Consulting is not for the broken: why great companies don't wait for a crisis

Consulting is usually sold as a fix, and a fix implies a fault. Once that framing is in place, buying help starts to feel like a disclosure, and the decision gets postponed until the situation is uncomfortable enough to justify it. By then the options have narrowed, the budget is under pressure, and the work is being judged by how quickly it produces relief rather than how well it resolves the cause.

That sequence is not a failure of intelligence. It is what happens when a normal management tool is treated as a verdict on the people using it.

Opportunity. A business is considering a move it has never made: a new market, a pricing change, a system replacement, an acquisition, a leadership structure that has to change before the next stage. The work is not repair. It is choosing well while the decision is still open.

Challenge. Something has resisted internal attempts, and the usual explanations have stopped being convincing. This is equally legitimate. A struggling business is not a defective one, and asking for outside perspective during difficulty is a normal, practical response. Nothing in this argument requires anyone to feel bad about the second reason in order to accept the first.

Before engaging anyone, test the gap honestly. Internal work is usually the better answer when all of the following are true:

• The team can describe the problem precisely enough to know what a good answer would look like.
• The constraint is time, capacity, or decision authority rather than knowledge.
• Someone credible can be freed to own it, with a date attached.
• The cost of a wrong answer is contained, and can be corrected without a rebuild.

In that situation advisory work often spends its first weeks learning what the team already knows, and ownership can drift to someone who will leave. Buying capacity, including temporary external capacity, is usually the better use of the money; buying advice is not.

• Nobody in the business has made this decision before, and the cost of learning by doing is high.
• The decision crosses functions and no single leader owns it, so it keeps being deferred.
• The team cannot get the real disagreement into the room, and everyone knows it.
• A dated, evidenced read is needed faster than internal work can produce one.
• A capability has to be built and handed over, not rented indefinitely.
• Leadership needs a mirror: someone with no stake in last year's decisions to say what the pattern looks like.

The most useful scoping question is not what the consultant will produce. It is what decision will be made differently because the work happened, by whom, and on what date. From that, four things follow.

First, name the internal owner of the decision, not just the project. Second, require the reasoning, not only the recommendation: a conclusion nobody can reconstruct will not survive contact with the next quarter. Third, write the exclusions down, because a scope that covers everything produces a report that decides nothing. Fourth, agree what the team will be able to do afterwards that it cannot do now.

This is the discipline that separates advice from dependency, with one qualification. Recurring outside support can be a deliberate choice rather than a failure: retained expertise, ongoing administration, or a specialist function that is not worth staffing internally may be worth keeping. The test is not whether outside help continues, but whether the business still owns its decisions and can state what the retained support is for. Agree review and renewal criteria when the engagement is scoped, including what would end it.

A useful test before signing: ask what the business will do if the recommendation turns out to be inconvenient. If the honest answer is that it will be shelved quietly, the engagement is a way of postponing a decision, and the money is better spent making it.

It states plainly what it does not know. It disagrees with the brief when the brief is the wrong question. It makes its thinking legible enough to be argued with. It leaves artefacts behind: definitions, a decision record, a process, a measure, a shortlist with the reasoning attached. It declines to own a decision that belongs to leadership, and says so out loud rather than making the choice quietly and attributing it to analysis.

Good work also has a stated end. An engagement that cannot describe what completion looks like is a subscription, whatever the paperwork calls it.

Waiting is rarely free. The decision keeps being remade in corridors. Workarounds accumulate and become the process. The people closest to the problem conclude that raising it changes nothing, and some of them leave. None of that appears on a budget line, which is exactly why it is easy to defer. A dated decision that can be corrected is usually cheaper than an undated one everyone is waiting on.

The argument is not that every company needs outside help. It is that the companies positioned to gain most from it bring it in while the question is still a choice. That work sits in Strategy & Transformation, and for growth questions in Growth & Revenue. Two practical next reads are what a first 90 days should settle and how to sequence initiatives.