Sixteen billion dollars. That was McKinsey & Company's revenue in 2023 — a record for a firm that sells nothing you can touch, ship, or install. No software license. No manufactured good. No patent portfolio generating royalty streams while its owners sleep. Just advice, delivered by roughly 30,000 consultants dispatched to 4,400 active engagements across more than 65 countries, billed at rates that can exceed $1 million per week for a single team. The number is staggering not for its absolute size — plenty of companies are larger — but for what it represents: the monetization of institutional confidence at a scale no competitor has matched in the nearly one hundred years since an accounting professor from the Ozarks decided that running a business could be taught as rigorously as medicine or law.
And yet that same year, McKinsey agreed to a $650 million federal settlement — the largest ever imposed on a management consulting firm — for its role in helping Purdue Pharma "turbocharge" sales of OxyContin during a public health catastrophe that has killed more than half a million Americans. A former senior partner pleaded guilty to a felony count of obstruction of justice for deleting documents. The Department of Justice called it "the first time a management consulting firm has been criminally responsible for advice it has given resulting in the commission of a crime by a client." The firm that built its brand on dispassionate excellence had become, in the eyes of federal prosecutors, an accessory.
This is the central paradox of McKinsey — and, in some sense, of the entire management consulting industry it invented. The same machine that produces eighteen sitting Fortune 500 CEOs, that has advised governments from Washington to Riyadh to Singapore, that publishes research shaping how the world thinks about productivity, AI, and economic growth, is also a machine that — by structural design — lacks the accountability mechanisms of the professions it has spent a century emulating. Doctors carry malpractice insurance. Lawyers face disbarment. Auditors are subject to SEC oversight. McKinsey operates behind non-disclosure agreements so comprehensive that, as Walt Bogdanich and Michael Forsythe documented in
When McKinsey Comes to Town, "Americans and, increasingly, people the world over are largely unaware of the profound influence McKinsey exerts over their lives."
The question is not whether McKinsey is powerful. It is whether a century-old organizational architecture designed to concentrate intellectual talent, maximize partner economics, and maintain radical client confidentiality can survive in an era that demands transparency — and whether the consulting model itself, facing the most significant technological disruption in its history, will prove as durable as the firm's mythology suggests.
By the Numbers
The McKinsey Machine
$16BRevenue in 2023, a record
~45,000Total employees worldwide
30,000+Consultants across 65+ countries
4,400Active client engagements
18Current Fortune 500 CEOs who are alumni
$650MFederal settlement over Purdue Pharma work
~700Senior partners who elect the managing partner
1M+Annual applicants for employment
The Accountant Who Wanted to Be a Doctor
James Oscar McKinsey was born in 1889 in rural Missouri, the son of a farmer, the kind of Ozarks upbringing that tended to produce schoolteachers or preachers but rarely the founder of an industry. He studied at Missouri State Teachers College, then the University of Oklahoma, then the University of Chicago, accumulating degrees and professional certifications — CPA, lawyer, professor — with the relentless acquisitiveness of someone determined to escape his origins by mastering every credentialing system available. By 1926, when he founded his eponymous firm in Chicago, McKinsey was not a consultant. The word barely existed in its modern sense. He was an accounting professor who had arrived at a radical insight: the same systematic analysis that governed auditing could be applied to the entire operation of a business. Budgets, organizational structure, strategy — these were not matters of executive instinct but of rigorous, structured inquiry.
His "General Survey Outline" — a comprehensive diagnostic framework for evaluating a company's competitive position, financial structure, and operating efficiency — became the ur-text of management consulting. It was, in essence, the first consulting methodology, and it reflected McKinsey's deepest conviction: that management was not an art but a profession, and that professionals required tools, standards, and independence from the clients who employed them.
McKinsey died in 1937, at forty-eight, of pneumonia contracted while serving as chairman of Marshall Field's — a client engagement that had become an operating role, a violation of the very principle of advisory independence he had championed. The irony was not lost on the partners who survived him, and it became the founding myth's dark footnote: the consultant who went native and paid with his life. The firm nearly collapsed in the succession struggle that followed.
From accounting professor to consulting pioneer
1889James O. McKinsey born in Mexia, Missouri — son of a farmer.
1926Founds James O. McKinsey & Company in Chicago, offering "management engineering" services.
1933Hired by Marshall Field's department store; eventually becomes its chairman.
1937McKinsey dies of pneumonia at 48. The firm fractures — the Chicago office splits off, becoming A.T. Kearney.
1939Marvin Bower assumes control of the New York office, rebranding it McKinsey & Company.
The Architect of the Professional Mystique
If James McKinsey invented the consulting engagement, Marvin Bower invented the consulting firm. Born in Cincinnati in 1903, a graduate of Brown and Harvard Law School who then added a Harvard MBA, Bower brought to McKinsey something far more consequential than analytical rigor — he brought the institutional theology of a profession. He had studied law at a moment when the great American law firms were consolidating their identity as partnerships governed by ethical codes, peer accountability, and a shared understanding that the profession's reputation was more valuable than any individual fee. Bower looked at these institutions and saw a template for what consulting could become.
The changes he imposed over his nearly three decades leading the firm — from 1939, when he took control of the surviving New York office, through his formal retirement from the managing directorship in 1967 — were less operational than civilizational. He banned the word "company" from internal usage, insisting on "firm" because that's what lawyers called their organizations. Clients were not "customers." Consultants did not "sell" — they "served." Engagements were "studies," never "jobs." This was not mere euphemism. It was a deliberate act of professional construction, a systematic effort to elevate consulting from the status of hired-hand advisory work to something that resembled — in its self-conception, at least — the prestige of medicine or law.
If a company is going to go somewhere, to really lead its industry, its people have to operate in a professionally creative way. And they won't do that unless they have a strong set of beliefs to guide them.
Bower's operating principles became McKinsey's constitution. The one-firm partnership: no separate profit centers, no regional fiefdoms, all revenue pooled globally and distributed by committee. The up-or-out promotion system: borrowed directly from the Cravath law firm model, it guaranteed that every consultant was either advancing toward partnership or departing — usually within five to seven years — thereby creating the alumni network that would become McKinsey's most powerful business-development engine. The CEO-only client relationship: McKinsey would advise only the chief executive, never subordinates, ensuring both strategic altitude and maximum pricing power. The obligation to dissent: the formal expectation that consultants would tell clients what they needed to hear, not what they wanted to hear, a principle that functioned as both ethical lodestar and sales proposition.
Bower's masterstroke was understanding that a professional services firm's primary asset is not its people per se — people leave — but its brand, and that brand is inseparable from the behavioral standards that every individual upholds. As Duff McDonald observed in
The Firm, Bower "wasn't selling strategy. He was selling McKinsey." The distinction is everything. Strategy can be replicated. The aura of institutional authority cannot.
When Bower reached the mandatory retirement age of sixty — a policy he himself had instituted — he sold his shares back to the firm at book value, reportedly leaving millions on the table. He could have sold them at a premium to an outside buyer. He chose not to, because doing so would have converted a partnership into a commodity. The gesture became myth, repeated to every incoming class of associates as proof that The Firm's values were not aspirational but structural. Bower lived until 2003, dying at ninety-nine, and the last four decades of his life were spent as a kind of consulting pope emeritus, radiating moral authority from an office on the thirty-fourth floor of McKinsey's New York headquarters.
The Knowledge Machine
The intellectual architecture that transformed McKinsey from a well-regarded strategy shop into the definitive global consulting institution was laid down in the 1970s and 1980s, a period when the firm's leadership grasped a truth that would later become axiomatic in the knowledge economy: in a professional services firm, the product is the process of developing the product.
Fred Gluck, who would become managing director in 1988, began in the late 1970s to formalize what had been informal — the creation of practice areas organized around industry sectors and functional disciplines, with codified knowledge bases that captured methodologies, case studies, and analytical frameworks. His 1978 staff paper, "The Evolution of Strategic Management," was both an intellectual contribution and an internal manifesto: consulting needed to move beyond ad hoc problem-solving toward the systematic development and dissemination of proprietary knowledge.
The frameworks that emerged from this era became the lingua franca of corporate strategy. The 7-S Model — developed in collaboration with Tom Peters and Robert Waterman, who would go on to publish In Search of Excellence — offered a holistic framework for organizational effectiveness that transcended the cold financial analysis of BCG-style portfolio matrices. The business-system approach. The three-horizons growth framework. The organizational health index. Each was a tool designed to do two things simultaneously: help clients think about their problems, and give McKinsey consultants a shared analytical vocabulary that made the firm's output feel systematically rigorous rather than individually idiosyncratic.
By 1996, when Rajat Gupta was managing director, McKinsey had 3,800 consultants in 69 offices worldwide, and the firm's knowledge-management apparatus had become a case study in its own right — literally. Christopher Bartlett's Harvard Business School case, "McKinsey & Co.: Managing Knowledge and Learning," documented the interlocking systems of practice development bulletins, knowledge resource directories, and competence centers that enabled a consultant in São Paulo to access the accumulated expertise of colleagues in London or Tokyo. The challenge Gupta faced was recursive: how do you manage knowledge creation at a firm whose product is knowledge creation, without bureaucratizing the creative process?
The answer, characteristically, was structural ambiguity. McKinsey organized simultaneously along three dimensions — local office, industry practice, and functional competence — creating a matrix in which any individual consultant sat at the intersection of multiple knowledge networks. This triple-axis design was intentionally complex, generating what Bartlett called "creative friction" but also making it nearly impossible for a competitor to replicate. The knowledge system was not a database. It was a culture.
The CEO Factory
The phrase has become so associated with McKinsey that it now appears in the firm's own promotional materials, stripped of the mild irony with which outside observers initially deployed it. But the numbers are genuinely remarkable.
As of 2025, McKinsey claims at least eighteen current Fortune 500 CEOs as alumni — more than any other company on Earth. Sundar Pichai at Alphabet.
Jane Fraser at Citigroup. Tony Xu at DoorDash. James Taiclet at Lockheed Martin. Ryan McInerney at Visa. Expand to the Fortune Global 500 and the count reaches twenty-eight, including Novartis's Vas Narasimhan and Allianz's Oliver Bäte. The firm's internal data — unverifiable by outsiders, naturally — suggests that more than 500 alumni have held C-suite roles at current Global 500 companies since the founding, and that more than half of its alumni over age forty have reached the C-suite somewhere.
You learn to structure issues from the ground up, to define what success looks like, and to pull the right people together to get there. That way of thinking has never left me.
The mechanism is not mysterious, but it is harder to replicate than it appears. First: exposure. A twenty-six-year-old McKinsey associate, two years out of business school, routinely sits in rooms with Fortune 500 executives discussing billion-dollar decisions. The sheer velocity of context-switching — a pharmaceutical merger this quarter, a retail supply chain optimization the next, a government digitization initiative the quarter after — develops a form of pattern recognition that specialists in a single industry cannot match. Fraser called it "problem structuring." Others call it the ability to walk into any boardroom and not be lost.
Second: feedback. McKinsey's evaluation system is relentless. Reviews are frequent, detailed, and paired with coaching from senior partners. The up-or-out model guarantees that only those who can absorb and act on criticism survive. What emerges, in the alumni who rise to CEO elsewhere, is not just analytical skill but a particular form of executive resilience — the habituated comfort with being told, repeatedly, that your work is not good enough.
Third, and most importantly: the network itself. The 50,000-strong alumni community operates as the world's most valuable informal professional network, a self-reinforcing loop in which former McKinseyites hire current McKinseyites, who later become former McKinseyites, who then hire more McKinseyites. The alumni network is not a side effect of the business model. It is the business model's most durable competitive advantage.
The Economics of Selling Certainty
Understanding McKinsey requires understanding the economics of high-end professional services — a business model that is, in its essentials, stunningly simple and stunningly difficult to sustain.
McKinsey sells human hours. That's it. There is no recurring revenue from software licenses, no installed base generating maintenance fees, no platform effects that reduce marginal costs over time. Every dollar of revenue requires a consultant to show up, do work, and convince a client that the work was worth what they paid. The firm reportedly charges between $60,000 and $500,000 per consultant per month, with a standard engagement deploying a team of three to seven consultants over eight to sixteen weeks. A typical project might run $2 million to $5 million; the largest transformational engagements stretch into the tens of millions.
The gross margin structure of consulting is deceptively attractive. The primary cost is compensation, and because McKinsey's brand allows it to underpay relative to the counterfactual earnings its analysts and associates could command in private equity or technology — the currency is career optionality, not current salary — margins on individual engagements are substantial. The problem is that there is no operating leverage in the traditional sense. To double revenue, you roughly need to double headcount. McKinsey has tried to bend this curve through knowledge management, methodological standardization, and, more recently, AI-augmented analytics, but the fundamental constraint persists: the business scales linearly with human capital.
This creates the tension that has defined McKinsey's strategic choices for a century. Growth requires hiring more people. Hiring more people dilutes the brand unless quality controls are exceptionally rigorous. Rigorous quality controls limit hiring. The up-or-out system was Marvin Bower's elegant solution: perpetual selectivity without the appearance of cruelty, continuous attrition disguised as a meritocratic tournament, a machine that processes raw intellectual talent into either partners or alumni, both of which generate revenue.
McKinsey's headcount swelled from roughly 28,000 employees in 2018 to approximately 45,000 by 2024 — a 60% increase in six years. This was the pandemic-era boom, when digital transformation budgets seemed limitless and every Fortune 500 company needed a consulting team to help navigate supply chain disruption, remote work, and the sudden urgency of "doing something about AI." The hangover arrived in 2023 and 2024: an uncertain macroeconomic environment, longer sales cycles, clients shelving discretionary projects. McKinsey eliminated roughly 1,400 roles in 2023 — an unusual move for a firm that traditionally managed headcount through the genteel mechanism of "counseling to leave." In early 2024, approximately 3,000 consultants — around 10% of the consulting workforce — received a "concerns" performance rating, the precursor to being counseled out.
The growth-quality tradeoff is not new. A 1993 Fortune profile described rivals warning that McKinsey risked becoming "the IBM of consulting," a giant too rigid to evolve. One competitor compared it to "fallen dinosaurs" — Eastman Kodak, Sears, Pan Am. Thirty-two years later, McKinsey is larger and more profitable than ever. The predictions of decline have, so far, been consistently wrong. But the anonymous letter circulated within the firm in March 2024 — penned, purportedly, by disaffected former partners — echoed the same critique: "self-inflicted overcapacity," "short-sighted commercialism," "a decline of both the quality and quantity of partner engagement." The letter's authors, whoever they were, titled it "An obligation to dissent." Bower's own phrase, weaponized against his heirs.
The One-Firm Firm
McKinsey's governance structure is genuinely unusual and genuinely consequential. The firm is a private partnership — not a corporation, not publicly traded, not subject to SEC disclosure requirements. There are no outside shareholders. No quarterly earnings calls. No obligation to disclose revenue, margins, partner compensation, or client lists. The opacity is total by design.
The managing partner — the firm's highest executive role — is elected by the approximately 700 senior partners for a three-year term, renewable once. This is not a CEO appointed by a board of directors; it is a peer-elected leader who governs by consensus, moral authority, and the implicit threat that the partnership can replace you. The position has no formal executive power in the corporate sense. The managing partner cannot unilaterally hire or fire partners, set compensation, or redirect the firm's strategy. They lead through influence, persuasion, and the careful cultivation of coalitions — skills that are, not coincidentally, the same skills McKinsey teaches its clients.
Bob Sternfels, the current global managing partner, was reelected to a second three-year term in early 2024. A Stanford and Oxford graduate (Rhodes Scholar), he joined McKinsey as an associate thirty-two years ago and has spent his tenure navigating the opioid fallout, the post-pandemic headcount correction, and the firm's bet on AI-augmented consulting. "We invest over a billion" in proprietary intellectual property each year, Sternfels told HBR in a January 2026 interview — a figure that, like all McKinsey numbers, is unaudited and unverifiable.
The one-firm structure — all revenue pooled globally, all partners compensated from a single pot — is both McKinsey's greatest cultural asset and its most significant structural constraint. The asset: it eliminates the regional profit-center competition that fragments competitors like Deloitte or Accenture, ensuring that a partner in Mumbai has every incentive to staff the best team for a client in Munich. The constraint: it means every partner has a de facto veto on institutional direction, creating the consensus-driven decision-making that alumni describe as either "profoundly democratic" or "glacially slow," depending on whether they benefited from or suffered under it.
McKinsey is a very kind place. McKinsey is a very cruel place.
The cruelty is structural. Up-or-out means that most people who enter McKinsey will leave it — not because they failed in any absolute sense, but because the pyramid narrows. Of every hundred associates hired, perhaps ten will make partner. Of those ten, perhaps two or three will make senior partner. The rest depart, bearing the McKinsey brand on their résumé like a war medal, and enter the alumni network where they become future clients, future board members, and future sources of revenue. The kindness, such as it is, lies in the effort McKinsey invests in making the departure feel like graduation rather than expulsion. "Counseled to leave" is the euphemism — a phrase that captures, in three words, the entire emotional architecture of the system.
The Shadow Civil Service
McKinsey's influence extends far beyond the corporate boardroom, and it is in the public sector that the firm's opacity becomes most controversial. The firm has advised dozens of national governments, central banks, and multilateral institutions — from the U.S. Department of Defense to Saudi Arabia's sovereign wealth fund to the U.K.'s National Health Service. The work is often invisible to the public. Non-disclosure agreements shield the scope, the recommendations, and the outcomes. The firm operates, in effect, as a shadow civil service — staffing the analytical functions that governments either cannot or will not develop internally, at fees that dwarf what any civil servant earns.
The conflict-of-interest problem is structural and unresolvable within the current business model. McKinsey simultaneously advises pharmaceutical companies and the government agencies that regulate them. It consults for oil companies and the climate policy teams tasked with reducing emissions. It works for competing firms within the same industry, relying on internal information barriers — so-called "ethical walls" — whose effectiveness is, by definition, unverifiable from the outside. As Bogdanich and Forsythe documented, the firm's internal records show it "has advised virtually every major pharmaceutical company — and their government regulators."
The problem is not that McKinsey employees are venal. Most are not. The problem is that the business model creates incentives that are structurally misaligned with the public interest, and the firm's radical opacity ensures that misalignment is invisible until something goes catastrophically wrong.
The Opioid Stain
Something went catastrophically wrong.
The details are by now extensively documented — in litigation records, in the Opioid Industry Documents Archive hosted by UCSF and Johns Hopkins (which released more than 114,000 McKinsey-related documents in 2022), in federal court filings, and in the firm's own admissions. Between 2004 and 2019, McKinsey advised Purdue Pharma, Endo Pharmaceuticals, Johnson & Johnson, and Mallinckrodt on strategies to increase opioid sales. The consulting was granular: how to target high-prescribing doctors, how to counter the efforts of DEA investigators, how to structure sales incentives to maximize prescriptions. One McKinsey proposal — outlined in a planning document but apparently never implemented — involved making payments to insurance companies of up to $14,810 whenever a patient became addicted or overdosed in an "event" linked to Purdue's opioids. The euphemism "event" is doing considerable work in that sentence.
In July 2018, as litigation against Purdue intensified, senior partner Martin Elling wrote in an internal email: "It probably makes sense to have a quick conversation with the risk committee to see if we should be doing anything other tha[n] eliminating all our documents and emails. As things get tougher here someone might turn to us." Elling subsequently deleted Purdue-related materials from his McKinsey laptop. In December 2024, he agreed to plead guilty to a felony count of obstruction of justice.
The total financial cost to McKinsey: roughly $1.55 billion — $650 million in the federal settlement, plus nearly $900 million in prior settlements with state and local governments. The reputational cost is harder to quantify but arguably larger. The federal deferred prosecution agreement requires McKinsey to forgo all work related to the marketing, sale, or distribution of controlled substances; to maintain an enhanced compliance program subject to DOJ and HHS oversight; and to cooperate with ongoing investigations. The firm stated: "We should have appreciated the harm opioids were causing in our society and we should not have undertaken sales and marketing work for Purdue Pharma."
The opioid case is not an aberration. It is the extreme case of a structural vulnerability that exists in every engagement McKinsey undertakes: the firm optimizes for the client's stated objective — in this case, increasing OxyContin prescriptions — without sufficient institutional mechanisms for asking whether the objective itself is harmful. The obligation to dissent, Bower's most celebrated principle, failed precisely when it was most needed. Not because individuals lacked the courage to speak up, but because the partnership's economic incentives — fees from Purdue were substantial — created a gravitational pull that overwhelmed the cultural norms designed to resist it.
This is the pattern Bogdanich and Forsythe identify across decades: tobacco companies in the 1950s, insurance cost-cutting schemes that denied accident victims fair settlements, consulting for repressive governments. The product McKinsey sells is analytical intelligence. But intelligence is morally neutral. It can be used to reduce childhood malaria deaths or to turbocharge opioid sales. The firm's architecture provides insufficient mechanisms for distinguishing between the two.
The Culture of Bright People Telling You Things
The 1993 Fortune profile remains the most psychologically acute portrait of McKinsey's internal culture ever published, and its observations have aged remarkably well. John Huey described a firm populated by "high-performing achievers with egos large enough to block the sun" who were then "forced to bow to the collective." The hierarchy was intellectual, not bureaucratic. Seniority mattered less than the quality of your argument. A second-year associate could challenge a senior partner's analysis — was expected to challenge it, in fact — as long as the challenge was analytically grounded.
This created an organizational personality that was simultaneously impressive and exhausting. The consultants were, by nearly every measurable standard, extraordinarily bright: Baker Scholars from Harvard, Rhodes Scholars, nuclear physicists, Ph.D.s in the hard sciences, graduates of the world's most selective universities. More than a million people apply annually; the acceptance rate makes Stanford look permissive. The sorting function is so powerful that simply having "McKinsey" on your résumé functions as a credential — a signal that you survived a selection process designed to identify not just intelligence but a particular kind of high-output, analytically rigorous, interpersonally polished intelligence.
But the culture's strengths are inseparable from its pathologies. The elevation of analytical rigor above all other virtues can produce a dangerous blindness to moral complexity. If a problem can be structured, analyzed, and optimized, McKinsey will structure, analyze, and optimize it — regardless of whether optimization is the appropriate response. Not every problem should be decomposed into a two-by-two matrix. Not every inefficiency should be eliminated. Not every margin should be maximized. The firm's instinct — relentlessly, culturally, almost genetically — is to optimize. The opioid case revealed what happens when that instinct is deployed without adequate moral guardrails.
The internal language reinforces the pattern. "Rightsizing" for mass layoffs. "Optimizing" for cost-cutting that eliminates jobs and degrades safety. "
Change champions" for the people tasked with implementing painful restructurings. The euphemisms are not innocent. They are load-bearing structures in the firm's epistemic architecture, allowing consultants to discuss human consequences in the abstract language of business process improvement. As one former consultant observed, quoted in Bogdanich and Forsythe's investigation: "The very word
commercial, when spoken about anyone at McKinsey, is akin to profanity." The firm's culture insisted that its work was noble — professional, analytical, in service of something larger — even as the commercial reality was that McKinsey's clients paid extraordinary fees for advice that often boiled down to: cut costs, increase prices, and lay people off.
The AI Reckoning
The question that now confronts McKinsey is whether artificial intelligence will do to consulting what consulting did to middle management: eliminate it.
The traditional consulting engagement follows a pattern that is, in its essentials, a knowledge-processing workflow. Consultants gather data — through interviews, document review, financial analysis, and benchmarking. They analyze the data using frameworks, both proprietary and generic. They synthesize the analysis into recommendations, presented in PowerPoint decks of architectural precision. They may stay to help implement. Then they leave.
Every step of this workflow is vulnerable to AI automation. Large language models can process documents, generate analyses, and produce structured recommendations at a speed and cost that no team of human consultants can match. The benchmarking that once required a McKinsey team to call thirty companies and aggregate responses can now be done in minutes using AI agents trained on public data. The analytical frameworks that were once proprietary intellectual property are now freely available — many of them described in McKinsey's own published research.
Sternfels has acknowledged this directly. In a January 2026 HBR interview, he described McKinsey as already viewing its "first AI agents as very much part of its workforce" and rapidly expanding that capability. The firm claims to invest "over a billion" annually in proprietary IP, including AI tools. The strategic bet is that AI will augment rather than replace the consultant — handling the data-gathering and initial analysis while freeing human consultants to focus on the higher-order work that AI cannot do: building relationships with executives, navigating organizational politics, exercising judgment in ambiguous situations, and — perhaps most importantly — providing the institutional credibility that makes a CEO feel comfortable spending $10 million on advice.
This is a plausible thesis. But it contains a dangerous assumption: that clients will continue to pay premium fees for the credibility layer once they realize the analytical layer can be purchased for a fraction of the cost. The moat around McKinsey's brand is deep, but it was built in an era when information asymmetry was the consultant's primary source of value. In an era of AI-enabled information parity, the brand must justify itself on different grounds — grounds that are, by nature, harder to defend.
McKinsey is also rethinking its talent model. Sternfels has suggested that the firm is expanding its recruiting aperture beyond the traditional economist-engineer-MBA pipeline to include humanities graduates, people with "out-of-the-box thinking" that AI cannot replicate. If this shift is real and durable, it would represent a genuine departure from the analytical-rigor-above-all-else culture that Bower built. Whether the partnership will tolerate that departure — whether a firm that has spent a century selecting for a particular cognitive profile can successfully select for a different one — is an open question.
Bower's Ghost
Every institution eventually confronts the gap between its founding mythology and its operational reality. For McKinsey, that gap has been widening for decades, and the opioid case blew it open. But the mythology persists — not because people at McKinsey are naive, but because the mythology is load-bearing. Remove it, and the economic machine stops working.
Consider the alumni network. Its value depends on former consultants retaining an emotional attachment to the firm — believing, even after they leave, that McKinsey is fundamentally different from other employers, that the experience was formative in ways that transcend a résumé line item. This emotional attachment generates referral revenue, board placements, and the soft power that allows McKinsey to command premium pricing. The mythology — Bower's values, the obligation to dissent, the one-firm ethos — is the substrate on which the emotional attachment grows.
The anonymous letter of March 2024 threatened this substrate. Its authors — whether genuine alumni or provocateurs — understood that the most powerful critique of McKinsey is not that it is greedy or unethical, but that it has become ordinary: a large, commercially driven professional services firm chasing growth at the expense of the cultural distinctiveness that justified its premium positioning. "Let us be distinctive again," the letter pleaded. The word "again" carries an enormous weight. It implies that distinction has already been lost.
Sternfels's response has been to tighten governance, introduce "industry-leading client service policies," enhance compliance investment, and revise the code of conduct. These are the standard institutional responses to scandal — necessary, probably insufficient, and fundamentally oriented toward preventing the last crisis rather than the next one. The deeper challenge is architectural. McKinsey's structure — the partnership model, the up-or-out pyramid, the revenue-pooling system, the radical opacity — was designed for a world in which consulting firms were small, elite, and operated at the highest levels of corporate decision-making. McKinsey now has 45,000 employees and $16 billion in revenue. It has become, despite its own mythology, a large corporation — with a large corporation's bureaucratic tendencies, political dynamics, and accountability gaps. The question is whether the partnership structure can govern an organization of this scale, or whether the scale has already outgrown the structure.
The centennial arrives in 2026. McKinsey's leaders will celebrate a century of influence, innovation, and institutional durability that has no parallel in professional services. They will not celebrate the opioid settlement, or the South Africa bribery scandal, or the Saudi government work that drew condemnation after the murder of Jamal Khashoggi, or the decades of work for tobacco companies that began in the 1950s. They will invoke Bower's name and Bower's principles and Bower's decision to sell his shares at book value. The ghost will be present, as always — radiating moral authority, insisting on standards, reminding everyone that The Firm is not a business but a profession.
Whether anyone still believes that, in a year when the DOJ settlement requires federal oversight of the firm's compliance program, may be the defining question of McKinsey's second century.
In a conference room somewhere — São Paulo or Singapore or San Francisco, it doesn't matter, they all look the same — a team of McKinsey consultants is preparing a deck. The slide is titled "Strategic Transformation Roadmap." The client is a Fortune 200 company facing disruption from AI. The team includes a former Rhodes Scholar, a Stanford MBA, and an AI agent that completed the data analysis in twenty minutes. The partner leading the engagement billed at $12,000 a day. The AI agent cost nothing. The question of which one the client is actually paying for — the human judgment or the institutional brand — hangs in the recycled air like an obligation no one wants to dissent from.
