The investigation
A bounded inquiry, run the same way every time.
Twelve to 16 weeks. Four phases applied identically to each engagement, so that a finding in one organisation can be read against a finding in another.
01 / Terms of reference
Three conditions, agreed before the first session.
Blame is outside the terms
The question is what the system lacked, rather than who is answerable. Findings are written as system statements and carry no individual name.
Restructuring pauses
Announcements, personnel moves and communication tidying are held for the duration of the first phase. Each of them alters the record the investigation is reading.
A sponsor with authority
One executive who can act on what the investigation finds. Where the sponsor lacks that authority, the finding lands somewhere it cannot be used.
The moment an organisation knows what is being looked for, it begins managing what is seen.
02 / The four phases
What happens, and in what order.
Scene preservation
The evidence already sitting in the organisation is retrieved before anything moves: board packs, minutes, exit interviews, prior survey data, the formal chart and the working chart.
Baselines are recorded alongside it. Meeting cancellation patterns, response times, and who is consulted before a decision reaches a room.
Black box listening
One-to-one sessions across the leadership team and the layer below it. Non-attribution, verbatim record, one consistent question set mapped to what a person perceived, understood and projected.
Every claim is provisional until a second source carries it: another session, a document or a metric.
Causal chain
A dated timeline of decisions and structural changes. Each visible symptom is traced four levels past itself, and each factor is sorted into an active failure or a latent condition.
A failure of this kind rests on seven to 12 contributing factors. A chain that stops at one has stopped early.
Defence layers
Strategy, communication, trust, accountability and escalation are each scored from the evidence gathered. A single gap is survivable, and most organisations carry several.
The finding is the execution path where the gaps line up, because that is the path with no remaining defence on it.
03 / What the record shows
Five signals that survive a transcript.
- “We decided” becomes “it was decided”, and the transcript marks when.
- A topic is approached and left, in the same place, by four separate people.
- A decision is minuted as unanimous and described differently in three sessions.
- A concern raised once, years ago, is still cited as the reason concerns are not raised.
- A challenge to a senior view is put twice and dropped, rather than resolved.
Each of these is a structural signal rather than a personal one. The same transcript from a different organisation produces a different set.
04 / The documents
Four documents, and one that outlives the engagement.
Preliminary report
States whether the conditions indicate an alignment failure, what an investigation would examine and what it would cost. Where the conditions are absent, it says so.
Findings report
Numbered findings, each opening a fresh page, carrying pattern, evidence and consequence together. Sample size is carried in the sentence.
Alignment register
Every adjustment scored by severity, likelihood and detectability, owned by one named person, and given a completion indicator a third party could verify.
Embedding protocol
A declared schedule for withdrawing support, with the register re-scored at each step. The board keeps the instrument once the engagement closes.
An investigation without an intervention is an expensive document.
05 / Next
Establish whether the conditions are present.
Four minutes against the five protective layers, or 90 minutes with the executive who would sponsor the work.
The diagnostic is an indication from one source. A finding requires evidence from more than one, and that is what the investigation produces.