The Track Record Was Real

In 1946, ten former officers from the Army Air Forces Office of Statistical Control walked into the Ford Motor Company and offered to fix it. Henry Ford II hired all of them. Ford employees called them the Whiz Kids. The company had been run on instinct. The Whiz Kids brought data.¹

Robert Strange McNamara was the most disciplined of the ten. He built statistical systems for tracking costs, forecasting demand, and measuring output. Ford's market share climbed. Waste came under control. On November 9, 1960, the board named him president of the Ford Motor Company. He was forty-four years old, the first person outside the Ford family to hold the position.²

Five weeks later, President-elect Kennedy offered him the Department of Defense.

McNamara brought the Ford playbook to the Pentagon. Systems analysis. Planning-Programming-Budgeting. Quantitative management applied to the largest bureaucracy on earth. Then he applied it to a war.³

Body counts. Kill ratios. Sortie rates. Hamlet Evaluation System scores. Eighteen indicators across twelve thousand hamlets, computed with Bayes' Theorem. Each variable could be measured. Each variable went up.⁴

The variables that determined the outcome could not be measured by the Ford playbook. Population loyalty. Political legitimacy. Local knowledge of terrain and supply networks. The measurable variables were proxies. The proxies went up. The thing they were supposed to measure went down.⁵

Only two percent of American generals surveyed after the war considered the body count a valid way to measure progress. Sixty-one percent said the numbers were routinely inflated.⁶

McNamara knew. By mid-1965, within months of advocating escalation, he had privately concluded the war was militarily unwinnable. In April 1966, he told his aide John McNaughton: "I want to give the order to get our troops out of there so bad that I can hardly stand it." He did not give the order. In November 1967, he sent President Johnson a memorandum recommending withdrawal. Johnson rejected it. Twenty-eight days later, McNamara's resignation was announced.⁷

The credential holder was inside his own belief, knew the belief was failing, and could not break it. The cost of breaking it, admitting the war was unwinnable while serving as Secretary of Defense, while half a million American troops were deployed, was higher than the cost of holding it.

He did not say so publicly until 1995. "We were wrong, terribly wrong." Thirty years of silence.⁸

58,220 American dead.⁹

The premise

The board reviews the candidate. Built a division. Turned around a company. Delivered record revenue. The references confirm it. The track record is documented, verified, and real.

The hiring decision follows. The premise, that what this person did there they will do here, is never examined. The credential is so strong it becomes the examination.

The outcome was always two things multiplied together. Talent and conditions. The resume shows only the product. It never shows the multiplication. When you hire the person, you are betting the talent was the entire product.

It never is.

Revenue at a company with a monopoly is talent multiplied by the monopoly. The resume says the revenue number. The monopoly stays behind. Cost reduction at a company in crisis is talent multiplied by the crisis. The resume says the turnaround, but the crisis that made it possible does not follow. The credential carries the outcome across the boundary. The conditions do not cross.

Spencer Stuart studied 855 S&P 500 CEOs appointed over twenty years. Executives with prior CEO experience consistently underperformed first-timers on market-adjusted shareholder returns.¹⁰

Seventy percent performed better in their first CEO job than their second. The share of CEOs hired for prior experience quadrupled since 1997. The performance went the other direction.

The researchers had a phrase for it. "Experienced executives rely too much on old playbooks."

The wrong industry

Sculley, Apple, 1983

John Sculley was president of Pepsi-Cola at thirty-seven, the youngest in the company's history. He had taken the international food division from $156.5 million in losses to $40 million in profit within three years. He launched the Pepsi Challenge in 1975. His credential was consumer marketing, brand management, and retail distribution.¹¹

Steve Jobs recruited him with a question now repeated in every business school case study. "Do you want to sell sugar water for the rest of your life, or do you want to come with me and change the world?" Sculley accepted.¹²

Within two years, Sculley ousted Jobs. The board backed him unanimously. Then he applied the Pepsi playbook. Line extensions. Brand proliferation. Retail channel expansion across every possible market segment. Forty-four Performa models. Eleven Quadra models. Three Centris models that shipped and were discontinued in the same year. Nine LC models. Multiple machines with nearly identical hardware sold at different price points to different retailers.¹³

Revenue grew from roughly $800 million to $8 billion. Tenfold. The credential appeared to be working.

The Newton shipped in 1993 with handwriting recognition that barely worked. The Simpsons mocked it. Doonesbury mocked it. Sculley was forced out the same year. His successors accelerated the decline. By the time Steve Jobs returned in 1997, Apple was ninety days from insolvency. Jobs drew a two-by-two grid on a whiteboard, consumer and professional across the top, desktop and portable down the side, and cut every product line that did not fit in one of the four boxes.¹⁴

The Pepsi credential had built a consumer packaged goods company. Apple needed someone to build computers people loved. Revenue and product identity had moved in opposite directions for a decade. The credential could not see the difference because the credential measured the wrong variable.

The same industry, opposite conditions

Mayer, Yahoo, 2012

Marissa Mayer was Google's twentieth employee. She joined in 1999 and rose to Vice President of Search Products and User Experience. She sat on the operating committee. She was part of the three-person team that built AdWords, the advertising platform that became Google's revenue engine. She created the Associate Product Manager program, which produced five hundred alumni, many of them now leading companies across Silicon Valley.¹⁵

Google's conditions during Mayer's tenure: a search monopoly generating billions in ad revenue, an engineering-first culture, a product people already loved. Mayer built features on top of a platform that was already winning. The platform did most of the work. The credential recorded the output.

Yahoo's conditions: declining display advertising revenue, no mobile strategy, four CEOs in five years, and an identity crisis the company had been unable to resolve since 2001. Media company or tech company. Nobody knew.¹⁶

Mayer acquired Tumblr for $1.1 billion in May 2013. Yahoo wrote down $712 million of it within three years. In 2019, Verizon sold Tumblr to the parent company of WordPress for less than $3 million.¹⁷

She made approximately fifty acquisitions totaling $2.8 billion. Most were acqui-hires. The products were shut down. Yahoo's revenue was $4.98 billion the year she arrived. It was $5.16 billion when she left. Flat across five years. The stock rose fifty percent, but analysts attributed the rise almost entirely to Yahoo's stakes in Alibaba and Yahoo Japan, not to the core business.¹⁸

Yahoo had been worth $125 billion in January 2000. In June 2017, Verizon closed the acquisition of Yahoo's core internet business for $4.48 billion. A ninety-six percent decline from peak.¹⁹

The credential was from the same industry. The conditions were opposite. Google needed optimization inside a monopoly. Yahoo needed reinvention inside a decline. The Google resume measured one. The job required the other.

The same building

Chapek, Disney, 2020

Bob Chapek joined Disney in 1993 as a marketing director in home entertainment. Over twenty-seven years, he rose through distribution, consumer products, and parks. In February 2015, he became Chairman of Walt Disney Parks and Resorts. By fiscal 2019, his last full year running parks, the division generated $26.2 billion in revenue and $6.8 billion in operating income. The most profitable division in the company. The credential was operational: throughput, pricing optimization, capacity management.²⁰

Bob Iger handpicked him as CEO. Chapek was named to the position on February 25, 2020.

The CEO role needed something else. Creative talent relationships, content vision, public navigation. Chapek raised the Disney+ ad-free price from $7.99 to $10.99, a thirty-eight percent increase, calling the original launch price "absurd." The streaming division lost $4 billion in fiscal 2022.

When Scarlett Johansson sued over the simultaneous streaming release of Black Widow, Disney's public response called the lawsuit "especially sad and distressing" and disclosed her $20 million salary. When the Florida legislature passed the Parental Rights in Education Act, Chapek initially refused to take a public position, then reversed days later, then provoked a retaliatory dissolution of Disney's self-governing district that had existed since 1967.²¹

Two years and nine months. Iger came back. The company reversed the credential hire on November 20, 2022. Iger's first act was to fire Chapek's distribution chief and dismantle the centralized content structure Chapek had built.²²

Twenty-seven years inside Disney. The conditions on the parks floor and the conditions in the CEO suite were different enough that twenty-seven years of proximity did not close the gap. The resume was internal. The conditions were still different.

The record confirmed it

Tavares, Stellantis, 2021

Carlos Tavares saved PSA Group from bankruptcy. The French automaker had lost approximately five billion euros in 2012. Factory productivity was at sixty-five percent. In February 2014, Dongfeng Motor and the French government each invested 800 million euros to keep the company alive. Tavares launched a plan called "Back in the Race." He cut costs, streamlined products, raised margins. Factory productivity reached ninety percent by 2016. The operating margin hit five percent in 2015, more than double the target. By 2018, PSA posted record revenue. The credential was earned in a crisis.²³

Then Tavares engineered the merger of PSA and Fiat Chrysler, creating Stellantis in January 2021. He applied the same playbook. Aggressive cost reduction. Premium pricing. Margin expansion.

2023. Stellantis reported net profit of 18.6 billion euros. Record. The credential looked validated.²⁴

Then U.S. market share collapsed. Dealer inventories ballooned to over 150 days' supply. The Dodge Hornet sat on lots for 323 days. The Jeep Renegade, 332. Product quality declined. The Charger Daytona shipped with software problems. Consumer Reports called the Jeep "unfinished." Stellantis broke contract commitments to the UAW. The union filed grievances. Stellantis sued the union.²⁵

Net profit fell seventy percent in one year. Free cash flow went negative. The stock lost more than forty percent of its value. The board accepted Tavares's resignation on December 1, 2024.²⁶

The 18.6 billion euros is the data point that should have raised the alarm. The credential confirming itself. The metric the credential knew how to measure, profit, rising while the variables that would determine the outcome, market share, inventory, quality, labor relations, deteriorated underneath. Same structure as the body counts. The number going up. The war going down.

The playbook that saves a dying company is the opposite of the playbook that grows a healthy one. Saving requires cutting. Growing requires investing. The credential measured the outcome without decomposing the conditions.

What the system selects

Five cases. McNamara, 1961. Sculley, 1983. Mayer, 2012. Chapek, 2020. Tavares, 2021. Sixty years. Each credential was real. Each track record was documented. Each time, the talent was present. The conditions changed. The outcome changed with them.

Boris Groysberg, a Harvard Business School professor, studied more than a thousand star equity analysts at Wall Street investment banks. Stars who switched firms suffered an immediate and lasting performance decline that persisted for at least five years. Their earlier excellence, Groysberg found, "appeared to have depended heavily on their former firms' general and proprietary resources, organizational cultures, networks, and colleagues." Stars who moved with their teams showed no decline. Stars who moved alone became, in Groysberg's word, meteors.²⁷

The hiring system selects for the credential. The board reads the resume. The search committee reviews the track record. The executive recruiter matches achievements to job descriptions. Each step evaluates what the person has done. No step evaluates whether the conditions that enabled what they did are present in the new environment. The system selects for legibility and ignores the variable that determines the outcome.

Three payoffs keep the system in place.

The first is social safety. Nobody gets fired for hiring the most impressive resume. The credential functions as a consensus.

The second is cognitive ease. The resume saves the board the work of analyzing conditions. The credential substitutes for the analysis.

The third is delegated responsibility. If the hire fails, the board hired the best person available. The failure belongs to the executive. The selection process is clean because the credential was real.²⁸

The credential holder is also inside the belief. McNamara applied Ford metrics to Vietnam because Ford metrics were who he was. Tavares applied cost-cutting to Stellantis because cost-cutting saved PSA and saving PSA defined him. Sculley built a consumer packaged goods company inside Apple because building consumer packaged goods companies was the only thing the credential had ever measured. The playbook becomes the person who carries it. The self-belief the credential creates prevents the adjustment.

The premise I held

We trusted the resume instead of challenging the premise.

There was a partnership. The track record was real. We stopped doing the one thing we tell every client to do: check whether the conditions that produced the past result exist in the present one. The system kept running. It was not profitable. The credential was so strong it suppressed the question that would have caught the mismatch. The most expensive lesson either of us ever paid for. It is why MMK Retail exists.

The counter-case

Mulally, Ford, 2006

Alan Mulally spent thirty-seven years at Boeing. He ran the 777 program, the first commercial airliner designed entirely with digital software. He became President and CEO of Boeing Commercial Airplanes. Then, in September 2006, he was hired as CEO of Ford Motor Company. An aerospace executive running an automaker. By every rule of this piece, a credential failure waiting to happen.²⁹

Ford had lost $12.7 billion in 2006, the worst year in the company's 103-year history. The stock was at $8.39. The debt was rated junk. U.S. market share had fallen from twenty-five percent in the late 1990s to 17.5 percent.

Mulally's first structural decision was diagnostic, not operational. He created weekly Business Plan Review meetings. Every Thursday. Same time, same room. Color-coded status reports. Green, yellow, red. For weeks, every report came back green. In a company losing $12.7 billion a year.³⁰

Mulally kept asking. We are losing billions. Is there nothing that is not going well?

Mark Fields, the president of Ford Americas, was the first to break. The Ford Edge had a problem in production. Fields turned his slide red. The room went silent. Everyone waited for the response that had always come in Ford's culture: punishment.

Mulally clapped. "Great visibility."

The next week, the charts were covered in red and yellow. The culture of hiding problems broke because one person was rewarded for telling the truth about the conditions.³¹

Before the financial crisis arrived, Mulally had already mortgaged virtually every asset Ford owned, including the Blue Oval logo, the Mustang trademark, the F-150 trademark, the company's headquarters, its factories, and Ford Motor Credit, to secure a $23.5 billion credit line. His stated rationale: "A cushion to protect for a recession or other unexpected event." The economy was still healthy when he made that decision. He was diagnosing conditions that had not yet arrived.³²

Ford was the only Big Three automaker that did not go through bankruptcy in 2008-2009. The stock fell to $1.01 in October 2008. By January 2011, it was $18.79. Ford posted a $2.7 billion profit in 2009, a $6.6 billion profit in 2010. The credit rating returned to investment grade. The company logged nineteen consecutive profitable quarters before Mulally retired in July 2014.³³

Mulally diagnosed the conditions before importing a playbook. He spent months learning what Ford's problems were, then built a system for Ford, on Ford's terms.

The difference between Mulally and every other case in this piece: each of them imported their playbook. Mulally checked the conditions first.

The credential gets you in the room. The premise check determines what happens next.

The desk

The body count reports arrived on the desk at the Pentagon where production numbers from Ford used to sit. Same columns. Same precision. Different stakes.

The metrics went up every quarter. The war went down every quarter. The credential holder knew. By April 1966, he knew. The cost of saying so was too high. The numbers that saved Detroit could not save Saigon. The conditions were different. The playbook was the same.

The conditions were never portable. The credential carries the outcome across the boundary. The conditions stay behind.

Someone in every room holds the signal that the premise is wrong. Whether they surface it depends on the cost of challenging the credential.

The track record was real. The premise was unchecked.