AI Engines Know Tony Uphoff. They Recommend His Businesses Almost Never

Tony Uphoff asked ChatGPT, Gemini, Perplexity and Google AI Overviews the questions his prospective clients ask, and his businesses almost never made the list. Across 59 answers this month, his newsletter Uphoff on Media appeared in zero of 12 category answers. His firm, Uphoff Advisory, appeared in one of 43. Yet every engine that answered "Who is Tony Uphoff?" recited his career accurately. "The models know who I am. They almost never recommend what I've built," Tony said.

That gap, between being known and being recommended, is the AI search measurement problem most marketing teams have not solved. Tony's October posts lay out why CFOs, acquirers and employers will start asking about it — and what practitioners should change now.

Tony has been CEO of five businesses, including Business.com and ThomasNet.com, led the carve-out of Pipeline360, where he remains a board advisor, and now runs Uphoff Advisory. His October 7 post, "AI Stopped Being a Story. Now It's in Your Multiple," makes the core argument: when AI shows up in company valuations, it has stopped being an event and become a trend.

What does the buyer data show?

Tony's starting point is a sequence change in B2B buying: "The machine drafts the shortlist. The human validates it."

G2's March 2026 survey of 1,076 B2B software buyers supports him:

• 51% now start research with an AI chatbot more often than with Google, up from 29% in G2's 2025 report.
• 69% said a chatbot led them to choose a different vendor than planned.
• One in three bought from a vendor they had never heard of before.

"For two decades the scoreboard was discoverability. SEO taught every marketing team to rank, earn the click and get found. Most companies carried that scoreboard straight into AI," Tony said. Teams ask a model about themselves by name, get a correct answer, and conclude they are covered. Tony calls this a false positive. Consideration, not awareness, is the metric that shows who is winning.

Why do engines describe the company you used to be?

The second finding matters to anyone who has survived a rebrand, merger or leadership change. "All three engines that described me still call me CEO of Pipeline360," Tony said. Only one mentioned the firm he runs now. The sources were org-chart aggregators, executive databases and old press releases. His rule: answer engines default to what the web has said about you most often, "and most often usually means longest ago."

Few teams watch this. A Corporate Ink study reported by eMarketer in May found only 29% of marketers actively track and measure AI visibility.

His test also carried a content strategy lesson. Asked "What is Demand-as-a-Service?" all three engines defined it accurately using Pipeline360's own pages — and never mentioned Tony, who helped create the strategy. "The models credit whoever publishes the clearest explanation, not whoever had the idea first," he said.

Who builds the shortlist?

When Uphoff Advisory did appear, ChatGPT ranked it second of four for "Who advises founders of B2B information and data businesses?" Broad questions returned nothing. The citations explain why: the models cited roundups from Zapier, Content Allies and others, not the firms they named.

A May 2026 study by Maryanna Franco and João da Silva on Friction AI's platform, covering 14,140 queries about activewear brands, found the same mechanism at scale. Nike and New Balance share the same Google Knowledge Graph description, yet Nike appeared in roughly 71% of athleisure recommendation answers while New Balance almost never did. Nineteen prompts over one week is a snapshot, and answers shift run to run. Tony plans to rerun the same prompts in January and publish the numbers either way; tracking weekly would make Citation Share of Voice meaningful as a trend line.

What does this change for valuation and the CFO?

Tony's Q3 2026 Uphoff Value Index now scores AI Disruption Protection, including this question: "Relationship ownership. Do you own the customer relationship, or rent it from search and platforms?"

Collingwood's 2026-27 Market Report, surveying about 100 active acquirers, found 91% call the AI resilience of a business model essential or important, and 74% said the same of a direct customer relationship. In Tony's worked example, two data businesses with identical P&Ls are valued at 5.6 times EBITDA ($28 million) and 4.0 times ($20 million). The only difference is whether buyers believe AI can route around them.

His October 19 post turns to revenue per employee using Block: $10.36 billion of gross profit with 10,205 people in 2025, about $1 million per person, versus raised 2026 guidance of $12.51 billion with just under 6,000 people, about $2.1 million per person. The average marketing agency does about $163,000 of revenue per employee; the median public software company about $395,000. In his index, two agencies with identical P&Ls, one with 68 people and one with 40, sit about $1 million apart in value.

Tony ranks signals from honest to performative: contract value, expansion revenue and churn near money; form fills and self-reported data at the cheap-to-fake end. "A lead score isn't a fact about someone's business. It's a fact about how well your content ranks," he said.

How should teams apply this?

Tony's October 14 post, "Your AI Pilot Doesn't Have a Job," borrows Darshan Deshmukh of ProcureAbility's phrase, "death by a thousand pilots." He cites BCG's 2024 finding that roughly 70% of AI value challenges come from people and process, 20% from technology and 10% from algorithms. His Application Test requires three gates before spend: state the job without a vendor name, name the owner and baseline, and set the one signal to check in 90 days with a pre-made scale, fix or stop decision. "Track our AI mentions" fails the first gate.

His five-move sequence, adapted for SEO teams:

• Add a recommendation rate to the scorecard and report it first. Write 10 to 15 buyer-style prompts plus two or three branded ones, run them weekly across all four engines, and track list inclusion and description accuracy.
• Fix the record the models read. Update bios, LinkedIn and press pages; request corrections from aggregators; assume models have not heard about any rebrand.
• Own the narrow phrase. Publish the canonical explanation of the buyer's most specific problem, then target the 10 to 20 roundup pages the engines cite — a digital PR job.
• Translate visibility into owned relationships. Tie recommendation rate to signups, demos and pipeline.

Teams still reporting rankings first will spend 2027 explaining a pipeline gap their dashboards never showed them. The models are drafting the buyer's list now; the practical question is whether any dashboard measures whether you are on it.