Ten phone calls will tell you what your position actually is. A dead radio fad from 1976 will tell you why.

Your positioning is not a document. It is the sentence one distracted person says to another over dinner, somewhere between the entree and the check, when a name is needed and yours comes up. That sentence was written by someone half-listening on a kickoff call eighteen months ago. You did not approve it, you will never see a draft, and it is outselling your website.

1. THE TEN-CALL AUDIT: RUN IT THIS WEEK, IT COSTS YOU NOTHING

1.1  Call ten past clients. Ask exactly one question: “If a friend at dinner said they needed someone like me, what would you tell them?”

1.2  Then stop talking. Do not clarify. Do not correct. Do not rescue the pause. The silence is the instrument.

1.3  Write down every noun they use, verbatim, including the wrong ones. Especially the wrong ones.

1.4  Tally them. The noun that appears in six or seven of ten calls is your position. The gorgeous phrase from your homepage that appears zero times is a hobby you have been funding.

What you hear back	Times in 10	On your website?	Verdict
Noun A (example: “the permitting guy”)	7	No	This is your real position. Start here.
Noun B (example: “construction stuff”)	4	Yes	Too broad to survive a retelling. It will keep drifting.
Your official positioning line	0	Yes, twice	Beautiful, expensive, and dying in transit every single time.

TABLE 1. The tally sheet. Ten calls, one column of nouns, zero interruptions from you.

2. BARTLETT PROVED YOUR MESSAGE HAS PHYSICS, AND HE DID IT IN 1932

Frederic Bartlett sat twenty Cambridge students down with an unfamiliar folk tale and passed it along a chain, one person to the next. It opened at about 330 words and was down to 180 by roughly the seventh retelling. Length was the boring half of the finding. The interesting half: the story shortened and rationalized at the same time. Odd details got smoothed into familiar shapes, supernatural elements quietly evaporated, and anything a British undergraduate could not place got explained into something he could.

The modern replication is more brutal. Mesoudi and Whiten (2004) ran forty participants through ten four-person chains. One published chain opened at 130 words and closed at eleven: ninety-two percent gone in four hops, with nobody in the chain being careless.

Retelling	What the person actually wrote (published chain, abridged)	Words	Survives
Original	Rachel parked outside the supermarket, got a trolley, checked her list, went down the aisles, joined the fastest line, unloaded onto the belt, the cashier rang up the total, she paid, got change, bagged it, loaded the trunk…	130	100%
First	Rachel went shopping, parked, got a trolley, collected the food she wanted, paid the cashier, got change and a receipt, drove off.	59	45%
Second	Rachel drove to a supermarket, parked, got a trolley, chose some food, paid the cashier, got change, drove away.	41	31%
Third	Rachel drove to the supermarket, parked, got a trolley, chose some food, paid the cashier and drove home.	22	17%
Fourth	“Rachel went to the supermarket, got some food and went home.”	11	8%

TABLE 2. Mesoudi & Whiten, Journal of Cognition and Culture (2004), Appendix 1. Word counts are exact.

Now the finding that should move money in your budget. The gist survived intact, and so did sequence: exactly one of 163 mid-level statements came back out of order, along with three of 241 low-level ones. Detail died. Summary lived. Every retelling shoves your message one rung up the ladder of abstraction, and it never climbs back down.

So hand people a paragraph and they will pass along a sentence. Hand them a sentence and they will pass along a noun. Budget for one level of loss, and write the noun yourself.

3. THE CORPORATE VERSION IS FUNNIER, AND MUCH WORSE

Donald Sull, Charles Sull and James Yoder surveyed 132 leaders at a technology company with more than 10,000 employees. On the annual engagement survey, 97 percent said they clearly understood company priorities. Then the researchers asked those same people to list them. The five had not changed in two years and the CEO repeated them constantly. One quarter could name three. One third could not name one.

Across 124 organizations, 28 percent of the executives and middle managers responsible for executing strategy could list three priorities. The decline is a cliff: 51 percent of top teams converge on the same list, and 22 percent of their direct reports do. One step down the org chart cuts comprehension in half, among salaried people whose bonuses depend on knowing. Your referral sources have no such incentive.

4. BREAKER ONE-NINE: CB RADIO WAS A POSITIONING LABORATORY WITH 20 MILLION SUBJECTS

CB radio is the cleanest natural experiment in message relay this country ever ran. By May 1, 1977 the FCC had issued 9,830,122 licenses and was stamping 300,000 more a month. The Electronic Industries Association counted over 20 million actual users, up from under a million before 1970. The paperwork got so heavy the FCC gave CB applications their own ZIP code in Gettysburg, Pennsylvania (17326).

Range was three to twenty miles, so every message got relayed by strangers to strangers with no authority anywhere in the chain. The vocabulary mutated exactly the way Bartlett’s folk tale did.

Term	What it meant in 1976	What happened to it	Lesson for your positioning
“Good buddy”	The standard warm greeting between drivers.	Flipped into an insult among truckers. “Good neighbor” replaced it.	A phrase can keep its words and lose its meaning. No committee will notify you.
“Smokey”	A state trooper, after the hat.	Survived fifty years and a Burt Reynolds movie. Still legible.	Concrete, visual, and one word. That is what survives a fifth retelling.
“10-50”	Depends entirely on which county you were in.	Traffic accident in one state, officer down in another. DHS issued a plain-language guide in 2008 telling responders to drop codes.	Insider shorthand fragments the moment it crosses a border. Yours has borders too.
“Channel 19”	Nothing official. Drivers just picked it.	Became the de facto highway channel by consensus, never by rule.	The market assigns the meaning. You get a vote, not a veto.

TABLE 3. Four pieces of vocabulary, four different fates, one relay network.

Then the commercial half. Americans bought roughly 25 million CB sets between 1974 and 1977, and imports climbed from $37 million to $839 million in four years. The category collapsed anyway. Hy-Gain did $96.8 million in 1976 and $50.4 million the year after, watching its stock fall from $28 to about fifty cents. Japanese makers held orders for 8.6 million sets in July 1976 and saw 58 percent canceled. Capacity use fell from 83 percent to 29, margins from plus 8.4 to minus 27.5, and half the factory jobs vanished.

Fame arrived early. Comprehension never showed up. “CB radio” meant freedom, trucker rebellion, gadget, toy, safety tool or party line, depending on who was retelling it. When the FCC changed the channel count in 1977, the category had no defensible meaning left to fall back on. Twenty million advocates, and not one durable sentence among them.

5. THE MONEY RIDING ON A SENTENCE YOU DID NOT WRITE

Nielsen asked 40,000 consumers in 56 countries which channels they trust. Word of mouth won at 88 to 89 percent globally, and it is the only channel on the list where you do not write the copy.

5.1  Hinge Research surveyed 523 professional services firms. 81.5 percent had gotten referrals from people who never worked with them. Most of your advocates hold a copy of a copy of a copy.

5.2  Asked why they do not get more referrals, 72 percent gave the same answer: no one asked them.

5.3  51.6 percent of buyers report ruling out firms without ever speaking to them. The surviving sentence gets a vote before you do.

5.4  Schmitt, Skiera and Van den Bulte tracked about 10,000 German bank customers for 33 months. Referred customers churned 18 percent less and were worth 16 to 25 percent more over their lifetime. The margin edge faded after roughly 2.5 years; the loyalty edge never did.

6. WRITE NEXT YEAR IN THE SURVIVING SENTENCE, NOT THE FLATTERING ONE

6.1  Adopt the winning noun, even if it embarrasses you. “The permitting guy” beats “integrated regulatory advisory services” because it already survived seven retellings without your help.

6.2  Write one level below where you want to land. Every hop climbs a rung, so aim low on purpose. Say “we fix payroll for restaurant groups with more than twelve locations” and let it abstract up to “restaurant payroll.”

6.3  Attach one number. Numbers are concrete, and concrete details ride along with the gist the way “Smokey” did. “Twelve locations” survives. “Scalable” does not.

6.4  Apply the fifth-retelling test to every headline, deck and bio. Read it, wait a day, write it from memory, hand that to someone else, repeat. Whatever survives five rounds is what you are buying media against.

6.5  Feed the strangers. Since 81.5 percent of referrals come from people who never hired you, publish, speak and post in the surviving sentence. Visible expertise is the largest single driver of those referrals in Hinge’s data.

6.6  Re-run the ten calls in six months. If the tally has not moved, your marketing is talking to an audience of one, and that one is you.

7. READ THESE NUMBERS LIKE A GROWN-UP

A few honest asterisks, since a piece about message distortion should not distort its own evidence. Bartlett published sample protocols rather than aggregate statistics, and his repeated-reproduction result went unreplicated until 1999. Nielsen measures stated attitudes. The referral study is one bank in one country. Take the direction and skip the decimal points. It has held since 1932.

SOURCES

Bartlett, F.C., Remembering (Cambridge, 1932). • Mesoudi & Whiten, Journal of Cognition and Culture 4:1 (2004). • Bergman & Roediger, Memory & Cognition 27:6 (1999). • Sull, Sull & Yoder, MIT Sloan Management Review (Feb. 2018). • U.S. International Trade Commission, Citizens Band (CB) Radio Transceivers, Pub. 852 (Feb. 1978). • DHS/FEMA, Plain Language Guide (July 2008); NIMS Alert NA 06-09. • Hinge Research Institute, Referral Marketing for Professional Services Firms (523 firms, 2015-16). • Schmitt, Skiera & Van den Bulte, Journal of Marketing 75 (Jan. 2011), 46-59. • Nielsen, Trust in Advertising (2015 and 2021).