The dominant structure in strategy advisory is leverage: senior partners sell and supervise, and the work is executed by a team several layers below them. There are good reasons for this. It scales, it develops people, and for a large class of problems, those requiring substantial analytical throughput, it produces better results than any individual could.
It is also poorly matched to a specific kind of engagement: the one where the value lies in judgment rather than analysis, and where the client's real question is what a person with three decades of relevant experience would actually do.
What leverage costs
In a leveraged model the senior judgment is applied at review points, a handful of hours across an engagement, exercised on material already shaped by others. The partner sees a filtered version of the situation. The filtering is competent, and it still removes exactly the ambiguous, atmospheric, half-substantiated signals that experienced judgment is best at reading.
This matters most in cross-border work, where a great deal of what determines the outcome is not in the data: how a counterparty's silence should be read, whether a relationship is actually forming, what a particular hesitation in a meeting signified. These are not analytical questions and they do not survive summarization.
A great deal of what determines a cross-border outcome is not in the data, and does not survive summarization.
The trade-off we accept
Principal-led engagement has an obvious constraint: capacity. We take on a small number of engagements because the model does not permit otherwise, and we decline work that genuinely requires analytical scale rather than judgment, for which a leveraged firm is the better answer, and we will say so.
What the structure makes possible is a different relationship. The person in the first meeting is the person in the difficult meeting eight months later. Recommendations are not abstracted from execution, because the same individual is present for both. And the advice can be direct in a way that a consensus-managed team product usually cannot be. That includes, when it is warranted, advising a client not to proceed.
