Two finalists, the same three words, and the client who kept mixing them up

Years ago I sat in on a decision I was not supposed to influence. A client of mine was choosing between two finalists for a fractional CMO seat. Both were genuinely excellent. Real track records, real judgment, the kind of people you would hire again.

And she kept mixing them up.

Not their names. Their value. Both had described themselves the same way, the same three words on the same kind of headline, so by the second round she could no longer hold onto what made either of them different. She chose on a coin-flip of chemistry. I am fairly sure the better fit for the actual problem lost. Not because she was weaker. Because nothing about how she described herself let the client keep hold of what set her apart.

That is the whole problem in one room. And the machine your buyers now use makes it sharper.

When a buyer asks an AI tool who to hire, it does not return the best expert. It returns the one it can most clearly place.

Answer engines do not browse the way we picture. They sort. Before ChatGPT, Perplexity, or Gemini writes a single word, it files the buyer’s question into a category, then pulls the sources that clearly belong to it. If your public positioning does not name a category the engine recognizes, you are hard to retrieve and easy to skip. You are not being judged as not good enough. You are being judged as not placeable.

So it is kind of like a librarian. Before she can hand someone the right book, she has to know which shelf it lives on. If your work has no clear shelf, she cannot pass it to the person at the desk asking for exactly what you do. Not because the book is weak. Because she does not know where to file it.

We know the machine works this way because someone read the wiring. A working SEO practitioner spent several days in June 2026 reading the raw network traffic underneath ChatGPT’s replies and documented that the engine stamps every query with an internal use-case label, and every source with a source label, before it answers. It categorizes first, then speaks. (Suganthan Mohanadasan, Search Engine Journal, June 2026: “How ChatGPT Actually Picks Sources.”) And the academic work agrees clarity gets rewarded: a Princeton-led study found the right visibility moves can lift a source’s presence in generative-engine answers by up to 40%. (GEO: Generative Engine Optimization, Aggarwal et al., arXiv 2311.09735, KDD 2024. A research benchmark, not a promised result for any one person.)

You did not get rejected. You got averaged.

You are genuinely senior. But your public positioning reads as a job title, not a category. “Fractional CMO.” “Fractional CFO.” “Leadership advisor.” Ten other capable people hold the same role, with the same three words, on the same kind of headline. So when a buyer asks an AI tool what to look for, the tool reaches for the cleanest, most generic summary of the role, and you get folded into it. The title did not lose you to the worst person in your field. It lost you to the average one.

Here is why this stings, and why it is not your fault. This is not a competence problem, and it is not a volume problem. You can post four times a week and still be invisible in the way that matters, because posting more of a generic category does not make you a distinct one. A clear category is not marketing gloss. It is the thing that makes your expertise retrievable: by an AI tool, by a referrer describing you over coffee, by a buyer trying to remember why you stood out. Nothing gets chosen until it can first be placed.

Name and claim one distinct category this week. One buyer, one situation, one consistent set of words.

This is the next rung of The Chosen Authority Visibility Stack™. Issue #1 asked the first question: do you appear at all? This one narrows it. When you do appear, do you appear as a category, or as a generic role? The move that closes the gap is the same one the research rewards: stop publishing your title and start publishing your category. Do it in four steps, in order.

• Pick one lane you are genuinely best at. Not the widest. The sharpest.
• Name it in five or six plain words. Not “fractional CMO,” but “the embedded growth-stage marketing leader for B2B SaaS before a Series B.” A buyer can repeat it. A machine can file it.
• Name the buyer and the situation inside it. A category with no buyer is still just a label.
• Publish the same words everywhere. Your site, your headline, your bylines, your proof. A category the machine sees in one place it ignores. A category it sees everywhere it learns.

Four buyer-style prompts. Run them in order. For each, you are watching for one pattern: does the AI hand back a definition, or a name. Results shift by tool and by day, so treat this as a snapshot and re-run it monthly.

• 1 · Role-versus-adjacent: “What’s the difference between hiring a fractional CMO and a marketing consultant, and how do I know which one I need?”
• 2 · Specific-situation compare: “Compare the types of fractional CFO a B2B SaaS company should consider right before a Series B raise. How are they different?”
• 3 · What kind do I need: “What kind of fractional COO does a founder-led services firm actually need: systems, team-building, or scale?”
• 4 · Named-category: “Who are the fractional CHROs known specifically for fixing culture and retention in fast-scaling companies, not generalist HR leaders?”

Read it in one line. A strong result draws the distinction and names a real person in a distinct lane, and that person is you. A weak result hands back a clean definition with no one named, stitched from agency blogs and comparison sites. Ask about a title and you get a definition. Ask about a category and you get names, if the names exist.

Two live tests, for the shape of it. While building this section, two of these searches were run for real. The CMO-versus-consultant question came back almost entirely as definitional comparison content, agency blogs and one advisory explainer, not one individual fractional CMO named. The founder-led COO question came back as “what is a fractional COO” explainers, rate guides, and talent platforms, naming no category owner. (Standard web search, which approximates what these tools retrieve from but is not identical to any one model’s output.)

This week, write your category in one sentence and publish it in one place. Use this shape: “I am the [specific kind] of [your role] for [specific buyer] who needs [specific situation handled].” Five or six plain words at the core. Then put it live where the machine can read it: a clear page on your site, or your LinkedIn headline and About section, in the exact same words. Not your whole brand rebuilt. One sentence, one place, this week.

You already did the hard part. You became genuinely excellent. This is the part most experts skip, and it is the part that decides whether the machine forwards you or averages you out. A title gets you summarized. A category gets you chosen. The goal is not louder. The goal is placeable, and then chosen.