After running AI search visibility audits for five different companies and sectors, we discovered that one pattern remained consistent: AI reliably cites products, services and pricing, but almost never a company’s reputation. This post examines why the thing your company is proudest of is often the thing that is least visible to the platforms now shaping how your firm gets discovered.

Listen to “What Makes Your Company Different? AI Can’t Tell You” Deep Dive

Ask your favorite AI platform this question: what makes your company different? Then see if it matches the answer you’d have given.

In an AI search visibility audit that we ran across biotech, healthcare and industrial manufacturing companies, ChatGPT, Claude, Perplexity and Gemini consistently cited these companies for their specific services—e.g., a provider’s services by location, a drug by name, a manufacturer’s product catalog. Yet, no prompt returned anything about these companies’ reputations. One company that prides itself on being one of the top 100 places to work in its sector produced zero citations about this. Unbeknownst to the company, that silence was creating a recruitment void, making the thing that they were rightfully most proud of invisible to every prospect job searching in a competitive field.

Consider this: companies will spend over $100 billion this year protecting or amplifying the reputation that sits on their About Us page.

So how is it that a company’s best story isn’t being retold by AI?

“Ask your favorite AI platform this question: what makes your company different? Then see if it matches the answer you’d have given.”

The Reputation Gap in AI Search Results

In our previous post, we made the case that more people, no matter the sector, are starting with AI rather than a company’s or product website. Supporting this are two studies: Bain & Company found that 80% of consumers rely on AI search results for at least 40% of their routine information discovery, while McKinsey found that among people who use AI-powered search, 44% now call it their primary and preferred source of insight, which they’re turning to first ahead of traditional search, retailer websites and review sites.

This is a stark reminder that the person doing the search isn’t the first one forming an opinion about your company—the AI platform is, and before your prospect ever sees a result. For an AI search audit that we ran for a national B2B franchise in the repair industry, we found that despite the company having the digital infrastructure of a category leader and an active franchise-development presence, almost none of this information ever appears in AI search results. The audiences here—property managers and prospective franchisees—increasingly start their search with AI, which means the franchisees buying into the brand are unknowingly inheriting a visibility problem that’s just as invisible to them as it is to AI. Whatever you say about your reputation, if it’s not being cited, it’s lost.

This is what introduces the real differentiation between reputation and citation, and why public relations (PR) is playing a critical role in closing the gap between them.

Earned Media Doesn’t Create Your Reputation; It digitizes It

As I ran through these audits, I wanted to understand not just how and where prompts would result in AI platforms pulling in citations, but where those sources actually lived. One company’s own website was cited roughly 40%-50% of the time. Another sat closer to 25%. Two others fell under 10%, and one of them had zero citations. The other sources included media-specific sites, reviews and third-party sites, such as Reddit and YouTube.

The majority of citations, where these companies or their products were mentioned, came from peer-reviewed publications, press coverage and third-party recognition that AI treats as authoritative. That’s due to PR’s ability to not only drive the media coverage LLMs gravitate to, but to take a brand’s reputation and transpose it into a citation that AI can actually read. PR isn’t abandoning what it’s always done; rather, it’s extending it, from driving earned coverage that builds human awareness, to building a data footprint that expands a company’s share of voice inside AI search. These findings, and every audit we run, become part of the playbook that we hand to our PR partners—i.e., the specific gaps, the missing sources, the exact language AI is and isn’t picking up—so that the AEO strategy and the PR strategy move in lockstep instead of two teams optimizing for two different audiences.

A University of Toronto study found that AI search engines show a systematic and overwhelming bias toward earned media over brand-owned and social content. That finding backs up something worth saying plainly: if your reputation isn’t documented in the data, as far as AI is concerned, it doesn’t exist.

The Historical Connection to the Citation Problem

If anything, our recent audits revealed that what you’d expect to be cited—e.g., a company’s flagship products or brand name—is highly visible, while something entirely unexpected, like a citation for being one of the top 100 places to work, is nonexistent. While optimizing for AI search is still very much a moving target, it’s remarkably clear that the story each company is proudest of—and its actual reputation—are consistently what AI knows least about.

This disconnect isn’t new. It has a direct historical parallel in traditional search. In 1998, Google built its search foundation on an algorithm called PageRank, which treated an online link just like an academic citation. It treated links as a form of authority, or currency, if you will, that the algorithm could count. The AI models we use every day operate the same way. They don’t know who you are or what makes your business great. Instead, they focus on what’s being said about you, where it’s being said and how often. If your most valuable corporate stories, or the attributes that make your business unique, don’t have the repetition and amplification AI is scanning for, they’ll remain unknown to your customers and prospects.

ChatGPT Just Made AI Search Even More Complicated

On Wednesday, July 8th, OpenAI released GPT-Live, a full-duplex voice model that listens and speaks at the same time, which is much closer to an actual conversation than a typed query and response. Instead of typing in a question and reading an answer, you talk back and forth with ChatGPT in real time. The harder reasoning and search work still happens behind the scenes, handled by GPT-5.5, while the conversation keeps moving. As explained here, sometimes the answer shows up as a visual card on screen, sometimes it’s spoken directly, all without you ever visiting a source site. Claude, Perplexity and Gemini will likely follow with their own versions of this model over the next few months, if not weeks.

If you’re interested in seeing how AI currently cites your company and its reputation, where it’s visible and where it isn’t, I’d welcome the chance to show you. Let’s talk. Email me at mike@scheinerinc.com. As always, thank you for reading and/or listening.

Photo by Mihály Köles on Unsplash