Brand advertising
A Socratic walk-through of brand advertising — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a firm spend fortunes on advertising that says nothing about the product?
A brand, stripped of the design work, is only a name attached to goods — a way of saying this came from the same source as that. It tells you nothing about the contents of the tin. Yet firms defend a name with an energy they never spend defending a recipe, and buyers pay more for it than for a chemically identical own-label tin beside it. What is the name doing that the tin cannot do for itself?
Reasoning it through
REASONING #Start with a market where nobody knows who anybody is. A seller offers a tin of something you cannot inspect. He has every reason to cut the contents and keep the difference, because tomorrow he is indistinguishable from every other seller. You know this, so you will not pay for quality; so nobody supplies it. The failure is not dishonesty — it is that misbehaviour has nowhere to land.
Now give each seller a name he cannot shed. Consequences become attributable. Cut the contents and the loss shows up somewhere specific: buyers who leave, buyers who tell others, a premium that evaporates. The name is what collects the results of your conduct and hands them back to you.
But collecting consequences only bites if there is something to lose, and a name nobody has heard of collects almost nothing. So ask what a valuable brand is worth to its owner in cash: the stream of future sales at a price above what an unknown rival could charge. That stream exists only so long as the promise is kept. Benjamin Klein and Keith Leffler put the point precisely — the premium is not a rip-off but the mechanism, since a firm earning a rent on its reputation loses that rent the moment it is caught cheating, and the rent must exceed what cheating would gain.
Notice what has happened. The firm has posted something valuable that its customers can seize — not a deposit they receive, but a value they can destroy by walking away. Economists call this a hostage, and it is the strange fact at the centre of branding: the brand is credible precisely because it is vulnerable.
That reframes the advertising. If the campaign were trying to tell you something, its emptiness would be a puzzle. If its work is instead to enlarge the hostage — to make the name widely known, so that a public failure is publicly costly — the content matters far less than the scale. The firm is not describing its quality; it is enlarging the amount it stands to forfeit.
Does the theory predict anything awkward? It does, and this is where it earns its keep. If the brand is collateral, lending it out is dangerous. Extending a trusted name to a new product borrows credibility the new product has not earned, and if the new line disappoints, the damage travels back up the name to everything else sold under it. The collateral is pledged twice. Work on extensions by David Aaker and Kevin Keller found that they succeed where buyers perceive a fit between the parent and the new category and struggle where they do not — what you would expect if the borrowed asset is a specific promise rather than general goodwill.
The analogy
THE ANALOGY #Think of a security deposit on a flat. Nothing in it describes what kind of tenant you are, and the landlord learns nothing by counting it. What makes your promise credible is simply that a sum sits where you cannot reach it and he can, and that it is large enough that wrecking the place would cost you more than it gained.
A deposit is a fixed amount held by one identifiable party who can hand it back, whereas a brand's value is held by nobody, cannot be returned, and is not transferred to the wronged customer but destroyed — which is why a firm can lose everything without a single buyer being made whole.
Clarifying the model
THE MODEL #Three refinements keep this honest.
First, the hostage only works if buyers eventually learn the truth. For goods you can judge after use, that holds. For credence goods — a repair you cannot evaluate, a diagnosis, a pension recommendation — you may never discover you were shortchanged, and the mechanism weakens accordingly. Regulation and licensing cluster in exactly those markets, which is a mark in the theory's favour.
Second, an asset that can be built can also be cashed in. A firm expecting to exit can quietly cut quality and live for a while on a reputation it is no longer earning — the last-period problem. What partly repairs it is that brands are saleable: the price a buyer will pay capitalises the reputation, so running it down destroys value the owner would otherwise pocket.
Third, this account sits alongside others rather than replacing them, and testing it is hard. Advertising also makes a brand come to mind at the moment of choosing, and repeated exposure alone shifts how favourably people rate things. Which of these does more of the work in a given market is unsettled, and the measured relationship between advertising intensity and independently assessed quality has been weak and inconsistent.
A picture of it
THE PICTURE #How to readFollow a name, not a product. It begins with nothing worth protecting and can only reach Trusted by spending first and delivering afterwards — there is no shortcut edge. Note how cheap failure is on the left and how expensive it is on the right: the whole value of the brand is the size of the drop. The two edges out of Extended are the risk of brand extension, since the same arrow that carries credibility down to a new line can carry damage back up. Harvested is the last-period problem, and the back-edge from Damaged shows that recovery is not a repair but a restart.
What became clearer
WHAT CLEARED #A brand is not a message about quality; it is a store of accumulated reputation that the firm can lose. That is what makes quality credible where inspection is impossible — the seller has arranged to be punishable, and the advertising's real function is to make the punishable amount large. It also explains why extending a brand is genuinely risky rather than merely lazy: you are pledging the same collateral against a second promise you have not yet kept.
Where to go next
ONWARD #- Why supermarket own-label products, which post no hostage, can still charge more than the cheapest tier.
- How online platform ratings substitute for brands by making a small seller's history attributable.
Key terms
TERMS #| Term | What it means |
|---|---|
| Hostage | something valuable a party puts at risk to make its own promise credible, since breaking the promise destroys it. |
| Quality-assuring premium | a price above cost that generates a rent stream a firm forfeits if caught cutting quality. |
| Credence good | one whose quality the buyer cannot judge even after purchase and use. |
| Brand extension | selling a new product under an established name, borrowing its reputation and exposing it. |
| Last-period problem | the weakening of reputational discipline when a party expects no future dealings. |
Every term the collection defines is gathered in the glossary.