There is no such thing as a transformation programme that fails suddenly. They fail slowly, loudly, and in ways that were visible from the first governance meeting.
Most of the warning signs look like normal programme activity — until someone experienced enough to see the difference walks through the door.
The problem is that most organisations don't know what they're looking at. The warning signs look like normal programme activity: reports being produced, RAG statuses being updated, steering committees being held. It takes an experienced eye to see that the programme is producing governance theatre rather than delivery progress.
Here are the five signals we look for.
1. The programme has a plan but no theory of change
A plan tells you what will happen and when. A theory of change tells you why those things will produce the outcome you're aiming for.
Most transformation programmes have plans. Almost none have documented theories of change. This matters because when something doesn't go to plan — and something always doesn't go to plan — you need to understand which assumptions were wrong, not just which milestones were missed.
Ask the programme director: "If we deliver everything in this plan, why will the organisation be different?" If they can't answer in two sentences, the programme doesn't have a theory of change. It has an activity list.
2. Stakeholders attend but don't decide
Governance committees are supposed to make decisions. When they stop making decisions — when every agenda item results in an action to "take offline" or "bring back with more information" — the programme has a power problem.
Someone doesn't want the programme to succeed, or the committee doesn't have the authority it thinks it has, or the real decisions are being made somewhere else. Any of these is fatal.
A programme we were brought in to review had 24 steering committee meetings over 18 months. It had produced 0 formal decisions. Every major question was escalated upward, deferred, or quietly dropped. The programme was 14 months behind schedule and nobody could explain why, because the decisions that would have kept it on schedule had never been made.
3. The definition of done keeps moving
Scope creep is real but it's not the main problem. The main problem is unclear scope: requirements that are technically accepted but not genuinely agreed, deliverables that are declared complete before they've been tested, milestones that are marked green based on activity rather than outcomes.
When the definition of done moves, it almost always moves in the direction that makes the programme look better on a report. That's human nature. It's also a programme management failure.
The fix is to define acceptance criteria in advance, at the deliverable level, and to test them with the people who will actually use the output — not the people who produced it.
4. The programme team is isolated from the business
Transformation programmes are, by definition, trying to change the way an organisation works. That means the people who will live with the change need to be involved in designing it. Not consulted. Not updated. Involved.
When a programme team operates in isolation — producing designs, architectures, and process maps in a room without meaningful input from the operational staff who will implement them — the outputs are almost always wrong. Not wrong technically. Wrong for the context.
We've seen this pattern in three consecutive programmes at the same financial services firm. Each programme delivered its outputs on time. None of the outputs were used as intended.
5. Risk registers describe the past
A risk register that lists things that have already happened as "risks" isn't a risk management tool. It's a log. The same is true of a register that lists every possible bad thing with no priority, no owner, and no mitigation that anyone is actually doing.
Real risk management is forward-looking and uncomfortable. It involves saying "this specific thing is likely to go wrong in the next six weeks, here's why, and here's what we're doing about it." It involves surfacing risks that someone senior doesn't want to hear.
When risk registers become comfortable documents — when nobody reads them carefully in a steering committee because everyone knows they'll look the same next month — the programme has stopped doing risk management.
If you recognise more than two of these in a programme you're running, the time to act is now, not at the next quarterly review. Talk to us about programme health assessments.