The Indispensable Problem: Why the Paid Advisor Has Never Once Been Structured to Give You Good News
In the summer of 2001, Enron Corporation was engaged in the advanced stages of financial collapse. Its accounting was fraudulent, its business model was incoherent, and its leadership was either complicit or incompetent. It also had, at that moment, approximately two hundred McKinsey consultants on retainer.
This fact is cited occasionally as an embarrassment. It deserves citation as a data point in a much longer argument.
The management consulting industry — McKinsey, Bain, Boston Consulting Group, and their hundreds of competitors — is frequently analyzed as a modern phenomenon, a product of postwar corporate complexity and the MBA credential economy. This analysis is superficially accurate and fundamentally misleading. The paid strategic advisor is among the oldest professional categories in recorded history, and the structural flaw at the center of the modern consulting model is the same structural flaw that has existed since the first haruspex charged a Roman general for reading entrails before a campaign.
What the Romans Understood About Ambiguity
The haruspex was a specialist in divination — the reading of animal entrails, the interpretation of lightning, the decoding of unusual natural events as divine messages bearing on military and political decisions. Roman commanders and magistrates consulted them routinely before major undertakings, and the profession was sufficiently lucrative and prestigious that it maintained formal institutional structures for centuries.
The haruspex faced a professional problem that will be immediately recognizable to anyone who has sat across a conference table from a McKinsey partner: a clear, definitive answer was commercially dangerous. If the omens were unambiguously favorable and the campaign failed, the advisor bore responsibility. If the omens were unambiguously unfavorable and the commander proceeded anyway — as commanders frequently did — the relationship ended. The professionally optimal answer was one that acknowledged complexity, identified multiple contingencies, and positioned the advisor as essential to ongoing interpretation as events developed.
Ambiguity, in other words, was not a failure of the service. It was the product.
Medieval court astrologers operated under identical incentive structures. The most successful practitioners were not those who made accurate predictions — accuracy, over time, is statistically impossible for anyone genuinely claiming to read the future in celestial movements. The most successful were those who framed their counsel in terms that could be retrospectively validated regardless of outcome, while maintaining a posture of irreplaceable expertise that made dismissal feel riskier than retention. Kings who consulted astrologers were not purchasing forecasts. They were purchasing the psychological comfort of having consulted an expert, combined with a socially credible explanation for whatever decision they had already made.
The McKinsey Structural Problem
Management consulting's revenue model has been described in various ways by its practitioners and critics, but its essential architecture is not complicated: firms are paid by the day or by the project, and their long-term commercial health depends on repeat engagements and expanded scope. A client whose problem is cleanly resolved has no further need of the consultant. A client whose problem is partially addressed, rendered more complex, or replaced by a new set of problems requiring fresh analysis is a client with an ongoing relationship.
This is not an accusation of deliberate fraud. The incentive structure does not require bad faith to produce predictable outcomes. It simply requires that human beings — consultants included — respond to incentives, which five thousand years of evidence confirms they reliably do.
The consequences are visible in the historical record of major consulting engagements. Studies of large-scale consulting projects consistently find implementation rates for recommended changes that hover well below fifty percent. The consulting industry's own literature attributes this to client resistance, change management failures, and organizational complexity. An alternative explanation, less flattering to the profession, is that recommendations are sometimes designed to be sophisticated enough to justify the engagement without being specific enough to be implemented — and therefore evaluated.
The Advisor's Oldest Defense: Complexity
In fifteenth-century Florence, the Medici employed a network of commercial agents — factors — whose compensation depended on commissions from transactions they facilitated. The structural incentive of the factor was to generate transactions, not necessarily to generate good transactions for the principal. Florentine merchant manuals of the period contain extensive guidance on how to manage factors, how to audit their recommendations, and how to prevent the advisor relationship from drifting into something that served the advisor's interests at the principal's expense.
This is, in modern terminology, a principal-agent problem. It has been a known problem since the first moment one human being paid another to act on their behalf. The remarkable thing is not that it exists but that every generation appears to rediscover it as though it were new.
The consulting industry's primary defense against the principal-agent critique is complexity. Problems are presented as too intricate, too multidimensional, and too context-dependent for simple resolution — which is often true, but conveniently also makes the advisor's contribution impossible to evaluate against any objective standard. When a McKinsey engagement produces a four-hundred-page report recommending a strategic transformation that takes three years to implement, the causal chain between advice and outcome is sufficiently attenuated that attribution becomes a matter of negotiation rather than measurement.
The Roman haruspex, reading a sheep's liver before the Battle of Cannae, operated in a nearly identical epistemic environment.
The Whistleblower Problem
One consistent feature of the historical record is that advisors who give genuinely useful counsel — counsel that is specific, actionable, and honest about the client's actual situation — tend to have shorter engagements than those who manage relationships skillfully. The court physician who told Henry VIII that his dietary habits were dangerous was not retained for his candor. The advisor who tells a corporate client that their strategic problem is their current leadership is rarely invited back.
This is not a pathology of consulting specifically. It is a feature of the advisor relationship generally. The client controls the renewal decision, and clients, being human, are drawn toward advisors who confirm their existing instincts, provide sophisticated justifications for decisions already made, and identify the source of organizational dysfunction in departments or personnel that the client already dislikes.
The resulting dynamic is one in which the most commercially successful advisors are frequently those most skilled at reflecting the client's existing beliefs back to them in the language of rigorous analysis. This is not a new observation. Plutarch wrote about it. Machiavelli wrote about it. The Harvard Business Review writes about it approximately twice per decade, apparently without noticing the recursion.
What Five Thousand Years Suggests
The paid advisor is not going away. Organizations genuinely require outside perspective, specialized expertise, and analytical capacity that cannot be maintained internally. The advice industry fills a real need.
But the historical record offers no example of an advice industry that resolved this structural tension through professional norms, credentialing systems, or ethical codes. Every attempt to align the advisor's incentives with the client's actual interests has been defeated by the same force: the client controls the payment, and the payment depends on continuation.
The Roman Senate knew this. The Florentine merchant houses knew this. The corporate boards of the twenty-first century know this, and continue to write the checks anyway.
Five thousand years of data. Draw your own conclusions.