In 2012, one American retailer lost 985 million dollars in a single year. Sales fell 25 percent. Revenue dropped from about 17.3 billion dollars to 12.98 billion, the lowest since 1987, erasing 4.3 billion dollars in one trading year. The stock roughly halved. The chief executive who caused it, a man who had just built the most profitable shop in the world, was gone in seventeen months.
The retailer was JCPenney. And the strange part is what broke it. The new boss did not overspend on promotion. He switched the discounting off. To understand why that was fatal, and why discounting is roughly 8% of marketing but does most of the damage, you have to go back to a time when nobody could measure any of this at all.
The short version. Advertising and promotion are roughly 8% of marketing: one of four Ps, inside one of three stages. They are also the easiest part to measure, so they win the budget and slowly starve the brand. Kraft, JCPenney and Coach paid for it. Apple never did.
Act one: the age of feeling, when marketing could not be measured
For most of advertising’s history, you could not tell what worked. The department-store magnate John Wanamaker is supposed to have said it best, around the turn of the last century: “Half the money I spend on advertising is wasted. The trouble is, I don’t know which half.”
That was not a complaint. It was the operating condition of the whole business. In the era we now picture as Mad Men, advertising was a craft of feeling because feeling was all there was. You could not follow a shopper from a television spot to a till, so the great agencies bet on emotion, wit and distinctiveness, and they defended those bets with conviction rather than data. Bill Bernbach, the man behind Volkswagen’s “Think Small” in 1959 and Avis’s “We Try Harder” in 1962, put the creed plainly: persuasion, he said, “happens to be not a science, but an art.”
Those campaigns would never survive a modern dashboard. “Think Small” told Americans to buy a smaller car in the age of the tail fin. No metric could have approved it. It ran on nerve, and it built one of the most valuable car brands on earth. The age of feeling was romantic, and it was also genuinely wasteful, exactly as Wanamaker said. But it had one accidental virtue. When you cannot measure the short term, you are forced to invest in the long one. You build something that lasts, because lasting is the only return you can see.
Act two: the age of measurement, and how discounting erodes a brand
Then the counting began.
When supermarket scanners spread in the early 1980s, marketers could suddenly run controlled experiments on real purchases. The branding scholar David Aaker watched what happened next. Every time a brand ran a cents-off deal, sales jumped, visibly, tomorrow. So, in his words, “they put all their money into promotion and they took it out of image advertising. They destroy brands. The Kraft brand was so destroyed it took them three years to bring it back.”
Look at the trap in that sentence. The promotion showed a wonderful return. The brand advertising showed almost none. And the brand that chased the wonderful return nearly killed itself. The measurement was working perfectly. It was measuring the wrong thing.
Here is the mechanism, named plainly. Marketing has three roughly equal stages: diagnosis, strategy and tactics. Tactics is the four Ps, product, price, place and promotion. Advertising and promotion are one P inside one of the three stages. One third times one quarter is about 8 percent of the machine. The frame comes from Andrew Tindall at System1, built on the marketing economics of Les Binet.
Eight percent is not nothing. Great promotion earns its place. But promotion is the one part of the 8 percent you can measure cleanly and fast, so it always looks like the smart spend, while the slow work of building a brand has no tidy number this quarter and always looks like the soft spend. So the money migrates, quarter after quarter, toward what the dashboard rewards, and the brand thins out underneath. Marketers have a name for the cost of that thinning. They call it the dullness tax: the price you pay, in attention and in money, for a brand that no longer means anything beyond its next offer.
Scanner data was only the start. Digital turned the trickle of measurement into a flood, until every click and every return on ad spend could be counted to two decimal places. And the more you could measure, the more the meaningful got demoted to make room for the measurable.
Act three: what the end of the trap looks like (the JCPenney collapse)
Which brings us back to JCPenney. For decades it had run a permanent sale: mark the price up, mark it down, coupon after coupon. Every quarter the promotions drove the traffic, and every quarter the numbers said keep going. It worked, in the way the trap always works, and it slowly hollowed the brand out until the discount was the only reason anyone walked in.
In 2011 the board hired Ron Johnson, the man who built the Apple Store, and he saw the addiction clearly. He scrapped the fake markdowns for honest everyday prices. On paper, the right diagnosis. In practice, a patient going cold turkey with nothing left in the tank. Customers did not feel honesty. They felt the loss of the small win that behavioural economists call transaction utility, and the loss of the price anchor they used to judge value at all. Take those away and there was nothing else. No brand underneath. Only the sale.
The fall was instant: comparable sales down 25.2 percent for the year, a 985 million dollar loss, the stock from about 43 dollars to under 18 (Bloomberg; Forbes; Observer). Johnson admitted it before the board removed him. The customer, he said, “loves a sale, at times she loves a coupon, and she always wants a reference price.” And here is the detail that should stay with you. When JCPenney restored the discounts, one analysis found it was often charging higher prices than during the honest-pricing experiment, simply framed as markdowns from an inflated anchor. The customers cheered. The company had proven, at a cost of nearly a billion dollars, that its brand sold nothing except the feeling of a discount. That is not a pricing story. It is a brand that promotion had already eaten, years before anyone arrived to read the will.
The trap is reversible, and the reversal is just as measurable. Coach spent the 2000s chasing distribution into more than a thousand department stores, lost control of its own price, and got discounted into a mass label. So in 2014 it deliberately pulled its bags out of the discount channel to rebuild the one thing a luxury brand actually sells, which is that not everyone has one. The recovery shows in the numbers: handbags at 400 dollars or more went from about 30 percent of sales in 2015 to 40 percent in 2016 to more than 55 percent by 2017, while the old logo bags fell below 5 percent of its North American shop sales (SAGE Business Cases; CNBC). And Apple simply never entered the trap. It does not run a mid-season sale. It holds what the pricing expert Ron Baker calls pricing integrity: the price reflects the value, and it is never rescued by a coupon. A brand that never teaches its customers to wait for the discount never has to survive the withdrawal.
Coach’s climb out of the discount channel, in one line:
Source: SAGE Business Cases; CNBC. Over the same period, its old logo bags fell below 5% of North American shop sales.
Act four: the age of the machine, and what AI changes
Now the counting is about to become total, and that changes the stakes.
Artificial intelligence is a measurement engine of a kind Wanamaker could not have dreamed of. It optimises a campaign toward return on ad spend in real time, attributes a sale across a dozen touchpoints, and finds the cheapest possible click before you finish your coffee. Every pull that dragged the budget toward the measurable 8 percent for the last forty years just got a thousand times stronger. The machine is very good at the short term, and the short term is exactly what erodes the brand.
There is a second edge to it. AI has made sameness cheap. When every company runs the same models on the same prompts, they converge on the same voice, the same look, the same competent, forgettable copy. The dullness tax stops being a slow leak and becomes the default setting of a whole market. The thing no machine can measure, and no machine can copy, is the distinctive feeling of a brand a buyer already trusts.
And that feeling is now, quietly, the thing that gets you found. Buyers increasingly ask an assistant who runs their category before they ever run a search, and the answer is a position no promotion can buy. You earn it the slow way, through a brand strong enough to be remembered and repeated. The age of the machine does not retire the age of feeling. It makes feeling the last defensible asset you own.
What to do about it
The lesson is not “stop promoting.” Promotion is 8 percent of the job and it earns its keep. The lesson is to stop letting the measurable 8 percent starve the 92 you cannot count as cleanly, especially now that the machine will optimise the 8 percent harder than any human ever could.
Ring-fence the brand. Protect a budget for the slow work that builds mental availability, the degree to which your name comes to mind when a buyer is finally ready. It will never show a clean quarterly return. That is precisely why it needs protecting from the numbers that do.
Respect the 60:40 rule. Binet and Field found the durable split is roughly 60 percent long-term brand building to 40 percent short-term activation. Most companies in the trap run closer to 10:90 and call it efficiency.
Measure the unmeasured. If you only score what promotion does, promotion will always win the argument. Track the harder things too: whether people can find you, whether they remember you, whether they will pay without a prompt.
Wanamaker could not tell which half of his spending was wasted, so he was forced to build something that lasted. We can now measure almost everything, which tempts us to build nothing that does. The machine is about to make that temptation overwhelming. The brands that survive it will remember the oldest lesson in the business: the number on the dashboard looks its most convincing right up until the moment the brand behind it falls over.
Does discounting damage a brand? Not in a single sale. It damages the brand over time by training customers to buy only on promotion, which erodes pricing power and mental availability. JCPenney is the clearest case: decades of couponing left it unable to sell at full price at all.
What is the 8% figure? Advertising and promotion are one of the four Ps (product, price, place, promotion), inside one of the three stages of marketing (diagnosis, strategy, tactics). One third times one quarter is roughly 8 percent of total marketing.
Brave New, plainly
Brave New is the strategy-led growth partner for high-stakes B2B. We build brands that get found, remembered, believed and valued.
What we believe: strategy is the moat and story is the weapon. AI made sameness free, which makes distinctiveness the most expensive asset in your category and the only one that compounds. So the machine does the volume and the versions, never the thinking. As Google and BCG put it in 2026, AI is an engine for growth, not a replacement for talent. The strategy holds the reins.
And what we refuse to do. We don’t do boring. We don’t do vanilla. We don’t do borrowed voices, templated copy, safe work that reads like the category, or dashboards that flatter while the brand quietly fades. Dull is a tax. Our clients don’t pay it.
Insight, with no mercy for the dull. If your brand sounds like everyone else, that is the problem we exist to fix.
Read next: “The Long Game Is a Pricing Strategy”, on why brand building is the only reliable route to a firm price. And “How to tell your marketing budget is going south”, the seven warning signs.
Sources: David Aaker, “20 Principles That Drive Success”; System1 / Andrew Tindall and Les Binet on the structure of marketing; the Wanamaker line as widely attributed; Bill Bernbach and DDB on advertising as art; JCPenney figures via Bloomberg, JCK, Forbes and Observer; Coach via SAGE Business Cases and CNBC; Ron Baker on pricing integrity; Binet and Field on the 60:40 rule.
