What Is the Ghostwritten Shortlist?

The fundamental premise

Here's a question worth sitting with. When your buyer decided which three companies to call, where did that list come from?

Not a spreadsheet. Not a forty-row matrix. Nobody in the room wrote a single criterion down.

They asked ChatGPT, and the answer came back already sorted, already reasoned, already cut to a handful of names, with a paragraph explaining why those names and not others.

That paragraph is the shortlist now. And somebody wrote it.

Not the buyer. Not the model, which invents nothing. The model absorbed a standard from whatever your category put in writing, and the companies that published most confidently shaped it most. By the time it reaches your buyer it has lost its author and kept its bias. It arrives with no logo on it, so your buyer reads it as research instead of marketing.

That's the whole trick. A ghostwriter takes no credit, and the words land harder because of it.

Your Magnetic Messaging Framework opens on the line that explains this better than anything: you're not losing to a better product, you're losing to a better story. That was always true. What changed is that the story now runs through a machine that repeats it to every buyer in your category, all day, for free.

Does that make sense? Because if it does, the next question is the uncomfortable one. Whose story is it repeating?

Two gates, and most companies never learn they failed either one

Clearing gate one and losing gate two is the more expensive failure, because it looks like progress.

That's the old game your MMF describes, running at machine speed. Pitch your product and you're another option. Define the shift and you own the market.

We're in that number too

Let me stop and put our own receipt on the table before I teach anything.

We publish for a living and we still failed gate one on our own home topic. The engines resolved the whole subject to tracking tools and unbranded SEO content, because that's who built the corpus. Ask a question the vendors wrote the answer key for and you get the vendors' answer back.

Here's what's on the line in a number like that. Every buyer who opened a chat window and asked who helps with AI visibility got handed a list of software. Not one of them was told there's a messaging problem underneath it. We weren't rejected in those 184 answers. We weren't in them.

That number stung for a day, then it got useful. Keep reading, because it's the worked example at the bottom of this page too.

Why the Ghostwritten Shortlist exists

Somebody has to define what good looks like in a category, and models can only work from what exists in writing.

Most companies publish descriptions. What we built, how it works, who we serve. That's inventory. It tells a model what you are and gives it nothing to sort you by. Your MMF has a name for that habit, and it's the villain of the whole framework: me-centric marketing.

Fewer companies publish standards. How to think about this problem, what separates a real solution from a partial one, which tradeoffs matter and why. That's a shortlist in advance, whether or not anyone calls it one, and it's the raw material a model reaches for the second a buyer asks it to compare.

The vacuum gets filled by whoever showed up with a point of view. That was true when the artifact was a buyer's guide PDF. It's more true now that the artifact is a paragraph inside a chat window, because the chat window strips the byline off and hands the standard over as though nobody had an interest in it.

Surface area compounds it. Industry GEO research published in May 2026 found only 12 percent of top-ranking Google pages were cited by ChatGPT or Claude. Ranking on Google and being a source an engine builds an answer from are two different jobs. A company can be perfectly visible to buyers who search and invisible to buyers who ask.

The core mechanic

Criteria are authored. Authorship is a market position. Whoever writes the standard decides what counts as good, and the model repeats it.

A criterion isn't a fact. It's an argument that got promoted to a measurement. “Number of native integrations” reads like data. It's a claim that integration count predicts success, and somebody benefits from that claim being the thing your buyer asks about.

When a model recommends three vendors and explains why, it's applying claims like that one. It didn't verify them. It found them stated clearly, in more than one place, and treated the consensus as the shape of the category.

Two consequences follow, and both cost money.

First, the criteria that actually decide deals are usually the ones nobody published. Ask your won customers why they bought and you'll hear things no shortlist mentions. They understood our problem better than we did. They said out loud what our board has been saying for a year. We believed them.

Second, satisfying somebody else's standard makes you eligible, not preferred. If the model can only say you clear the same bar as everyone else it named, your buyer arrives treating you as one of three qualified options. Then it's price.

Three tests that expose one thing

Each takes under an hour. Run them in this order. The first usually ends the argument in the room.

Run the Default List twice, a month apart. The standard moves as the category publishes, and watching it move tells you who's writing it right now.

How this compares to what the industry already says

Adjacent ideas exist. Here's who owns them.

What PitchKitchen adds is authorship, and its consequence: a criteria set is a publishable asset, so the fix is publishing, not objection handling.

Two sibling concepts need separating cleanly, because they're the pages next to this one.

The Sorting Problem is what the model does with you once it has you: recommend you with a reason, or include you as filler. This page is where that reason came from.

Brand Signal Score measures whether your published material is readable and scoreable by those engines at all. That's the instrument for gate one. This page is gate two.

Who it's for

B2B founders and CEOs at $5M to $75M in revenue selling considered purchases over $25K, in categories where buyers research long before they make contact. Your MMF calls this person the hero of the story, and describes the exact frustration: competitors with weaker products are winning, just because they tell a better story.

It lands hardest for the CRO or VP Sales watching win rates hold while average deal size slides, and for the founder who can't explain why the pipeline keeps filling with companies that already decided what they want.

It matters less for transactional or self-serve motions where nobody researches a shortlist at all.

How it's used in practice

You don't beat a ghostwritten standard by satisfying it. You beat it by publishing a better one under your own name.

Put them in the room.Criteria only work if the rep opens with them and the deck is built around them. That's the Sales-Enablement Narrative layer.

Inside a Magnetic Messaging Sprint this falls out of the same extraction that produces the Magnetic Messaging Framework. The criteria set isn't a separate deliverable. It's your point of view, formatted so a machine can repeat it.

What we did about our own 184

Related concepts in the PitchKitchen universe

Frequently asked questions