The sequence
Five phases, and two of them are exits.
A build should not begin until there is evidence it is worth doing, and it should be possible to stop at the end of the diagnostic with something useful in hand.
-
00
Scoping conversation
We establish what the operation looks like and whether there is anything here worth investigating. If there is not, we say so and the conversation ends. This is a real exit, not a formality.
-
01
Business Health Check
Vertical I. Produces the findings register, the root-cause analysis and the agreed measurement baseline. You can stop here — the register is yours either way.
What the diagnostic produces -
02
Prioritisation
Findings ranked by value, effort and certainty, and mapped to the vertical that addresses them. You choose what proceeds. Not everything found should be fixed, and some of it is not a software problem at all.
-
03
Build
Foundation before automation where the diagnostic calls for it. Each slice reaches production and delivers before the next one starts — no eighteen-month programme where the first value arrives at the end.
Why foundation comes before automation -
04
Operate & measure
Recovered value measured against the Phase 01 baseline, on the method agreed before the build. Support and iteration run against what the measurement shows.
Why the sequence is the product
Almost every failed implementation we have examined failed at a decision made before any code was written: a system was selected against an assumed problem. Phases 00 to 02 exist to make that decision on evidence, and they are deliberately cheap and short so that making them properly is never the expensive option.