Is Your Message Easy to Repeat? Why Everyone Tells a Different Story, and AI Gets Your Value Wrong

The scene I'm in this week

Last week a founder forwarded me a screenshot and asked me to tell him why it made him want to throw his laptop.

It was a Slack message. His biggest customer, a genuinely happy one, two years in, renewed twice, answering a peer who'd asked 'hey, who do you use for this?' And the customer's answer about what this company does was wrong. Not hostile-wrong. Not angry-wrong. Fuzzy-wrong. He described them as 'kind of a dashboard thing for ops teams,' which is about a third of the truth and the least interesting third.

The founder was ready to blame the customer. Then I asked him a harder question. Forget the customer for a second. Ask your own three top salespeople to describe what you do, separately, and see what you get. He did it that afternoon. He got three different companies back. One led with the AI. One led with the integration. One led with the compliance angle. All three were technically true. None of them matched.

That's when it clicked for him. His customer didn't invent that fuzzy description. He inherited it. He'd been sold by a rep with one story, onboarded by a team with another, and emailed by marketing with a third. He repeated the blur he was handed. The message didn't break in the hallway. It broke inside the company, long before it ever reached him.

Naming what's actually broken

Here's the thing founders miss. Your message doesn't fail when a customer garbles it. It fails when everybody who touches your company is carrying a different version of it.

Walk the building. Your sales reps freelance the pitch, each one leaning on whatever they personally think closes. Marketing polishes a version for the website that nobody in sales actually says. The founder has the real, sharp story in his head, the one that never got written down, so it walks out the door every time he's not in the room. Your partners and resellers repeat a half-remembered rumor of any of it. And your customers, at the very end of that chain, get the average of all that noise. That's the version they repeat to the next buyer. That's the version they type into a review.

Count the tellers. Ten employees, a handful of partners, a few hundred customers, and every one of them is a broadcast tower pointing a slightly different signal at the market. You don't have a message. You have a rumor mill with your logo on it.

I call the fix the repeat test, and it's brutal because it strips out everything you can't control. In a demo you've got your deck, your voice, your comeback for the follow-up. In the retelling, whoever's carrying it has one sentence and a bad memory. If the story can't survive being told by your newest sales rep, or your most distracted partner, or your happiest-but-vaguest customer, you don't have a wording problem. You have a clarity problem at the level of who the company has actually decided to be. This is just truth: you can't make a message repeatable when the company underneath it never agreed on the story in the first place.

Why this is worse now than ever

For most of business history, a scattered internal story was a slow, quiet leak. Now the whole market reads every version at once.

The old way, the damage stayed local. A rep pitched it wrong on a Tuesday, a partner muddled it at a dinner, a customer fumbled a referral, and you never found out. The versions never met each other. AI ended that. ChatGPT, Claude, and Perplexity don't read your best salesperson's pitch. They read everything at once: your homepage, your reps' LinkedIn posts, your partners' listings, your customers' reviews, your old press. Then they average it and hand the buyer one confident summary before a human at your company ever gets a turn.

When all those internal stories disagree, the machine has no stable signal to lock onto. It hedges. It averages you into the category. Or it just names the competitor whose people all happened to say the same thing. AI brought the cost of content to zero, so volume isn't the moat anymore. Consistency is. Brand is the new backlink: in AI search, a clear and consistent narrative across every surface is what gets you cited, the way backlinks once drove rankings. A company where everyone tells a different story isn't sending a signal. It's sending static, and the machine tunes static out.

The diagnostic: run this on your company this week

You don't need to hire anyone to find out where you stand. Run these three tests.

1. 1The ten-tellers test. Ask three salespeople, one marketer, one partner, and three customers, separately, to tell you what you do, who it's for, and why you and not the other guy. One breath each. Don't coach them. Line up the eight answers. The number of genuinely different companies you see in that list is the size of your problem. If it's more than one, you found the leak.
2. 2The machine test. Paste your homepage into ChatGPT and ask it: 'What does this company do, who is it for, and when should I recommend it over a competitor?' Read the answer as if you were the buyer. That fuzzy, hedged, half-right paragraph is not the AI being dumb. It's a mirror. It's the average of every scattered signal you've put into the world, reflected back with nothing added.
3. 3The one-sentence test. Get your leadership team in a room and have each person write what you do in seven words or fewer, no jargon, without looking at each other. Compare. If the founders and the head of sales can't produce the same sentence, no rep, partner, customer, or AI ever will. We call that a 7-Word Verbal Identity, and most companies have never once made everyone write it down together.

What I see across 200+ B2B companies

The companies that get repeated aren't the ones with the best product. They're the ones where everybody tells the same story.

I've now sat inside more than 300 founder engagements, and the pattern holds every time. In the repeatable companies, you can ask the CEO, a two-week-old sales hire, and a channel partner the same question and get the same answer back, because they gave everyone one story and made it stick. In the stuck companies, every person is improvising off a homepage that lists nine capabilities and stands for none of them. April Dunford, who wrote the book on B2B positioning, makes this point cold: if the people around your company can't explain what you do to someone else, your positioning has failed, no matter how good the product is. Positioning lives or dies in the retelling, and the retelling happens hundreds of times a week without you in the room.

The AI layer just made the cost visible. The companies ChatGPT recommends by name are the ones it can summarize with confidence, and that confidence is built from consistency across every source it read. That's what an Army of Answers is: the deliberate footprint of clear, consistent answers a brand seeds across the web so the machines cite it when buyers ask. It only works when everyone is already telling the same story. Seed ten different versions across a hundred pages and you've just taught the machine your confusion at scale. Untrained AI already produces trendslop, the confident-sounding mush that describes half your category. A company that can't agree on its own story doesn't fight the trendslop. It becomes it.

A real example

A cybersecurity company, roughly $18M in revenue, came to me losing deals they should have won.

I didn't start with the homepage. I started by asking eight of their own people what the company did. I got eight answers. Sales led with speed, the founder led with the threat model, marketing led with 'AI-powered platform for modern security teams,' which describes about four hundred companies. Then I asked five of their customers, and got five more versions, two of which were basically describing a competitor's product. Then I asked ChatGPT who to consider for their exact use case, and it named three rivals and not them. Of course it did. There was no 'them' to name. There were thirteen different thems.

We ran the Magnetic Messaging Framework on them. Picked the one buyer they actually win with. Named the old way they were fighting against. Cut the nine capabilities down to the one transformation that mattered and wrote it in seven words. Then, and this was the real work, we got it into every hand: the sales team's talk track, the partner one-pager, the homepage, the onboarding deck, the founder's LinkedIn. Within a quarter, you could ask anyone in their orbit what they did and get the same sentence back. Customers started repeating it. The machine started putting them in the answer. Same product. One story instead of thirteen. That was the entire difference.

What changes when everyone tells the same story?

What this means for you

If any of this landed, the move isn't to go rewrite your headline tonight. A sharper headline on a company where everyone still tells a different story just gives everyone a cleverer thing to get wrong.

The fix is upstream. It's deciding the story once, in a form every single person can carry, and then getting it into every hand until your reps, your marketers, your partners, your customers, and the AI are all telling the same one. That's what the Magnetic Messaging Framework does. It's your business bible: who you're for, the problem you own, the villain you fight, the point of view only you hold, written down so the whole company stops improvising and starts repeating. Get that right and the repeat test stops being a threat and becomes your distribution. Every teller carries you intact. Every AI answer names you. That's why it matters, and it's why a scattered story quietly costs you deals you'll never even see.

Here's where to start this week:

1. 1Run the ten-tellers test. Ask three reps, a marketer, a partner, and three customers what you do. Line up the answers. Count the different companies.
2. 2Run the machine test on your homepage. Read what ChatGPT tells a buyer about you. Sit with how much of it came from your own scattered signals.
3. 3Get your leadership team to each write your 7-Word Verbal Identity, separately, then compare. If the top of the company can't agree on the sentence, that's the work, and it's exactly the work worth doing.