Anyone taking over a program that is going off the rails has an information problem, not a motivation problem. The first few weeks decide whether it becomes a controllable undertaking.
In the first few weeks, a program in trouble does not need a new plan but a robust picture of the situation. The reflex to present a revised schedule straight away produces the second commitment that is not kept — and costs what is left of your credibility.
The order I work in: week one, the assessment; week two, reconciling commitments with reality; week three, forcing the open decisions; week four, a new plan that rests on decisions rather than on hopes.
Programs in trouble almost always show the same symptoms: dates slip, status reports become vaguer, the people involved talk about each other instead of to each other, and committees discuss details because the big questions are uncomfortable.
The causes underneath are equally limited in number and usually four: a scope that was never properly delimited. Dependencies on parties that never gave a commitment. Decisions that have been open for weeks and that nobody insists on. And a reporting line that smooths problems on their way up.
Treating the symptoms — more reports, tighter cycles, more pressure — makes matters worse, because additional effort is created without a single cause disappearing. The work of the first few weeks consists of getting from symptoms to causes and documenting that in writing.
I look at the traffic light last. Three other signals are more reliable, because they cannot be dressed up.
The first: there is no reliable overall picture. Each workstream reports for itself, and no one can say where the program stands as a whole. Anyone who first has to reconstruct the overall status is not steering — they are administering individual reports. As long as that picture is missing, any statement about the final date is guesswork.
The second: decisions are deferred. The same open question is on the list for the third governance meeting in a row. That is not a capacity problem but a question of authority — and it costs more time than any technical difficulty, because in the meantime everyone carries on as if the question had been settled.
The third: dependencies are unresolved. Two areas are waiting for each other without it being written down anywhere. As a rule this only surfaces when there is no time left to replan. That is why, during the assessment, I ask everyone involved explicitly who they are currently waiting for and since when — the answers expose these gaps faster than any schedule.
The order is the actual substance: understand, reconcile, decide, plan. Reverse it and you produce the next broken commitment.
One-to-one conversations with every role — the business, development, operations, the supplier, the PMO. The same questions in each case: What has been committed? Which of it do you consider achievable? What are you missing? What is not being talked about? In this week I assess nothing and decide nothing — I listen and I take notes. Anyone who passes judgement in week one hears nothing from week two onwards.
All commitments are collected: dates, scopes, resources, dependencies. Then each one is checked against a source. For every commitment there are exactly three states: evidenced, unevidenced, disproved. The result is usually uncomfortable and always valuable — it is the first list in the program that nobody doubts.
The assessment produces a list of open decisions, each with options for action, consequences and a named person with decision-making authority. This list goes to a governance body with a clear message: without these decisions there is no reliable date. It is the most uncomfortable and the most effective step.
Only now does a plan emerge. It rests on decisions taken, confirmed dependencies and effort estimates that come from the people who will do the work. Everything not confirmed is stated in it as an assumption with a date — visible, not hidden. Such a plan is often more uncomfortable than the old one, but it holds.
The assessment depends on a few questions that are always the same. Always the same, so that the answers become comparable — the contradictions between the answers are the actual finding. I allow around 45 minutes per conversation:
The suggestion to clarify the situation in one large workshop comes up reliably in situations like this. It is well meant and, at this stage, wrong. In a group where mistrust and blame are already in the room, nobody says what they really see. You get the official version — precisely what is already in the reports anyway.
One-to-one conversations deliver the opposite: differing versions of the same set of facts. These contradictions are the most valuable raw material of the first few weeks, because they lie exactly where the causes sit. The workshop has its place later — when it is about options and decisions, not about describing the situation.
On confidentiality I say the same sentence in every conversation: statements are consolidated by topic, not by name. I hold to that even when attribution is demanded. Anyone who uses the assessment to attribute statements to individuals immediately loses all further access to information — and with it the basis of their own work.
The most common and most expensive reaction. A new date without a new basis for decisions is the same date with a different number on it. In week one, the right answer to the question "when will it be finished?" is: "I can tell you reliably in four weeks — before that it would be guesswork."
Weekly instead of monthly reports create more effort for the people who are supposed to deliver and change nothing about the causes. Consolidating reporting can make sense — but only once it is clear which information is actually missing.
Before the assessment there is no way to tell whether someone is not delivering or cannot deliver because a decision is missing. Personnel changes at the outset cost knowledge and deter exactly the people who would speak openly.
Reducing scope is often the right measure — but as a visible decision with agreement, not as a quiet omission. Silent cuts come to light at acceptance, and by then the damage to trust is greater than the original delay.
I decline in these four cases, and I would rather say so in the first conversation than in the third month. A recovery engagement without these preconditions ends with someone carrying responsibility for a situation they are not permitted to change.
If the person engaging me can neither take the pending decisions nor reliably bring them about, the leverage is missing. So I ask early: who decides on scope, date and budget — by name? And can I get to that person?
If the engagement arrives with the cause already in its luggage — the supplier is to blame, the team is too slow, all it needs is a new plan — then there is no room for an assessment. I am then meant to confirm an existing opinion, not to clarify the situation.
A politically set date is normal and often understandable. It becomes a problem when it is immovable before anyone even knows what is still open. The outcome is then predetermined, and my role would be to certify it.
Steering without access to scope, prioritisation and staffing is a title, not a role. Before I give my commitment, I clarify which decisions I take myself, which I prepare and who takes them. That delineation belongs in the engagement in writing.
The success of the first thirty days does not show in a new schedule but in three observations. First: governance bodies discuss decisions instead of status. Second: those involved name risks before they materialise, because naming them no longer has consequences. Third: when I ask what the most critical open point currently is, I get the same answer from different people.
The third point is the most important. A programme does not become manageable at the moment a plan exists, but at the moment everyone has the same picture of the situation. Everything before that is the administration of misunderstandings.
Describe the starting situation, the people involved and the role envisaged. You will receive a candid assessment of what needs to be done in the first weeks — and whether I am the right person for it.
Discuss your project