four of them had not run a single paid ad in the past year.
The businesses that appeared most consistently across all five engines shared a specific profile. They were not the biggest companies in their category. They were not the ones spending the most on advertising. They were the ones whose digital footprint was structured for machine interpretation, not just human browsing.
Why Your Rankings Are Telling You a Story That Is No Longer True
For the past decade, a business owner could look at a Google rankings report and feel reasonably confident about visibility. If you were on page one for your core service terms, customers could find you. That model is breaking.
Google AI Overviews, now appearing on an estimated 38 percent of searches that previously returned a standard ten-link results page, do not show those ten links. They show a synthesized answer, sometimes with two or three business citations, sometimes with none. A business that ranks number two organically may not appear in the AI Overview at all. The traffic that used to flow to that ranking position now flows to the AI-generated answer, and the business in that answer is not necessarily the one that earned the ranking.
The more significant shift is happening across the non-Google engines. ChatGPT now handles more than one billion queries per week. Perplexity has grown its user base by over 400 percent in the past eighteen months. Gemini is embedded in Android devices as the default assistant for hundreds of millions of users. These are not niche tools. They are where buyers are going when they want a recommendation rather than a list of links to evaluate.
And none of them use the same algorithm. None of them weight the same signals. None of them are looking at the same data.
The practical consequence for a service business is this: your company can be well-optimized for traditional search and effectively invisible to AI recommendation engines at the same time. The two visibility systems are running in parallel, and most businesses have only prepared for one of them.
The Profit Gaps Hiding Inside the Disagreement
When we mapped the results of our five-engine test against the businesses that showed up most and least consistently, three specific profit gaps emerged. These are not abstract marketing failures. They are invisible breakdowns in how customers find, trust, and buy that translate directly into jobs that went to a competitor instead of you.
The findability gap. AI engines cite businesses they can describe accurately and specifically. A company whose digital presence describes it as a "full-service HVAC contractor" is harder for an AI engine to recommend confidently for a specific job, say, rooftop commercial unit replacement for a property management company, than a company whose content has addressed that specific job type in detail. The engines are not rewarding breadth. They are rewarding specificity, and most service businesses have built their online presence for the former.
The credibility gap. AI engines weight third-party evidence heavily. Reviews matter, but not just their star rating. The content of those reviews, whether they mention specific services, specific outcomes, specific locations, matters more. A business with 200 reviews that all say "great service, highly recommend" is less citable than a business with 80 reviews that describe specific jobs in specific language. The AI is looking for evidence it can quote, and most review profiles do not give it any.
The conversion gap. Being recommended by an AI engine is not the same as being chosen. The businesses that convert AI-driven inquiries most effectively have content that answers the questions a buyer asks after they receive the recommendation. What is the process? What does it cost? How long does it take? What do I need to prepare? A business that appears in an AI answer but has no content to support the next step of the buyer's journey is generating awareness without revenue.
These three gaps are what FreshThink's AI Visibility Check scores across nine specific dimensions. The diagnostic does not tell you whether AI search matters in general. It tells you exactly which of these gaps is costing your business the most, in ranked order, before a single dollar is spent trying to fix anything.
The AI Visibility Check does not tell you whether AI search matters in general. It tells you exactly which gap is costing your business the most, right now.
What the Businesses That Showed Up Consistently Were Doing Differently
Across every category we tested, the businesses that appeared in four or five of the five AI engines had built their digital presence around a specific set of practices. None of them were technically sophisticated. All of them were deliberate.
They had published content that answered the specific questions their best customers ask before hiring them. Not blog posts about industry trends. Not generic service pages. Specific, practical answers to the questions a commercial property manager or a general contractor asks before signing a maintenance agreement. The kind of content that positions a business as the expert answer rather than one of several options.
They had structured their review collection to capture specificity. They were asking customers to describe what was done, not just how they felt about the experience. That one change in how reviews are requested produces a completely different type of evidence for an AI engine to work with.
They had verified and expanded their presence on the data sources AI engines draw from. This is not just Google Business Profile. It includes industry directories, local business associations, trade publications, and the structured data signals that tell an AI engine what a business does, where it operates, and who it serves.
None of this is new marketing. It is existing marketing reoriented toward the systems that are now making the first recommendation. The businesses doing it well are not spending more. They are spending differently, and they started before their competitors did.
This is the window that is open right now. The AI search ecosystem is still early enough that a deliberate, well-structured push over the next 60 to 90 days can establish a visibility position that will compound for years. The businesses that act in this window will not need to explain their AI visibility strategy to their competitors. Their competitors will be explaining it to them.
The businesses that wait for the category to mature before acting will find that the recommendations are already set, and changing them is significantly harder than establishing them in the first place. The AI engines are building their recommendation models right now, and the businesses feeding them the right signals right now are the ones being built into those models.
What to Do Before Your Competitors Figure This Out
The path forward is not complicated, but it does require starting with a clear picture of where you actually stand. The instinct most business owners have is to start producing more content, to run more ads, to do more of whatever they were already doing. That instinct is the problem. More activity without a diagnosis of which gap is costing you the most is how marketing budgets get spent without producing traceable revenue.
The right starting point is an AI-powered diagnostic that scores your current visibility across the dimensions that AI recommendation engines actually weight. Not a generic website audit. Not a keyword ranking report. A scored assessment of whether customers can find you, whether the evidence available to AI engines is sufficient to recommend you, and whether the conversion path from recommendation to booked job is intact.
Once you know which of those three areas is your biggest gap, the fix is targeted rather than broad. A findability gap requires different work than a credibility gap. A credibility gap requires different work than a conversion gap. Treating all three as equally urgent is how businesses end up spending money on the wrong thing.
For an established service business with twenty or more years of reputation, the assets you need are already there. The jobs you have done, the customers who have trusted you, the expertise your team has built. The gap is not the substance. The gap is the structure around that substance that allows AI engines to find it, interpret it, and recommend it.
The businesses we tested that showed up in four or five AI engines were not smarter or better funded than the ones that did not. They had simply organized their existing reputation into a form that AI systems could read and cite. That is a structural problem with a structural fix, and it does not require starting over. It requires knowing exactly where the gaps are.
Find Out Where Your Revenue Is Leaking