A false story about stolen kidneys has out-survived every mission statement ever written. Chip and Dan Heath reverse-engineered why — six design choices that decide whether an idea lives in memory or dies in the deck.
Here's the whole chapter in one line: whether an idea survives has almost nothing to do with whether it's true, and almost everything to do with how it's built — and the six design choices that make ideas stick are learnable. The rest is the toolbox, plus the reason you can't use it on autopilot.
You already know the story. A traveler has a drink with a stranger, wakes up in a bathtub full of ice, and finds a note: call 911 — we took your kidney. You heard it once, years ago, probably secondhand, and you could retell it right now without notes. It has no media budget, no distribution deal, no brand team. It is also false. And it has been circulating, essentially unchanged, for decades.
Now try to recite any corporate mission statement — including your own employer's. Those sentences were written by professionals, approved by committees, printed on lobby walls, and budgeted for. Recall rate: approximately zero. Marketing spends billions pushing messages that evaporate on contact while an unfunded lie about organ theft replicates for free. That gap is the most interesting data point in communication, and the Heath brothers built a whole book on treating it seriously.
Their move was an engineer's move: urban legends are the control group. They control for truth (the kidney heist has none), for budget (none), for authority (none — "a friend of a friend" is the least credible byline imaginable). Whatever makes them immortal must live in the construction of the idea itself. Compare the survivors and the pattern falls out: they're simple, they violate an expectation, you can picture them, they carry their own proof, they make you feel something, and they arrive as a story. Six properties — Simple, Unexpected, Concrete, Credible, Emotional, Story — which the Heaths, in full airport-book tradition, filed under the acronym SUCCESs. Forgive the acronym; the checklist underneath is real.
Programmer's version: an idea has runtime characteristics independent of its correctness. Code can be right and unmaintainable; a message can be true and unmemorable. Stickiness is the memetic equivalent of cache-friendliness — the six checks are what let an idea fit in a human's L1 and survive eviction. This chapter is a profiler for messages: where they're slow, and the specific optimizations that fix them.
Before the checklist, the enemy. In 1990 a Stanford graduate student named Elizabeth Newton ran what may be the highest-value-per-dollar experiment in this whole atlas. She split subjects into tappers and listeners. Tappers picked a famous song — "Happy Birthday," the national anthem — and tapped its rhythm on a table. Listeners had one job: name the tune.
Before tapping, tappers predicted listeners would get it about half the time. The actual rate: 2.5% — three songs out of a hundred and twenty. The gap is the finding. When you tap "Happy Birthday," the full arrangement plays in your head — melody, lyrics, candles. The listener gets knuckle Morse. And here's the cruel part: the tapper cannot switch it off. Once the song is in your head, you physically cannot experience your own taps as the listener does.
That is the curse of knowledge, and it is the villain of every marketing message ever drafted. You know your product — its architecture, its roadmap, the three years of decisions behind the pricing page. So you write "seamlessly integrates with your existing workflow" and hear a symphony. Your prospect hears taps. The curse explains a mystery from Part I: it's not that companies don't know their positioning — it's that the people writing the copy are constitutionally unable to hear it cold. The disease isn't stupidity. It's knowledge. Which is why being smarter doesn't cure it, and why every expert gets it worse, not better.
Programmer's version: you can't judge your own documentation, because you read it with the entire codebase already paged into memory. The new hire cold-starts with an empty cache, and the docs that felt complete to you are a series of dangling pointers to context only you hold. The curse of knowledge is leaking local state into a public interface. The SUCCESs checklist is the discipline that forces the flush — try the experiment on yourself first.
Simple does not mean dumbed down. It means you found the core. The model is the proverb: "a bird in the hand is worth two in the bush" has shipped unchanged for centuries, across languages, with zero maintenance — it is the most battle-tested one-liner in existence. Compact, yes, but compact and load-bearing: a complete theory of risk in ten words. That's the target — not a sound bite, a proverb.
The military figured out the mechanism. Every battle plan dies on first contact with the enemy, so officers distill each order to a commander's intent: the single sentence that must remain true when everything else changes. Southwest Airlines runs on one: "We are THE low-fare airline." Herb Kelleher's famous test — someone proposes adding chicken salad to a route; does serving it help us be THE low-fare airline? No? Then no chicken salad. The core sentence isn't decoration; it's a decision procedure that thousands of employees can execute without asking permission.
The craft skill is the journalist's: don't bury the lead. Nora Ephron's high-school journalism teacher gave the class the facts for a story — the whole faculty travels to Sacramento Thursday for a colloquium with Margaret Mead, etc. The students wrote dutiful who-what-when paragraphs. The teacher read them and said the lead was: "There will be no school Thursday." Everything you write has a no-school-Thursday sentence in it somewhere, usually in paragraph four, wearing a lanyard.
And forced prioritization is the whole game: if you say three things, you say nothing. Chapter 1 made this argument about positioning — one word in the mind — and here it is again at the level of the sentence. Programmer's version: one enforced invariant beats ten guidelines, and an API with one obvious entry point beats a rich surface nobody can hold in their head. Simplicity is what survives transmission; everything else is dropped packets.
Unexpected is how you get attention without renting it. Brains run on prediction: whatever matches the expected pattern gets processed on the cheap path and discarded. To be noticed, break the schema — then, to hold attention, open a gap. A minivan glides down a suburban street in a car ad, dad chatting, kids in the back — and a truck T-bones it mid-sentence. Cut to black: "Didn't see that coming? No one ever does. Buckle up." The ad works because it violates the one pattern you were sure of — that car ads don't crash the car.
Holding attention is a different mechanism: the curiosity gap. Behavioral economist George Loewenstein's observation is that curiosity isn't about what you know — it's about a hole in what you know, and holes itch. Pose the question before the answer. "Saturn has rings; here's what they're made of" is a lecture. "Something out there is slowly tearing Saturn's rings apart — and it took forty years to find out what" is a gap. Same content; only one gets read past the colon.
Concrete is the check most corporate copy fails hardest, because abstraction is the curse of knowledge's fingerprint. Experts compress experience into abstractions — that's what expertise is — and then forget the compression isn't shared. Nordstrom doesn't train "world-class customer service" into staff; that phrase is taps. They tell the story of the clerk who cheerfully refunded a set of snow tires — which Nordstrom has never sold. You can picture the tires on the counter. Now "world-class service" has a test case attached, and ten thousand employees can pattern-match new situations against it.
The Heaths call memory Velcro: the more sensory hooks an idea has, the more loops it catches. "Blue kettle, Denver, 9:04 a.m." has hooks; "optimized inventory outcomes" is a smooth surface sliding off a smooth surface. Chapter 15 taught the ad-copy face of this same law — Hopkins' specifics beat superlatives because specifics can be pictured and doubted and checked. Watch it work rung by rung:
Credible is the check you can't phrase your way into — belief needs somewhere to stand. The strongest floor is a testable credential: a claim the audience can verify themselves. Wendy's 1984 "Where's the beef?" wasn't a slogan about quality; it was an invitation to look at the competitor's bun and check. Chapter 15's reason-why copy is this check industrialized — every specific, checkable claim ("washed with live steam") deposits credibility that "highest quality" never touches.
Statistics can carry credibility, but only at human scale. During the Cold War, arms-reduction advocates stopped saying "30,000 warheads" — a number with no sensory meaning — and started dropping one BB into a metal bucket: Hiroshima. Then 5,000 BBs, the roar going on and on. Nobody remembers the integer; everyone remembers the bucket. A statistic's job isn't precision, it's relationship — recast it until a body can feel it. And when one vivid case can stand for the whole claim, use the Sinatra test: "if I can make it there, I'll make it anywhere." One flagship client that everyone agrees is the hardest test beats a logo wall of forty.
Emotional is the check with the most uncomfortable evidence behind it. Researchers asked people to donate to fight African hunger. One group saw the statistics — millions affected, across multiple countries. Another saw one seven-year-old girl from Mali: Rokia. The Rokia group gave more than twice as much. The brutal twist: a third group saw Rokia and the statistics — and gave barely more than the stats group. The numbers didn't add credibility; they switched off the feeling. Shift a mind into calculation mode and the part that reaches for a wallet goes quiet.
Emotional doesn't mean melodrama. It means associating the idea with something the audience already cares about — one person, not a mean; identity, not arithmetic. People don't buy a drill bit because of torque specs; Part I already made the deeper point that they're hiring progress. Feeling is the scheduler: analysis ranks the options, but emotion is what moves one of them to the front of the run queue.
The last check is the container the other five ship in. A story isn't entertainment garnish — it's a flight simulator for the listener's brain. When nurses swap stories about the infant whose monitor said one thing while the baby's color said another, every listener mentally runs the scenario — and acquires a unit of judgment without living through the emergency. Xerox repair techs learned more over shop talk about the weird machine that lied about its error codes than from the official manual, which is why smart organizations stopped treating war stories as wasted time.
Stories also do something exposition can't: they carry all six checks at once without feeling like persuasion. Jared Fogle was a college student who weighed 425 pounds and lived above a Subway. He walked in, invented his own regimen — a turkey sub and a veggie sub a day — and lost 245 pounds. A franchisee spotted the local story; the ad agency fought head office to run it nationally; sales jumped double digits. Notice the checklist audit itself: simple (eat subs, lose weight), unexpected (fast food as a diet), concrete (the enormous old jeans held up to camera), credible (a Sinatra-test case in the flesh), emotional (one identifiable person, not a nutrition table), story (a challenge plot, the oldest one we have). Nobody at Subway wrote a value proposition; they found one walking around.
The footnote is the risk clause: tie the brand to a person and you inherit the person, and Subway inherited Jared's crimes when they surfaced years later. Borrowed stories are borrowed equity — Chapter 12's loan logic, with a human counterparty. And one more boundary: this chapter treats story as a property of sticky messages. Story as an operating system — narrative frameworks, the customer as hero, your brand demoted to supporting cast — is big enough to be its own chapter, and it's the very next one.
The checklist earns nothing as trivia; it pays as an audit. Take the sentences your company actually ships — the homepage headline, the first line of the sales deck, the quote in the press release — and score each against the six checks. Not "is it good," which invites taste and rank, but "which boxes does it pass," which invites evidence. Most corporate copy scores zero to one, and the failure is always the same shape: abstract, expected, unfeeling, storyless — a smooth surface engineered by committee to catch on nothing.
Two rules make the audit honest. First, you can't score your own copy — you're the tapper, and every abstraction you wrote expands to a symphony only you can hear. Hand it to someone with an empty cache: a new hire, a customer, a smart friend from another industry. Whatever they play back to you is what the message actually says. Second, score against the best in class, not against your competitors — most category copy is a tie at zero, and beating it proves nothing. That's why the scorer below includes a control from 1961.
Compress the chapter to its own core: human attention is a lossy channel, and SUCCESs is the encoding that survives it. Simple is compression — one core, nothing to drop. Unexpected is the interrupt that gets you scheduled at all. Concrete is a shared instruction set — images run on every listener's hardware; abstractions only run on yours. Credible is the checksum, emotional is the priority bit, and story is the packaging that gets the whole payload executed instead of merely received.
The checklist also closes a loop this book opened in Chapter 1. A position is an idea that has to survive in a mind that doesn't care, doesn't listen, and won't be corrected — Part I established what should survive; this chapter is how anything survives at all. And it hands a tool to its neighbors: Chapter 13's levers decide whether people say yes, but stickiness decides whether the message is still in memory when the decision arrives. (This chapter has also been quietly eating its own cooking — every section here led with a bathtub, a tapper, or a kettle before it dared name a principle. If any of it sticks, that's why.)
One warning label before the next chapter. The kidney heist passes five checks and is a lie; stickiness is a property of construction, not of honesty, and the checklist works just as well for misinformation — which is precisely why yours can't afford to skip it. Next: story stops being one checkbox and becomes the whole architecture — the customer as hero, and your brand in the one role every founder resists.