When a consultancy engagement produces no lasting change, the standard diagnosis is implementation failure. The recommendations were right; the organisation just didn’t follow through. So the prescription is: do the implementation better. Clearer owners. Simpler priorities. Shorter cycles. Aligned incentives. Hold people accountable.
This sounds reasonable. But it starts from an assumption that deserves scrutiny: that the consultancy’s diagnosis was correct.
A consultancy doesn’t arrive without a frame. It arrives with one. The framework it uses to define what kind of problem you have, which levers matter, what good looks like — that was built somewhere else, for a class of organisations, and applied to yours. The report it produces is generated by that frame applied to your situation, not from your situation directly.
This matters because a frame shapes what is visible and what isn’t. The consultancy’s interviews, its data gathering, its synthesis — all of it happens through assumptions it brought with it. Those assumptions determine what gets registered as a problem and what gets filtered out as noise. If the real problem lies outside the frame, the report won’t find it. It will find the problems the frame is designed to find.
The organisation then tries to implement a solution to a problem it was handed rather than one it discovered. And the reason that better implementation — tighter accountability, shorter cycles, cleaner ownership — rarely breaks the cycle is that it operates at the wrong level. It assumes the diagnosis is sound and the gap is execution. But you cannot genuinely own a solution to a problem you haven’t recognised yourself.
The same structure appears when you hire an intelligent person and then micromanage them. Micromanagement feels like delegation — the task goes to someone else. But what actually gets delegated is the task plus a fixed worldview: how to do it, in what order, according to which priorities. The person’s own understanding of the work, which is usually what you hired them for, is never engaged. They execute your frame rather than apply their own judgment. The work gets done, but the capacity you brought in is never used. You end up with someone technically present but epistemically absent.
The consultancy situation scales this up. The organisation brings in an external party for their expertise and independent view. But it then treats the output as a fixed answer to be implemented rather than a starting point to be interrogated. The expertise was hired; the frame was not examined. So the organisation finds itself trying to execute a worldview it doesn’t actually hold.
Edgar Schein spent much of his career distinguishing what he called expert consultation from process consultation. In the expert model, the consultant diagnoses and prescribes; the client implements. In the process model, the consultant helps the client develop the capacity to diagnose and act for themselves. Schein argued that the expert model consistently underperforms, not because consultants are wrong, but because it produces solutions the client cannot sustain without the consultant present. The diagnosis was never internalised because the diagnostic work was never done inside.
This is not an argument against using consultants. External perspective is genuinely useful, particularly when the problem is structural and the people inside share the same frame. But there is a difference between using external input to sharpen your own diagnosis and renting a frame wholesale as a substitute for one.
A rented frame cannot be owned. Better implementation, tighter accountability, more energetic follow-through — none of that changes the underlying situation. If the recommendations don’t describe a problem the organisation actually recognises, the gap will keep returning. Not because of cultural resistance or execution failure, but because the diagnosis was never really theirs.
The question to ask when nothing changes is not: how do we implement this better? It is: did this diagnosis actually describe what is happening to us? That is harder to ask after commissioning an expensive report. But it is the prior question.
Further reading:
↳ The Consulting Paradox — on why organisations keep choosing large consultancies even when the pattern of non-change is visible
Garden notes
- Frame failure — what happens when the frame stops fitting and analysis inside it cannot detect the mismatch
- Why the frame cannot see itself — the instruments used to diagnose a situation are built from the same assumptions as the frame
- The consulting paradox — why organisations keep hiring large consultancies even when the pattern of non-change recurs
- The diagnosis is yours — the same problem from the other direction: a diagnosis formed entirely inside has its own distortions
- Separated knowledge — how the people who can see what is wrong are structurally separated from those with authority to act