Programmes and projects rarely fail everywhere. They fail at a few points: the architecture decision nobody takes; the cutover nobody wants to own; the stream that holds up all the others. That is where we start. We work inside your team, take over the leadership or the critical stream, and stay until every signed success criterion has been accepted and the programme has been handed over to your operations.
Almost every programme we are called into looks the same in the SteerCo: the report shows progress, the dates are “on track”, the budget is “within limits”. Six weeks before go-live it all tips at once. Not because anyone lied — but because nobody ever established what success is supposed to look like.
The last row is empty because it does not exist in most reports. That is exactly where our model starts.
“We are at 80 per cent.”
For four months now. Progress is counted in completed tasks, not in results somebody has accepted and can use.The architecture decision sits with the SteerCo.
For three meetings now. Nobody feels responsible, everyone waits for the other, and meanwhile the team is building on an assumption.The service provider delivers to contract.
Just not what the project needs. The statement of work dates from a time when the goal was different.The business units have “no capacity”.
For testing, acceptance, data cleansing. The project depends on people who were never in the project plan.The cutover date is fixed — the way back is not.
There is a go-live plan, but no abort criterion, no rollback rehearsal and nobody who takes the decision at three in the morning.Every stakeholder means something different by “success”.
The board is thinking about the date, the CIO about stability, the business unit about function, operations about handover. None of these definitions is written down anywhere.Five stages, one principle: before we steer, we establish with everyone involved how the success of this programme will be recognised — measurable, accepted, with a date. Stage two is the core. Everything after it is steering against those criteria instead of against a plan that was still right at kick-off.
Two to three weeks inside the project, not above it: conversations with the team, the business units, the service providers, the sponsor. We check status reports against actual work products and find the points where the programme is really stuck.
Situation report · pressure pointsWith the sponsor, the CIO, the business unit and operations: what must be true at the end for this programme to have been a success? One metric, one person to sign it off and one date per criterion. What does not make this list does not get built.
Success criteria map · signedTest the plan, the scope, the contracts and the roles against the criteria. What does not contribute is cut or deferred. Open decisions get a decision-maker and a date — not an agenda item.
Cleaned-up plan · decision registerWe take over the leadership or the critical stream and are present where things tip: architecture decision, provider escalation, test acceptance, cutover. The SteerCo sees criteria status, not percentage of tasks done.
Criteria status at every meetingA criterion is met when the person signing it off has confirmed it — not when the team is finished. Once all of them are met, the programme goes to your operations, documented. We stay until it runs without us.
Acceptance record per criterionThe success criteria map is the one document that stays with the SteerCo through the whole programme. On the left, what project mandates usually say. On the right, what we turn that into before we steer.
The map is signed at the realignment kick-off by the sponsor, the CIO and those who sign off the criteria. Changes to the goal are allowed — but only as a change to this map, visible to everyone.
We are not equally present everywhere in the programme. We are where experience says the outcome is decided — and at those points we take over ourselves instead of advising.
Signed before anything is built. Renegotiated as soon as the goal changes — not quietly adjusted.
One person, one date, one rationale. We prepare it and we get the decision taken — even against the wish to defer it.
Check the statement of work against the criteria, run the escalation, negotiate change orders. The provider delivers what the programme needs.
Go/no-go with a written abort criterion, a rehearsed rollback, named decision-makers for the night. We are in the room.
Criteria accepted by operations, not by the project. Knowledge transfer with evidence. Only then do we leave.
Both steered against signed success criteria. Anonymised; reference conversations after pre-qualification.
Six streams — AD, server, client, network & applications, Exchange, remediation. New ESAE tiering architecture, PAWs with smart cards for all domain admins, 300 DMZ servers hardened, krbtgt reset against golden and silver tickets.
Insourcing strategy, governance framework, CI/CD with GitLab and integration with Microsoft Entra ID, 73 workloads across three tiers, a team of 23 FTE, 25+ knowledge transfer workshops, hypercare with simulated incidents. Budget held.
