When a project falls behind plan, what should change?

Every engineering organisation measures plan versus reality in some form — a burn-up chart, a sprint report, a slipping roadmap. Most of it is consumed as status: a number to be noted, occasionally worried about, rarely acted upon. That is the waste. A plan-versus-reality signal is only worth producing if it changes a decision.

The three decisions it should feed

• Reallocate. The plan assumed a distribution of effort that reality has moved. Where is capacity now most valuable, and what should stop?
• Rescope. The gap is not a delivery failure but an estimate that met new information. Which scope should be renegotiated before more is spent against it?
• Renegotiate. The cause is outside engineering — a customer delay, a missing decision, a dependency. The right move is a conversation, not more hours.

If a plan-versus-reality report cannot be traced to one of these, it is a status update wearing a chart’s clothing.

Why it is hard

The gap is easy to compute and hard to attribute. Was the team slow, or the plan optimistic, or the customer late, or the context missing? Attribution is where the decision lives, and it is exactly what a raw velocity number cannot give you. Getting it right, without reducing an engineer to a single activity score, is the work.

This is an argument from practice, not a study. Its whole value rests on a single condition: that someone can act on the gap while it still matters.