A client once asked us to review a reorganisation they had completed eight months earlier. Performance had not changed. Engagement scores had fallen. They wanted to know whether to do another one.
We ran a decision audit: took the twelve decisions that most affected the outcomes they cared about, and traced who actually made each one, before and after. Eleven of the twelve were made by the same person in both states. The boxes had moved; the authority had not.
The diagnostic
Before any structural work, we now insist on the same exercise. It takes about three weeks.
List the decisions that determine the outcome the strategy depends on — pricing exceptions, capital allocation between units, hiring above a threshold, whatever they are for this business. Aim for ten to fifteen, and be concrete: not "product strategy" but "which features enter the roadmap for the next quarter".
For each one, establish four things through interviews with people at three levels: who is formally accountable, who actually decides, who can veto, and how long it takes. The gaps between the first two, and the existence of unlisted vetoes, are the finding.
If the target structure does not change the answer to at least half of those questions, it is not a reorganisation. It is a chart.
What it usually surfaces
Three patterns recur. Decisions formally delegated but escalated in practice, because the delegate has accountability without budget authority. Invisible vetoes — a functional leader nobody lists on the RACI who can stop anything by withholding resource. And decision latency concentrated in one or two forums that meet monthly and are the actual constraint on the whole operating rhythm.
None of these are fixed by redrawing an organisation chart. They are fixed by changing thresholds, budgets and forum cadence, which is much less disruptive and can usually be done without a consultation process.
About a third of the time, the honest answer at the end of the diagnostic is that the client does not need a reorganisation at all. That is not a comfortable thing to tell someone who has already announced one, which is why the diagnostic has to come first.