There’s a strategic gap between how leadership talks about the business and how customers actually see it. Ask an AI platform about a company by name, and it’ll describe it accurately and favorably. Ask an unbranded or buyer-intent question instead, and that same company no longer shows up. That’s not a visibility problem; it’s a positioning problem, and it traces back to a question Theodore Levitt asked in 1960: what business are you really in?

Listen to “AI Doesn’t Redefine Your Positioning. It Exposes the Wrong One.” (Audio Deep Dive)

For the majority of my career, I’ve helped companies identify where their brand awareness is being challenged, or where there is a disconnect between how the internal leadership of the company spoke about it versus its customers’ perception of it. As AI has become, in some ways, our new primary customer, it isn’t just deciding what we should see when we ask a question; rather, it’s filtering out companies, not because they aren’t relevant to the answer or even qualified, but because the company was never set up to be found as the answer to these questions in the first place. I’m not sure many companies even realize this, or how much revenue and opportunity is being diverted to competitors because of this positioning disconnect.

When was the last time you checked to see if your company is on AI’s recommended shortlist based on what your company does?
It’s not that AI doesn’t know your company. It most likely does, especially when asked about your company by name. But what happens when the question pertains to a specific attribute or criteria that’s part of a prospect’s checklist? That’s a clear pattern I’m seeing when I run these AI search audits.

Does your company get cited when asked about by name? Check. But say the question becomes broader and more industry-specific, something like “Which companies would make the shortlist as a full-service solution for X?” One company is explicit about its offerings, yet says nothing about being full service, meaning that there’s no reason it should be included in the shortlist answer, and because it positions itself using such a narrow term, AI doesn’t even consider it.

“When was the last time you checked to see if your company is on AI’s recommended shortlist based on what your company does?”

Here’s the part most companies miss: AI reads your site, reviews and third-party sites exactly as you describe yourself. The question companies need to start asking is how their description compares to the language their competitors are using, and whether that’s the reason they never make it into the consideration conversation at all.
It’s now about surfacing the language you think you should be known for, and making sure it sits alongside the language that gets you included on the shortlist.

AI Knows You; It Just Won’t Recommend You

I recently ran this exact audit for a global project advisory firm working across industry-specific construction projects. Here’s what I found:

The audit exposed not only a lack of visibility, but a bigger brand positioning challenge, where AI platforms were completely omitting this company when it wasn’t asked for by name.

To get a baseline view of how a company is getting cited by AI, I run prompts based on how the company’s audience would search: branded, using the name of the company in the prompt; unbranded, a more general question about the category to see who comes up; and buyer-intent, a more specific, detailed question that includes project type, size, budget and other parameters.

For example, I asked each of the AI search platforms a series of questions about the company. What drastically changes is when this company’s name is removed from the prompt entirely. Ask “who are the top project management firms for construction?”, and this company is completely absent against its four competitors. Now, as I’ve seen from running other AI audits for companies, this isn’t uncommon.

But what wasn’t obvious to this company is how its competitors are positioning themselves, which is more than an AI visibility issue; it’s a brand messaging challenge. These competitors aren’t only winning in visibility, they’re likely winning in the prospecting pipeline as well. The other thing these competitors are doing well is getting cited in industry-specific online media, where they were independently ranked, giving AI a third-party source to draw on that the company I audited doesn’t yet have.

So, I ran a second, more targeted test, this time using the company’s own specialty language for its own top-priority growth sector, in both the unbranded and buyer-intent prompts. I used the same language this company used to describe what it does, including its own specialty language for that specific sector. Yet, the company was still only visible on half of the platforms. The other thing worth mentioning is that all of its citations ladder back to its own website, which, while positive, is also limiting, in both exposure and real validation. When competitors were mentioned, it wasn’t just their own websites that were sourced, but third-party mentions as well.

The Citation Gap Is a Positioning Gap

This is why, when I run these audits, my findings aren’t just being validated by an LLM, but through qualitative interviews with leadership, board members and customers, as well as a deep competitive review. The LLMs show potential trends or common language being used. The interviews reveal what the LLMs, or the brand, are missing, especially through the lens of customers and their personal perception of the company or category. Then the competitive analysis shows how competitors are getting cited, and what language is associating them with unbranded and buyer-intent prompts that the client’s company may or may not be getting bucketed into.

Which is why a citation gap and a category or messaging gap are actually one and the same.

Say, for example, you’re a bread company whose entire strategy is built on freshness. Your competitors are also pushing freshness as part of their strategy, but also reference variety and free delivery. If I search for “bakeries near me with fresh bread,” your bread company will most likely be cited. But when I change that question to “Which bakeries near me have a variety of breads and offer delivery?”, your company is removed entirely, because strategically, its messaging is underselling itself against competitors’ messaging. Layer on top of that how AI treats third-party citations as validation, rather than a company’s own website that says it’s the best. Staying with the bakery example, are you going be swayed solely by a company’s website, or by reviews on Google or Yelp about that company? AI takes the same approach.

What Business Are You Really In?

In this well-documented Harvard Business Review (HBR) article, Marketing Myopia, which was written by Theodore Levitt in 1960, Levitt begins by asking, “What business are you really in?” This is the overarching premise that in order for a business to sustain growth, it should position itself very broadly against the customer’s needs. I found this question so appropriate for companies to describe what they do so that it’s broad enough for AI search and speaks to the needs of customers—and, more importantly, differentiates it from its competitors.

I believe the same can be said for strategy, messaging and AI search. Whether you’re focused on a set of products or on the depth of experience you have across several areas of business, how you position and talk about them needs to be relevant to your customers’ needs and wants, and it also needs to hold up against how your industry is positioning itself. What language are your competitors using to describe their services that not only makes them stand out, but also gets picked up by the LLMs? If companies don’t consider exploring these strategies, and instead focus only on their products or services, they risk not just sounding generic, but also their growth and visibility with LLMs.

Despite his article being written in 1960, Levitt built a case that still holds true today. His argument points at leadership, and while he doesn’t say it directly, most likely the board as well. Levitt argued that stalled growth is rarely about a saturated market; rather, it’s a failure of management, one that typically looks like companies selling products instead of marketing to customers what they actually need.

This goes back to one of my earlier questions: Does your leadership or board actually agree on what business you’re in, or just on what you call yourself? I believe that aligns closely with Levitt’s findings, and it’s why AI search escalates the stakes on getting it right.

For a company’s leadership or board of directors reading this and recognizing that there’s more than a little truth to it regarding their own business, the first step is asking this question: “When was the last time you checked to see if your company is on AI’s recommended shortlist based on what your company does?”

That’s the first step in understanding whether AI is exposing a company positioning that was already wrong. If you’re not sure whether this is happening to your company, I’d love to show you whether it is and how to fix it.

For more on AI strategy and its business impact, you can reach me at mike@scheinerinc.com. As always, thank you for reading and/or listening.