From the client’s side
An outcome no single function can deliver.
Sales to supply chain, a line no one owns, a number the executive committee has already agreed. This is the shape almost every engagement takes, whatever the sector, told as a pattern rather than a case.
The shape of the work
The same six disciplines, from your side.
The same six disciplines that The System describes, seen from your side and numbered to match. The numbers are the correspondence, not the order things happen in. The sequence itself is the five stages of COREV.
You hold the stop button, and you use it on day 90.
Five gates in 90 days, and none of them opens because a meeting went well. The problem and the measure come first, and the solution is chosen second. Your own people supply the evidence at each gate, your executive sponsor sees it at the midpoint, and on day 90 that sponsor makes the call. Scale or stop. What it asks of you is a sponsor who will attend and a willingness to stop.
You set the limit on what the machines may do.
The boundary is set by a named person on your side before the work starts, and it is enforced as machine readable policy rather than as team convention. Agents build inside it. Humans own Direct, Verify and Evolve. What changes for you is that the limit on autonomy is a document you authored and can point at in an audit, rather than a habit you have to trust.
Someone who has run your function sits inside the team.
Someone who has already run the function being fixed works inside the team while the work happens, drawn from the mentor network, not reviewing from a steering committee. What changes for you is who you argue with: a person who has carried the same decision, in the room, while it is still cheap to change.
The people who start are the people who finish.
One contract, one accountability, and the same operators from the first day to the running system. The people in the room at Demo Day are the people who were there in week one. What that spares you is the re tender, the re briefing, and the quarter most programmes lose between the strategy and the build.
You write the number before we write the plan.
It is expressed as a number the board is willing to be held to, not as a project with a deliverables list. The number comes before the plan. That number is written into the specification at Capture and it is the same yardstick at Value, which means you are held to it as well.
You may be told to buy, or to do nothing.
The build or buy call is decided on total cost of ownership and durability, not on speed to a demonstration. Sometimes the honest answer is to buy, or to do less. We hold no software licence revenue, so nothing on our side depends on the answer being build.
Across all six: adoption, control and the build are set up together on day one rather than bolted on once the demonstration works. A working thing that no one adopts is not an outcome.
One engagement, as illustration
AI everywhere, and no one holding the line.
An organisation was using AI across the business with no control, no governance and no documented process. The work ran from return on investment and business case definition through to a comprehensive technical architecture, with change management and governance built in from the first week rather than added at the end.
Details are generalised and the client is not identified, in line with our confidentiality obligations.