The Measure Always Wins: Goodhart's Law, Nike's $28 Billion Day, and How the Dashboard Took Over Business Strategy

The content and social media marketer did not seize strategic authority. Measurability conferred it. Their numbers arrive daily, denominated, and attributable; the numbers that describe the business arrive yearly, statistically, and late. An economist named the consequence in 1975: when a measure becomes a target, it ceases to be a good measure. The last decade of marketing is that sentence, operating at the level of the org chart — awareness manufactured in enormous volume, salience quietly unbuilt, and strategy inverted around whatever the dashboard could see. Nike paid roughly $28 billion in a single day to learn it. Airbnb ran the control experiment and kept the money.

The measure that became the strategy

In 1975, a British economist named Charles Goodhart, later of the Bank of England and the London School of Economics, set down an observation about monetary policy that has since escaped its origins entirely. Any observed statistical regularity, he wrote, will tend to collapse once pressure is placed upon it for control purposes.[1] The anthropologist Marilyn Strathern later compressed it into the form everyone now quotes: when a measure becomes a target, it ceases to be a good measure.[1]

Goodhart was writing about money supply targets. He might as well have been writing about marketing departments half a century later, because the story of the last decade in marketing is the story of his law operating not on a metric, but on an entire profession's chain of command.

Consider what actually happened. A new class of marketer rose, the content marketer, the social media marketer, the performance marketer, and they rose fast, from support function to strategy table in roughly ten/15 years. It is tempting to explain this as fashion, or as generational churn, or as the platforms' marketing of themselves. The Ledger's reading is colder. This class rose because it was measurable. Its numbers arrive daily, denominated, and attributable: reach, impressions, engagement, click-through, cost per acquisition, return on ad spend. Set against that, the numbers that describe the actual condition of the business — mental availability, pricing power, share of market- arrive slowly, statistically, and without a name attached to the win.

In any organisation, the function that can show its numbers fastest wins the budget argument. Win the budget argument for enough consecutive quarters and you are no longer reporting to the strategy. The strategy is reporting to you.

The content marketer did not seize strategic authority. Measurability conferred it, one budget round at a time.

That is the phenomenon this essay is about. Not the people — the best of this class are formidably skilled, and their discipline is genuine. The subject is the instrument error underneath their authority: the strategies now being set by what dashboards can see are systematically starving the one asset dashboards cannot see.

Awareness is being manufactured at industrial scale. Salience, the thing that actually converts to revenue, is quietly not being built in the same way. And the gap between those two words is where the next decade of enterprise value will be won and lost.

Realism's down payment: the rise was rational

An argument that cannot survive its own strongest counter-case is not worth making, so here is the counter-case, granted in full.

The rise of the content and social media marketer was not a con. It was a rational response to three real changes. Distribution collapsed to zero marginal cost, which made publishing a strategic capability rather than a media buy. Attribution technology has genuinely improved, making some marketing spend accountable in a way it had never been, and accountability is a virtue this publication exists to defend. And the platforms delivered reach at prices legacy media could not touch.

A marketer who mastered those mechanics was delivering real value, and the organisations that promoted them were not fools.

Nor is the work itself trivial. Capturing in-market demand is a legitimate and profitable activity; the roughly five per cent of buyers actively looking at any moment are real revenue, and the performance discipline harvests them with genuine efficiency.[2]

Content that is distinctive, consistent and broadly seen can lay down exactly the memory structures brands are made of. None of what follows argues that social media is a mistake, that content is froth, or that measurement is the enemy. This publication's entire premise is that marketing should be accountable to the P&L.

The argument is narrower, and it is structural: the measures that elevated this class do not measure the business. They measure activity within the platforms: how many saw, clicked, engaged, converted this week. Under Goodhart's law, the moment those numbers became targets, they stopped being information and started being the objective. And an organisation whose marketing objective is the dashboard will, with perfect internal logic and green numbers throughout, optimise itself away from the asset that determines what the whole enterprise is worth.

To see how, the two words doing the work here need to be pulled apart—because the entire error lies in the space between them.

Awareness is manufactured; salience is built

Awareness is a scalar: do they know you exist? It is one number, it is context-free, and above a modest threshold it saturates; in any serious category, everyone has heard of everyone, and the metric stops discriminating exactly where the competition starts mattering.

Salience is a different kind of object. In the framework built by Byron Sharp and Jenni Romaniuk at the Ehrenberg-Bass Institute, what converts to sales is not being known but being retrieved: mental availability, the probability that a brand comes to mind in a specific buying situation, built out of memory links between the brand and the moments buyers actually enter the category — the software failing again, the audit approaching, the trainers wearing through before the marathon.[3] Salience is a network, and networks are built the slow way: by attaching the brand to the situation, distinctively, repeatedly, across the broad mass of people who are not buying today so that the brand is present in memory on the day they do.

Here is the uncomfortable mapping. Every metric the platforms surface; reach, impressions, followers, engagement. And it is a proxy for the scalar. Not one of them can see the network. A post seen by a hundred thousand people that attaches the brand to no buying situation has added a large number to a dashboard and nothing to the asset; it has made the brand marginally more known and not one degree more retrievable. The dashboard cannot tell the difference. The buyer's memory can.

Awareness is a scalar: do they know you. Salience is a network: in how many buying situations do you come to mind. The dashboard measures the first. Only the second gets bought.

So the incentive structure completes itself. The marketer is rewarded weekly on numbers that move each week, and those numbers all describe awareness-shaped activity. Content drifts toward what engages rather than what encodes. Budgets drift from brand building toward activation, a drift Peter Field spent years documenting from the IPA's effectiveness databank as the quiet erosion of long-term effectiveness under short-termism.[4]

The strategy, increasingly set by the function with the freshest numbers, becomes a machine for manufacturing the measurable thing. Nobody decides to stop building salience. It is simply that no meeting ever shows the number for it, and what no meeting shows, no budget defends.

That is Goodhart's law as organisational design. And because it operates in daily increments, each one recorded as a win, the bill arrives all at once, years later, looking like a surprise. Which brings the argument to Beaverton, Oregon.

Case study: Nike and the dashboard's four-year decade

On 13 January 2020, John Donahoe became chief executive of Nike, a Bain, eBay, and ServiceNow man, only the second outsider to hold the job, explicitly hired for his digital credentials.[5] Within months, by the account of Massimo Giunco, who spent over two decades inside Nike's marketing organisation, the transformation was announced to staff by email: the category organisation — brand, product development, sales, the machinery built around athletes and sport — would be dismantled; Nike would become a direct-to-consumer led company, ending its wholesale leadership; and marketing would be centralised, data driven and digitally led.[6]

Read that last phrase through Goodhart. It does not say the marketing would be better. It says the marketing would be measurable, and that the measurable would lead.

What followed was internally coherent and, on the dashboard, successful for years. Digital sales grew. Direct margins looked handsome. The lifestyle retros — Dunks, Air Force 1s, Jordans — sold in measurable volume now, and so were fed, while innovation pipelines that pay off on a five-year horizon quietly thinned; even the product strategy had begun optimising for what the current quarter could count.[5] And in the two moves that matter most to this essay, Nike traded away both halves of the one asset Sharp's framework says a brand actually consists of, and it could do so without alarm, because neither half appears on any dashboard:

Figure 1 — What the dashboard could not see

The consequences accrued exactly as the theory predicts: invisibly, then suddenly. Occasional buyers, the light buyers who are, in the Ehrenberg-Bass account, where all growth lives, could not find Nike where they shopped, and bought what was on the shelf instead.[6]

The memory structures went unrefreshed. And on 28 June 2024, after another cut to guidance, the market repriced the accumulated error in one session: roughly $28 billion of market value gone in a day, the largest single-day loss in the company's history, around $70 billion over nine months, on Giunco's count, with the stock at its lowest since 2018.[7] A month later Giunco published his post-mortem on LinkedIn under a title this publication could not improve on — Nike: An Epic Saga of Value Destruction — and put the diagnosis in one sentence: the CEO of Nike doesn't come from the industry; at the end, he is a poorly advised, data-driven guy.[6]

Nike did not lose $28 billion in a day. It lost it over four years, with the dashboard recording daily increments as wins.

On 19 September 2024, Nike announced Donahoe's departure and the return of Elliott Hill, a thirty-two-year company veteran from the consumer and marketplace side, a sport and brand man.[8]

The stock rose seven per cent on the announcement. Sit with that for a moment, because it is the cleanest possible statement of this essay's thesis: the market added billions of dollars in enterprise value on the news that the dashboard would no longer set strategy. The correction was priced before Hill had made a single decision.

The fairness clause, because the house discipline requires it: Nike's decline was multi-causal, product staleness, China softness and post-pandemic demand all contributed, and Donahoe navigated the pandemic's digital surge genuinely well. But the causes are not independent. The wholesale retreat, the performance rotation and the innovation drought are the same decision wearing three costumes: in each case, the option with the visible near-term number beat the option whose value was real but unmeasured. That is not three mistakes. That is one instrument error, applied consistently.

The control group: Airbnb and the $541 million that bought five per cent

Every forensic case needs a control, and marketing rarely gets one as clean as this.

In 2020, with travel demand evaporating, Brian Chesky shut off Airbnb's marketing machine. Not trimmed, shut off. Total marketing spend fell 58 per cent, from $1.14 billion in 2019 to $482 million, and the deepest cut landed on performance marketing: $541 million of search bidding and paid acquisition, roughly four times the reduction in brand spend.[9] Every attribution dashboard Airbnb owned had spent years reporting that this spend was driving the traffic. The pandemic ran the experiment no marketer would have dared to run deliberately.

Traffic held at 95 per cent of the prior year.[9]

Let the arithmetic say it plainly: $541 million of annual performance spend had been buying roughly five percentage points of traffic. The rest — the overwhelming bulk of Airbnb's demand — was arriving because of the thing the dashboard could not see and therefore could not credit: a brand so mentally available that, in Chesky's phrase, it had become a noun and a verb in pop culture.[9] The attribution machine had been systematically claiming credit for demand the brand already owned — intercepting people who searched "Airbnb" and booking the click as a conversion it had caused.

Figure 2 — Two dashboards, one error, opposite polarity

In February 2021, at the first results presentation after flotation, Chesky made the cut permanent: marketing as a share of revenue would never return to pre-pandemic levels, and what remained would rotate from performance into brand and PR, the full-funnel investment his CFO later credited, in as many words, with improving the company's overall profitability.[9][10] Note carefully what Chesky did not conclude. He did not conclude that marketing was worthless; Airbnb's absolute spend later grew with the business. He concluded that the measure had been lying about which marketing was working — and he re-founded the strategy on the asset rather than the dashboard.

Here, then, is the structural asymmetry the two cases expose, and it is the quiet centre of this essay. The question that decided both outcomes was not creative quality, channel mix or budget size. It was governance: who held the authority to overrule the dashboard. At Airbnb, a founder-operator who understood what the business owned looked at years of attribution reporting and called its bluff. At Nike, authority had been formally transferred to the data — "this is what you asked for" — and there was no one left positioned to say that the most important numbers in the building were the ones not on the screen. The same instrument error ran in both companies. The difference was whether anyone with strategic authority could see past it. That difference was worth, at minimum, the gap between a seven-per-cent relief rally and a $28 billion day.

The implication for leaders: put the measure back in its seat

If you lead a business, the question this essay leaves you with is not whether your marketing is measured. It should be: this publication will never argue otherwise.

The question is the direction of accountability: does the measure report to the strategy, or has the strategy started reporting to the measure? There is a fast diagnostic. List the metrics that were actually discussed in your last three marketing reviews. If every number on the list describes platform activity — reach, engagement, followers, ROAS — and none describes the asset — share of search, mental availability, the buying situations your brand credibly owns in memory, pricing power — then the dashboard is in the chair, and Beaverton has shown you the invoice's final form.[11]

If you lead a marketing function, the deeper reading is in your interest, uncomfortable as it is. Authority built on the dashboard is rented at the dashboard's terms, and Goodhart forecloses on it eventually, every measure gamed into a target eventually detaches from the business, and when the gap is repriced, it takes the function's credibility with it. The marketers who will hold strategic authority in ten years are not the ones with the fastest numbers. They are the ones who can stand in front of a board and account for the slow ones — who can say what the brand owns in the buyer's memory, prove it moved, and connect it to the P&L. That is a harder job than reporting reach. It is also the only version of the job that survives contact with an Elliott Hill.

And if you are weighing the two words this essay began with, weigh them like an investor, because that is what they are: an accounting distinction. Awareness is activity; salience is the asset. One is generated, measured and gone by Friday. The other compounds quietly in millions of memories until the day a buyer walks into a category and your brand either arrives, or does not — and every prior week's dashboard is instantly irrelevant to which.

Goodhart's warning was never that measurement fails. It was that measurement succeeds, at reshaping whatever you point it at, until the number is thriving and the thing it once described is gone. Nike pointed it at marketing and lost $28 billion in a day. Airbnb pointed past it and found half a billion dollars a year that was never doing what the screen claimed.

The measure always wins. That is precisely why it must never be allowed to run the game.

Drawing on Charles Goodhart's 1975 observation on measures under pressure, Byron Sharp and Jenni Romaniuk's account of mental and physical availability, Peter Field's IPA work on short-termism, Massimo Giunco's insider account of Nike's transformation, and company-reported figures from Nike and Airbnb cited below. The application of Goodhart's law to the governance of the marketing function — the measure-rich function acquiring strategic authority — is the author's.

Sources & References

1. Goodhart, C. A. E. (1975). "Problems of Monetary Management: The UK Experience," Papers in Monetary Economics, Reserve Bank of Australia — origin of the observation that any observed statistical regularity tends to collapse once pressure is placed upon it for control purposes. The popularised form, "when a measure becomes a target, it ceases to be a good measure," is Marilyn Strathern's (1997, "'Improving ratings': audit in the British University system," European Review). Accessed July 2026.
2. Dawes, J. (2021). Advertising Effectiveness and the 95-5 Rule, Ehrenberg-Bass Institute for the LinkedIn B2B Institute — approximately 5% of category buyers are in-market in a given quarter. https://business.linkedin.com/content/dam/me/business/en-us/marketing-solutions/resources/pdfs/advertising-effectiveness-and-the-95-5-rule.pdf. Accessed July 2026.
3. Sharp, B. (2010). How Brands Grow, Oxford University Press; Romaniuk, J. & Sharp, B. (2016). How Brands Grow: Part 2, Oxford University Press; Romaniuk, J. (2023). Better Brand Health, Oxford University Press — mental availability, physical availability, category entry points, and the primacy of light buyers in growth.
4. Field, P. (2019). The Crisis in Creative Effectiveness, Institute of Practitioners in Advertising — documenting the erosion of long-term advertising effectiveness as budgets and attention rotated toward short-term activation; building on Binet, L. & Field, P. (2013), The Long and the Short of It, IPA.
5. Wahba, P. (2024). "Former Nike CEO John Donahoe's downfall is a brutal lesson in corporate leadership," Fortune, 20 September 2024 — Donahoe's appointment and background, the wholesale retreat and its beneficiaries (Hoka, On, New Balance), the reliance on lifestyle retros and the stalled innovation pipeline, and the June 2024 guidance cut. https://fortune.com/2024/09/20/former-nike-ceo-john-donahoe-leadership-lessons-choosing-ceo/. Accessed July 2026.
6. Giunco, M. (2024). "Nike: An Epic Saga of Value Destruction," LinkedIn, 28 July 2024 — insider account of the January 2020 transformation: elimination of the category organisation, the DTC-led pivot, marketing made "data driven and digitally led," the loss of occasional buyers through the wholesale withdrawal, and the "poorly advised, data-driven guy" verdict. Giunco's figures: ~$25bn of market value lost on 28 June 2024, ~$70bn over the preceding nine months. Accessed July 2026.
7. Press reporting of Nike's 28 June 2024 results: the largest single-day market-value loss in the company's history, with figures reported between $24bn and $28.4bn depending on measurement basis (Fortune; Finimize; Mi3, "The real reason behind Nike's US$27.5 billion loss," September 2024). The stock's ~20% single-session fall followed cut full-year guidance. Accessed July 2026.
8. Nike, Inc. (2024). Announcement of 19 September 2024: Elliott Hill, 32-year company veteran and former President of Consumer and Marketplace, to succeed John Donahoe as CEO effective 14 October 2024; Nike shares rose approximately 7% on the announcement (reported in Fortune, above). Accessed July 2026.
9. Campaign (2021). "Airbnb slashes spend in permanent shift from performance marketing to brand," June 2021 — from Airbnb's first annual report: total brand and performance marketing spend cut 58%, from $1.14bn (2019) to $482m (2020); performance marketing cut by $541m against a $121m brand cut; traffic retained at ~95% of the prior year; marketing as a share of revenue from 23.7% to 14.2%; Chesky: "we can take marketing down to zero and still have 95% of the same traffic… we're not going to forget that lesson"; "our brand is inherently strong — it's a noun and a verb in pop culture." https://www.campaignlive.co.uk/article/airbnb-slashes-spend-permanent-shift-performance-marketing-brand/1708621. Accessed July 2026.
10. Marketing Week (2023). "Airbnb CEO: Our brand strength allows us to keep marketing spend as a percentage of revenue consistent," August 2023 — CFO Dave Stephenson crediting the shift from performance to brand with improving overall profitability; the full-funnel approach and international brand-led growth. https://www.marketingweek.com/airbnb-brand-strength-marketing-spend-consistent/. Accessed July 2026.
11. On salience-side instrumentation: Binet, L. (2020). "Share of Search as a Predictive Measure," EffWorks Global, IPA — share of organic category search as a proxy for mental availability, correlating with and leading market share by up to a year in the categories studied; and Romaniuk's mental-availability metrics (Mental Penetration, Network Size, Mental Market Share) in Better Brand Health (2023). Accessed July 2026.