Evidence before opinion: how to diagnose a business in twelve weeks

Consultancy has a credibility problem, and it is largely self-inflicted. Too many diagnoses arrive as opinion wearing a lanyard: a senior person walks the floor for a fortnight, matches what they see to the framework they brought with them, and presents conclusions the organisation quietly rejects the moment the invoice clears. People do not resist change. They resist verdicts they had no hand in.

Over the years, across banks, law firms and trading platforms, I have refined a different approach. It is not clever. It is disciplined. And it works at any scale, from a three-firm financial group to a twelve-person business.

Listen at scale, in their words

The raw material is what the organisation already knows about itself. Structured interviews, working sessions and artefact review, captured as individual observations in the teams' own words. In a recent large engagement that meant over sixteen hundred observations from fifty-four contributors; on a Diagnostic Day it might mean forty from six people. The number matters less than the principle: every finding must trace back to something a real person actually said or a real document actually shows. When a leadership team asks "says who?", the answer is "your people, and here is the index".

Classify until patterns fall out

Individual observations are anecdotes. Grouped honestly, they become evidence. The observations get clustered into categories, categories into themes, themes into a handful of structural patterns, and suddenly three hundred complaints about different things reveal themselves as five instances of the same thing. The most useful discipline at this stage is writing each theme as behaviour and consequence: because the business does X, Y follows. That framing describes the system rather than blaming the people, which is precisely why people accept it.

Play it back honestly, including the good news

A credible diagnosis names strengths with the same rigour as weaknesses. "Operationally strong but structurally fragmented" lands, where a deficit-only report gets defended against. The playback should be recognisable to the people who contributed; the moment a team sees its own words in the picture, the picture stops being the consultant's and starts being theirs. That transfer of ownership is the whole game.

Show your work, weekly

Trust is built in transit, not at the destination. A short written update every week, same format, same day, showing what was done, what emerged and what comes next. No surprises at the final playback, because leadership watched the evidence accumulate in real time. It also imposes a useful discipline on the consultant: if you cannot say what you learned this week in two pages, you did not learn much.

Sequence the fix in dependency order

The output is a roadmap in which each step creates the foundation the next depends on: inventory before automation, ownership before workflow, baseline before measurement. An ambition-ordered plan reads well and fails quietly; a dependency-ordered plan is executable. And every line of it should trace back through the themes to the original evidence, so that a year later, when someone asks why a workstream exists, the answer is still on the record.

None of this requires an enterprise budget. It requires the discipline to gather evidence before forming opinion, and the honesty to play it back whole. That travels to any size of business, which is rather the point.