Concept · Strategy
Brand
Why customers pay for a promise they cannot inspect, and what a company must spend to keep it

A promise you can be held to.
On the morning of September 29, 1982, a 12-year-old girl in Elk Grove Village, Illinois, told her parents she had a sore throat and a runny nose. They gave her one Extra-Strength Tylenol capsule. Mary Kellerman was dead by 7 a.m. Within days, seven people in the Chicago area had died from capsules laced with potassium cyanide. Before the deaths, Tylenol held 35% of a $1.2 billion analgesic market. Afterward its share fell to 7%. Johnson & Johnson pulled 31 million bottles, spent $100 million doing it, and relaunched the product two months later in tamper-proof packaging. By mid-1983 Tylenol’s share was back to 30%, and it reached 35% by the end of the year.
A brand is the set of expectations customers associate with an offer, shaped by experience and signals and economically useful only when it changes a decision.
The molecule in the capsule never changed, and the poisoner was never found. What customers had stopped trusting was a prediction: that a bottle with this name on the shelf was safe to swallow. A brand is that prediction. It is the set of expectations people hold about an offer before they use it, assembled from every earlier encounter with the product, its packaging, its staff, its advertising and the people who recommend it. Its economic function is to let a buyer decide without inspecting, and so to skip evaluation that would otherwise cost time, money or risk.
That is why a brand is not the logo and not the campaign. A logo is the handle that lets memory find the prediction. The prediction itself is built by behavior, and it is rebuilt by behavior when it is damaged. Advertising can introduce an association or remind people of one, but it cannot repair an experience that contradicts it. Tylenol’s recovery came from a recall larger than anyone required and a package that made tampering visible, not from a slogan.
The mechanism is easiest to see in a laboratory. In a 2004 study in Neuron, researchers at Baylor College of Medicine noted that Coke and Pepsi are nearly identical in chemical composition, yet people routinely express strong preferences for one. They gave subjects both drinks in two ways, anonymously and with a brand cue. Anonymous preferences tracked activity in the ventromedial prefrontal cortex. When the brand was revealed for one of the drinks, the authors reported that brand knowledge had a dramatic influence on both stated preference and measured brain response. The label was not decoration added to the drink. It changed the drink the person experienced.
A brand therefore has economic force only when it changes a decision. Awareness alone does not. People can know a name and distrust it, or know it and confuse it with three competitors. The useful question about any brand is what specific expectation it creates, in which customers, and what those customers do differently because of it.
Forms of brand
A brand can do three different jobs for a buyer. They rest on different evidence and fail in different ways, so a claim about brand strength should say which one it means.
RecognitionA familiar name and mark let a buyer find the offer among alternatives and connect it to what they already know. Recognition is the entry ticket: it creates no preference on its own.
01
A familiar name and mark let a buyer find the offer among alternatives and connect it to what they already know. Recognition is the entry ticket: it creates no preference on its own.
TrustExperience and credible signals help a buyer predict quality or safety without inspection. Market share can reflect many things, so it is not a direct measure of trust by itself.
02
Experience and credible signals help a buyer predict quality or safety without inspection. Market share can reflect many things, so it is not a direct measure of trust by itself.
Identity and affiliationBuying or recommending the product says something about the buyer: taste, community, values. In 1985 Coca-Cola drinkers formed protest groups and wrote songs to bring back the old taste, which is identity defending itself.
03
Buying or recommending the product says something about the buyer: taste, community, values. In 1985 Coca-Cola drinkers formed protest groups and wrote songs to bring back the old taste, which is identity defending itself.
A continuum, not a switch
Brand strength differs by customer, category and risk. It may be more valuable where buyers cannot inspect quality before use or where the cost of a bad outcome is high; it lasts only as long as behavior supports the promise.
“Perception that compounds.”
Why it matters
The largest practical benefit of a brand is that it lowers what a customer must know to say yes. A buyer facing a medicine, a car or a fleece jacket cannot test every claim. Where the cost of being wrong is high and the product is hard to inspect, the trusted name does the inspecting. The logic points to health, safety, finance and any purchase that is expensive to reverse as the places where brands should matter most.
The second benefit is that a brand changes what a company is allowed to do. If it cannot decide a hard case, it is only a tagline.
The third is resilience. Tylenol’s recovery to a 35% share within about fifteen months is consistent with restored trust, but it cannot isolate the effects of the recall, new packaging, product availability or other market changes. The recall cost $100 million, on top of the sales lost while the product was off the shelf; that is evidence of a costly response, not a direct measure of customer perception.
Finally, a brand can carry meaning beyond the product. On September 14, 2022, Yvon Chouinard announced that the Chouinard family had transferred all of Patagonia’s ownership to a purpose trust and a nonprofit, so that profits not reinvested would go to fighting the climate crisis. That is the extreme case of a company treating its promise as an operating constraint rather than a marketing position. Whether customers reward it is a separate, unproven question. The point for strategy is that the company chose a structure in which it cannot easily break the promise.
Real-world examples
The same concept shows up in different ways across industries.
After seven deaths from cyanide-laced Tylenol capsules sold in five Chicago stores, J&J recalled about 31 million bottles at a cost of $100 million.…
On April 23, 1985, Coca-Cola changed its formula for the first time in 99 years, to one preferred in taste tests with nearly 200,000…
In September 2022 the Chouinard family gave Patagonia to the Patagonia Purpose Trust, which holds all voting stock (2% of the total), and to…
When it breaks
The sharpest failure is a promise the company knew was false. On September 18, 2015, the EPA said more than 480,000 cars contained software that turned on full emissions controls only during testing, letting them emit up to 40 times the allowed pollution. In January 2017 Volkswagen agreed to plead guilty to three felony counts over about 590,000 vehicles and to pay $4.3 billion: $2.8 billion in criminal penalties and $1.5 billion in civil resolutions. The buyers had paid for a prediction that turned out to be a fabrication. Tylenol’s case is the mirror image: J&J was the victim of a crime and took the loss anyway, and the name survived. Volkswagen was the perpetrator.
A brand also breaks when management measures the wrong thing. Coca-Cola’s 1985 reformulation was preferred in taste tests of nearly 200,000 consumers, and it still failed in the market, because a sip test with the name removed measures the drink but not the brand. The Neuron study points the same way: the label can change the experience. A blind test answers a question about the product without its brand cue; to estimate what the brand adds, compare otherwise similar branded and blind conditions. Coca-Cola’s own account says the tests missed the bond many consumers felt with the original formula, a limit for any survey that measures product attributes without asking what the name means.
Finally, a brand is easy to overclaim. Recognition is not trust, and trust is not a price premium. Patagonia’s ownership structure shows conviction, but it does not show how much of the company’s sales come from its stance rather than its jackets, and the same caution applies to most brand valuations. A brand promise also has a cost. J&J’s $100 million recall was a real cash outlay made on the belief that the name was worth more. Companies that cannot or will not pay when the promise is tested hold a logo, not a brand.
Key takeaways
- 01
What specific expectation does the name create for each target customer, and what choice changes because of it? Separate awareness from preference and trust.
- 02
What costly action would show that the company keeps this promise when it is tested? Treat a stated value as intent until behavior supports it.
- 03
Does the brand cue change recognition, trust, conversion or willingness to pay in a comparison with and without the cue? Measure those outcomes separately.
Sources
- Tylenol and the Legacy of J&J’s James Burke · Knowledge at Wharton, 2012-10-02. Seven deaths, five Chicago stores; 35% of $1.2 billion market fell to 7%; $100 million recall of 31 million bottles; share 30% by mid-1983 and 35% by year-end; Burke quotes on the credo and on saving the brand
- How the Tylenol murders of 1982 changed the way we consume medication · PBS NewsHour (Howard Markel), 2014-09-29. Mary Kellerman died by 7 a.m. on Sept. 29, 1982; tampering occurred after cases left the factory; share above 35% fell below 8%; tamper-proof packaging and caplet; perpetrators never found
- New Coke: The Most Memorable Marketing Blunder Ever? · The Coca-Cola Company, 2015-04-23. First formula change in 99 years; taste tests of nearly 200,000 consumers; 79 days; hotline calls 400 to 1,500 a day; the tests did not show consumers’ bond with Coca-Cola
- Neural correlates of behavioral preference for culturally familiar drinks · Neuron 44(2):379-387 (McClure, Li, Tomlin, Cypert, Montague, Montague), via PubMed, 2004-10-14. Abstract: Coke and Pepsi nearly identical in chemical composition; anonymous versus brand-cued delivery; brand knowledge had a dramatic influence on behavioral preferences and brain responses
- Patagonia’s Next Chapter: Earth is Now Our Only Shareholder · Patagonia Works press release, 2022-09-14. Patagonia Purpose Trust (all voting stock, 2% of total) and Holdfast Collective (98%); roughly $100 million projected annual dividend; Chouinard statement
- Volkswagen Used ‘Defeat Device’ To Skirt Emissions Rules, EPA Says · NPR (Bill Chappell), via KUNM, 2015-09-18. More than 480,000 cars; up to 40 times allowed pollution; TDI; $1,300 federal tax credit in 2009; buyers motivated by perceived environmental impact
- Volkswagen Agrees to Plead Guilty, Pay $4.3 Billion in Criminal and Civil Penalties · U.S. Environmental Protection Agency, 2017-01-11. Three felony counts; $2.8 billion criminal penalty; $1.5 billion civil resolutions; approximately 590,000 diesel vehicles