You Own Google. You're Invisible in AI Search.
A major Planet Fitness franchisee and a financial services market leader ran the same test this year. Neither showed up in AI answers. MIT Sloan just published why — and the data kills the assumption that Google dominance transfers.
The test took ten minutes. Kate Klein, EVP of Marketing at Houston Fitness Partners — a major Planet Fitness franchisee with a serious search investment — ran some queries through ChatGPT and Perplexity to see how her brand showed up.
A small local company was showing up where they weren't.
"That was a wake-up call," she said. "We were shocked when a small, local company in Houston was landing better in AI searches."
Around the same time, a financial services executive watched a consumer research vendors on ChatGPT. Her company held the largest market share in its category. It had spent more on traditional media, digital marketing, and SEO than any competitor. The AI didn't recommend it. A much smaller player appeared instead.
Both cases appear in a January 2026 MIT Sloan Management Review paper by researchers from UVA's Darden School of Business and the University of Houston. The thesis: even market-leading brands risk becoming invisible in AI search if they keep running the playbook that built their position in traditional search.
The conclusion is uncomfortable but specific. Market leadership in Google doesn't transfer to AI search. The moat you spent a decade building may not protect you in the channel absorbing more top-of-funnel intent every quarter.
By the numbers
- 37% of domains cited by AI engines don't appear in any traditional search results — Zhang et al., arXiv:2512.09483
- 88% of Google AI Mode citations don't appear in the SERP top 10 — Moz, 2026 (40,000 queries)
- 65% of ChatGPT's top-cited pages come from DR80+ domains — Ahrefs (1,000 citation analysis)
- 85%+ of non-paid AI citations originate from earned media, not brand-owned content — Muck Rack
- 60% of Google searches already end without a click — SparkToro, 2024
Why SEO doesn't buy AI visibility
SEO is a ranking game. You optimize content for an algorithm that produces a list. A user clicks. That chain assumes the algorithm reads your content, weights it against competitors, and surfaces you near the top.
AI search skips the list. It synthesizes. When someone asks ChatGPT or Perplexity who leads a category, the model draws from what it learned during training and what recent sources it can retrieve, prioritizing publications it already treats as credible. Your 10,000 optimized blog posts don't enter the calculation.
The 37% finding from Zhang et al. makes this concrete: more than a third of the domains AI engines cite don't appear in traditional search results at all. These aren't brands that "also" won AI search alongside their Google rankings. They're brands with entirely different authority profiles — documented in publications, cited by journalists, recognized by the sources AI engines treat as ground truth.
The small Houston company that beat a major franchisee likely had something more useful than any SEO budget: a few editorial placements in local health and wellness publications that AI systems already indexed as authoritative. Three earned media mentions in the right publications can outperform 10,000 SEO-optimized pages when the mentions live inside sources the AI treats as credible.
The allocation problem
Most brands have the inverse of what AI search requires. They've over-invested in content optimized for algorithm position and under-invested in earned editorial presence in publications that AI engines treat as ground-truth sources.
That wasn't a mistake — it made sense for two decades. The ROI calculation just changed.
Muck Rack's analysis of AI citation behavior found that more than 85% of non-paid AI citations originate from earned media — third-party editorial placements, not brand-owned content. The brands winning AI search aren't doing anything exotic. They built real relationships with editors and publications. Those placements are now being read by machines deciding who shows up when a buyer asks "who's the best [category] company?"
Meanwhile, brands that optimized for click-through rates and keyword density are invisible in the channel taking share from traditional search. SparkToro's 2024 zero-click study found roughly 60% of Google searches end without a click — a number that grows as AI Overviews expand. The traditional SEO model requires a click to work. AI search doesn't need one.
What doesn't transfer — and what does
There's an assumption embedded in most marketing organizations: that the authority built in traditional search would transfer to AI search. That domain rating, content volume, and keyword coverage would carry over.
The MIT Sloan data says it's not happening. For both the fitness franchisee and the financial services firm, years of SEO investment and market leadership produced zero protection in AI answers. A competitor with a fraction of their brand equity showed up where they didn't.
This isn't an argument that SEO is finished. Technical hygiene — crawlability, site structure, schema — still matters for baseline visibility. But the authority AI engines use to decide who to cite comes from a different source entirely.
Research from Princeton and Georgia Tech found that adding verified statistics cited from credible external sources improves AI visibility by 30-40% — Aggarwal et al., SIGKDD 2024. The variables that move AI citation rates are named sources, external validation, and placement in publications that AI already indexes as authoritative. Keyword density is not on the list.
Two-thirds of Gen Zers and more than half of Millennials now use LLMs to research products before buying, according to Harvard Business Review citing Pernod Ricard's 2026 analysis. If your brand doesn't appear in those answers, you're not in consideration. The investment that earns you a place in those answers is not paid search — it's earned editorial presence in publications AI engines trust.
The mechanism behind it
What's actually happening isn't an SEO failure. It's a category confusion about what's being optimized.
SEO optimized for human readers clicking links. Earned media built authority with human readers who trusted the publications covering you. AI search inherited the second mechanism, not the first. The publications that shaped human brand perception for decades are the same ones AI systems treat as authoritative sources. When a prospect asks Perplexity who leads your category, the answer is downstream of your editorial presence — not your ad spend, not your content volume.
This is what Machine Relations is built around: the discipline of ensuring your brand is documented by the publications AI engines already index as credible. PR's original mechanism — earned coverage in trusted publications — is the mechanism that works for machine readers too. Machine Relations applies it systematically for an era where the first reader of that coverage is often an AI engine building a shortlist before a buyer opens a browser.
The fitness brand case isn't a warning about the future. It's a report from the present.
To see where your brand currently stands in AI answers: https://app.authoritytech.io/visibility-audit