A crown made of paper only holds its shape indoors.

Kings do not crown themselves. That part did not travel.

What traveled was the shortcut. Do not join a category, make one. And since making one starts with naming one, naming became the work, because naming is the part you can finish by Friday. Everything else in category design takes three years and a fight you might lose.
So the ritual set in. A team argues over words, picks one, ships a new homepage. The deck gets a slide that says we are not X, we are Y. The founder says it on a podcast. Sales gets new talk tracks and a laminated card.

Then the company waits for the market to show up. It does not.

What you built instead

You changed a word. You kept the same competitors.

Call it the Paper Crown. You cut it out yourself, put it on your own head, and announced a kingdom. Everyone in the building can see it. Nobody outside the building can, because a crown made of paper only holds its shape indoors.

The tell is simple and you can check it in a minute. Say your category name out loud. Now name one person who does not work for you, does not sell to you, and has never been paid by you, who has said that name back to you without prompting.

Most companies cannot name one. That is not a marketing gap. That is the whole verdict.

The most expensive version on record is a public document. In 2019 WeWork went to the market wearing technology framing. Investors read the filing, saw a business that signs long leases and sublets them short, and declined. The offering got pulled and the valuation fell from 47 billion dollars to about 8 billion inside a year. Two years later the company filed again, and that filing still described it as the preeminent brand within the space as a service category. The crown stayed on after the kingdom got repossessed.
This is not a software problem or a startup problem. Restaurants do it. Apparel does it. Beverage brands do it every time a drink becomes a wellness platform. Any company that decides the label is the strategy runs the same machinery, and the machinery does not care what you sell.

Some names do escape. Fast casual and athleisure got picked up by the press, then by rivals, then by shoppers, until the words belonged to the market rather than to whoever coined them. That is what ratification looks like, and notice who performed it. Never the originator.

Why it felt like it worked

Because everybody you asked said yes.

Your team loved the name, which proves nothing, since they helped pick it. Your friendly customers nodded, which proves nothing, since they already buy from you and would nod at a menu. Your investors said it sounded sharp, which proves nothing, since they read the same book you did.

Nobody in that group is a market. Every one of them is a mirror.

And no one gets assigned to check the outside, so nobody checks. The company grades its own claim, awards itself a passing mark, and moves to execution. This is the same failure I described in the piece on how good strategy gets killed, showing up in a different room. An idea that might have worked, judged by a process that cannot tell adoption from applause.

The verdict moved, and it moved somewhere you do not own.

This is the part that changed in the last year and a half, and it is why the Paper Crown got expensive.

Your category name lives on property you control. Your site, your deck, your posts. That used to be enough, because buyers landed on your page and read your framing first.

They do not land on your page first now.

Muck Rack’s research team ran the numbers across more than 25 million links cited by ChatGPT, Claude, and Gemini across 17 industries. Earned coverage accounts for 84 percent of AI citations. Paid and advertorial content accounts for 0.3 percent. Those proportions have held steady across three editions of the study going back to July 2025.

Read that against your category launch. The systems now describing your market to your buyers weight the sources you do not control and discount the ones you own. Your new word lives in the exact place the machine trusts least, which means it never enters the answer. So the answer keeps using the old word, and the old word comes with your old competitors attached to it.

Sell to companies and the damage multiplies. Forrester’s State of Business Buying, 2026 found generative AI searches are now the starting point for business buyers, and that a typical purchase pulls in 13 internal stakeholders and nine outside influencers. Twenty two people forming a view of your market from a summary nobody on your team has read. Sell to consumers and the count drops but the mechanism holds, because one shopper asking one question gets the same manufactured answer.

It gets worse when you do get mentioned. A Semrush study run with Kevin Indig analyzed almost 4,000 domain appearances and found 62 percent are ghost citations, where the system links your page as a source and never says your name. Your category page becomes the footnote under a paragraph recommending the incumbent you spent a year distancing yourself from.

The test. Ninety seconds.

Open ChatGPT, Perplexity, or Gemini. Ask one question.

What companies compete in [your category name]?

Use your exact category name, nothing else, no company name attached. Read the list.

Two things can happen, and they mean the same thing.

The list comes back full of the competitors you claimed to leave. That is a rename. The market placed you where it always had you, and your homepage did not get a vote.

Or the list comes back empty, hedged, or invented on the spot. That is a Paper Crown with nobody in the room. There is no outside evidence your category exists, so the system made something up, and inventions drift toward whoever has more material out there.

There is a third outcome, and it is rare. The system names your category, uses your framing, and puts you at the center of a coherent set of players. If that happens, you built something real, and your job changed from creating to defending.

Run it right now, before you finish reading. It takes less time than the meeting where you picked the name.

What the rename costs while you wait

Here is the line nobody puts in a spreadsheet.

Every sales call opens with translation. Your rep says the new category name, the buyer does not recognize it, and the rep explains it by naming the old category. Five minutes, every call, forever. Multiply that by your call volume and you have a permanent tax on the entire pipeline, paid before the conversation about your product starts. A buyer who has to translate you before they can consider you is a buyer you made work for the privilege.

Then count the rest. Analyst briefings aimed at firms whose report titles still use the old word. Paid search bidding on terms with no volume, while you go missing on the terms buyers type today. A content program teaching a vocabulary the market never requested. A booth with a category name on it that nobody walking past has heard.

None of that spend is wrong in principle. Category creation is supposed to cost more, because you are funding market education instead of lead generation. It goes wrong when you fund education nobody enrolled in.

And the meter runs in a direction most teams never check. Your competitors are not paying that tax. They kept the word buyers already use, so their search terms have volume, their analyst reports exist, and their sales calls start at minute zero instead of minute five. While you teach, they sell. That gap does not show up in a brand tracker. It shows up in a win rate, one quarter late, with no obvious cause attached to it.

Two moves, and the second one hurts

Win the old word while you build the new one. Rank and get cited for the term your buyers use today, because the category you are building does not pay rent yet and the old term does. Give it up too soon and you go dark during the exact window you needed pipeline to fund the education. This is a timing problem before it is a positioning problem, which is the same argument I made about why product market fit is survivorship bias.

Then hand your category language to a competitor, on purpose, for free.

That is the one that makes people flinch, and it is the whole game. A market with one company in it is not a market. It is a product with ambitions. You need rivals using your word, because a category with competitors in it is a category, and a category with only you in it is a slogan. Feed the language to press and analysts with no strings on it. Publish the framework, not just the pitch. If a rival puts your category name on their site next quarter, you did not lose ground. Somebody handed you a crown you did not make.

You cannot name your way out of a crowded market. You get named out of one, by people you do not employ.