Nine signs a program is in trouble before the date slips

Programs rarely fail on the day the date moves. By then the evidence has usually been accumulating for a quarter or more. These are the signals that show up first, in roughly the order they appear.

1. The status has been amber for months

Amber is a stable state in most reporting, which is the problem. It means “we are managing it” and it can hold that meaning indefinitely. Ask what would have to change for it to go green, and by when. If nobody can answer in a sentence, it is red.

2. The plan has stopped changing

A plan that has not been re-baselined while the work underneath it changed is no longer a plan. It is a record of what was believed at the start. Healthy programs replan; the absence of change usually means the plan stopped being used rather than that everything is going well.

3. Testing keeps moving right

Every phase before testing absorbs a little slippage. Testing absorbs it all, because it sits at the end and the date after it is fixed. When the test window shrinks twice, the quality of what goes live is already decided.

4. Change requests arrive faster than they are closed

The count matters less than the trend. A rising open balance means scope is being agreed faster than it is being delivered, and the gap will be paid for at the end — in cost, in date, or in what gets quietly dropped.

5. The people who know the process stopped attending

Business subject-matter experts disengage when they conclude the design is set and their input will not change it. It is a quiet signal and a serious one, because it usually predicts a user acceptance phase that finds things the design should have caught.

6. Scope is moving to a phase two nobody has costed

Deferring scope is legitimate. Deferring it to an unfunded, undated phase two is how a program hits its date and still fails to deliver the benefit. Ask what phase two contains, what it costs, and who has agreed to fund it.

7. Defects are being reclassified rather than fixed

When severity definitions start moving, or when a growing number of defects become “known issues” or “documentation items”, the quality picture in the pack is no longer the quality picture in the product.

8. Reporting only travels up through the delivery team

Status that is written by the people responsible for the outcome will be optimistic. Not dishonestly — it is how confidence works under pressure. A program without a reporting line independent of delivery has no mechanism for bad news to arrive early.

9. Nobody can state what “done” means

Ask three people what has to be true for the program to be finished. If the answers differ, or if they are phrased as activities rather than as testable outcomes, the end date is an estimate of effort rather than a commitment to a result.

What to do with this

Any one of these is worth a question. Three or more together means the date on the pack is not the date that will happen, and the earlier that is said out loud, the more options remain. The cheapest intervention is almost always a re-plan with the current facts; the most expensive is discovering the same facts during testing.