When a major programme is clearly in trouble – deadlines missed, costs rising, confidence falling – there’s a strong pull towards starting again. New leadership, a new plan, perhaps a new supplier. A fresh start feels decisive.

It’s also expensive, slow and often unnecessary. A restart throws away the work that’s going well alongside the work that isn’t, resets the learning curve, and usually adds months before anything useful happens. Most troubled programmes can be recovered without one.

Here’s how we approach programme recovery.

The warning signs of a programme in trouble

Recovery is far easier when it starts early. The warning signs are usually visible well before a programme is formally declared red:

• milestones are being met, but nobody can say what has improved
• the same risks appear on the RAID log month after month
• the benefits case hasn’t been revisited since approval
• key people have left and haven’t been replaced
• steering meetings are spent explaining rather than deciding
• confidence among senior sponsors is quietly falling

If 3 or more of these are true, it’s time to act – whatever the status report says.

Why a restart is tempting – and usually wrong

A restart offers a clean narrative: what went before was the problem, and new people with a new plan will fix it. For sponsors under pressure, that’s attractive.

But most troubled programmes aren’t failing everywhere. Some workstreams are delivering. Valuable knowledge has been built up – about the systems, the data, the stakeholders and what doesn’t work. A restart discards all of it and pays to rebuild it. It also tends to repeat the original mistakes, because the underlying causes – unclear ownership, weak decision-making, unrealistic benefits – haven’t been diagnosed.

“A restart discards what you’ve learnt and usually repeats what went wrong.”

7 steps to recover a programme

1. Establish what’s actually true

The first step is an independent, rapid baseline – done in weeks, not months. What has been delivered? What’s working? What’s the real status of the plan, the budget and the benefits? Which assumptions have turned out to be wrong?

This needs fresh eyes and the standing to say uncomfortable things. Its purpose isn’t to assign blame but to replace competing narratives with a shared, evidenced picture leadership can act on.

2. Separate what’s working from what isn’t

Protect the workstreams that are on track and focus recovery effort on the ones causing most of the problems, rather than disrupting everything.

3. Re-anchor the benefits and their owners

Programmes drift when they lose sight of why they exist. Revisit the benefits case: which benefits still hold, which need to change and which should be dropped. Make sure each material benefit has a named business owner who has accepted it.

4. Reset the rhythm and the decisions

Troubled programmes often have too much governance and too few decisions. Simplify. Establish a clear operating rhythm, a single reporting model and forums that exist to decide. Then make the decisions that have been waiting.

5. Bring in the specific expertise the problem needs

Recovery rarely needs a large new team. It needs the right specialists, quickly, aimed at the actual problem. When a master-data risk surfaced in one live transition programme, 2 specialists were mobilised from our network within days to address it.

6. Show progress quickly

Confidence returns when people see things improving. Identify a small number of visible wins that can be delivered in the first few weeks of recovery, and make sure they’re noticed.

7. Decide on evidence – including whether to stop

Occasionally a recovery baseline shows the programme should be fundamentally changed or stopped. That’s a legitimate outcome, and a far better one than continuing out of sunk cost. The point of recovery isn’t to save the programme at all costs; it’s to make sure the organisation is investing in something that will deliver.

When a restart is right

There are times when starting again is the correct answer: when the original objective is no longer valid, when the chosen solution can’t meet the need, or when trust has broken down beyond repair. But that should be a conclusion reached from evidence, not a reflex.

In most cases, the faster, cheaper and less disruptive route is to find out what’s true, keep what’s working, fix what isn’t and get decisions moving again.

Common questions

How do you recover a failing programme?

Start with a rapid, independent baseline of what’s true; protect what’s working; re-anchor the benefits and their owners; reset governance so forums decide; add targeted expertise; deliver visible early wins; and decide on evidence whether to continue, change or stop.

When should a programme be stopped rather than recovered?

When the original objective is no longer valid, the solution can’t meet the need, or the evidence shows the benefits no longer justify the remaining cost. Stopping on evidence is a sign of good governance, not failure.

How long does programme recovery take?

The diagnostic baseline typically takes weeks. Visible improvement should follow within the first few months; full recovery depends on the programme’s size and the issues found.

What is programme assurance?

Independent, evidence-based review of whether a programme is on track to deliver its intended outcomes, providing early warning of risk and challenge to leadership.

What this looks like in practice

When a water utility’s internal benefits team shrank from 5 people to 2, with roughly half its change portfolio on hold, a single senior practitioner embedded alongside the project lead protected momentum on critical workstreams and restored confident year-end reporting – leaving a benefits framework and reporting templates the client owned afterwards. Read the case study →