Ask for Help
Programmes do not fail because problems occur; they fail because problems are known for weeks before anyone says so. Asking for help early is a capability, and it is one a programme either builds deliberately or destroys accidentally.
Why it matters
Every recovered programme has the same forensic finding: the issue that sank the schedule was visible to someone months before it was raised. The question is never why nobody knew. It is why nobody said.
- Cost compounds with delay: A design problem found in workshop is a conversation; found in testing it is rework; found at cutover it is the date.
- Silence is rational when raising is punished: If the first person who escalated was blamed, nobody else will.
- Status becomes fiction: A programme where nothing is ever red is not a healthy programme.
Putting it into practice
- 01
Make raising cheap
A low-ceremony route that does not require a paper, a forum or a confrontation.
- 02
Respond to the first one well
The treatment of the first escalation sets the behaviour for the rest.
- 03
Separate the problem from the person
The issue is about the work, and it is discussed that way.
- 04
Give it a service level
Raised issues get a response in a defined time, or people stop raising them.
- 05
Close the loop
The person who raised it is told what happened.
- 06
Watch the leading indicators
All-green status and zero escalations from a large workstream are warnings.
A programme with no red items and a slipping date does not have a reporting problem. It has a truth problem, and the reporting is the symptom.
What you need in place
- A route that is not the line management chain: Some problems are about the chain.
- Response service levels: Raised issues acknowledged and actioned in a defined time.
- Visible response: People need to see that raising something worked.
- Independent assurance: Someone whose role is to find what is not being said.
- Leaders who raise their own problems: Behaviour is set from the top, not from the policy.
Questions to ask
01First escalation
What happened to the last person who raised something unwelcome?
Everyone in the programme already knows the answer.
02Route
How does someone raise a concern without going through their manager?
If there is no route, some concerns will not be raised.
03Latency
How long between a problem being known and being reported?
Ask the team, not the register.
04Status
Is anything red, and has it ever been?
Perpetual green in a complex programme is not good news.
05Response
What service level applies to a raised issue?
Unanswered escalations train people not to escalate.
Metrics and KPIs to track
Issue age at raise
Known to reported, sampled by asking.
Escalation volume
By workstream — silence from a large one is a signal.
Response time
Raised to acknowledged and actioned.
Status distribution
Red, amber and green over time.
Late surprises
Issues first reported within a month of a gate.
Alongside this