We are often brought into projects already underway and going badly. The causes repeat so consistently that the warning signs are almost a checklist, and every one of them is visible well before the deadline everyone is worried about.
The signs, in the order they appear
| Warning sign | What it usually means | When it appears |
|---|---|---|
| Meetings attended by deputies | The sponsor has disengaged | Early |
| Decisions deferred to next week, twice | No one is empowered to decide | Early |
| "We will handle that in phase two" rising | Scope is being hidden, not managed | Middle |
| Test data still not ready | The data problem is bigger than admitted | Middle |
| Training postponed | Configuration is not actually finished | Late |
| Go-live date moved without a new plan | The plan was never bottom-up | Late |
Scope is rarely the real cause
Scope gets blamed because it is measurable. The actual cause is usually ownership: no single person inside the business is accountable for the outcome, so decisions queue and nobody has authority to break the tie.
The question that unblocks most stalls
Who can decide this today, without asking anyone else? If the answer is nobody, the problem is not the software or the scope.
Restarting a stalled project
- Stop building. Nothing improves while work continues on an unclear target.
- Write down what is genuinely live and used, honestly.
- Cut scope to one process that can go live within eight weeks.
- Name one accountable owner inside the business, not the vendor.
- Re-baseline the plan bottom-up, then commit to it publicly.
Restarted projects that follow this get moving again more often than not. The ones that do not are almost always the ones that skipped step four.
