When to stop a project

A project that should stop often carries on instead. It slows, changes shape and absorbs more money and attention, until it is either finished at a cost nobody would have approved at the start or abandoned without a decision ever being made.

Stopping a project is not a failure of the people running it. It is a decision, and it deserves the same preparation as the decision to start. The best time to prepare for it is before any work begins.

Write down the reasons the project exists

At the start, write a short statement of what the project is for. The result it must deliver, the benefit the sponsor expects from that result, and the limits of time and money within which the benefit is still worth having.

This statement is the measure against which every later decision is taken. When a project runs into trouble, the question is not whether it can still be finished. Almost anything can be finished with enough money and time. The question is whether finishing it still delivers the benefit, within limits the sponsor would accept.

Set the conditions for stopping in advance

Alongside the purpose, agree the conditions under which the sponsor will consider stopping. A forecast cost beyond a stated ceiling. A completion date after which the result is no longer useful. A change in the sponsor's own plans that removes the need. A dependency that fails, such as a permit refused or a site that cannot be secured.

Written in advance, these conditions are calm and impersonal. Raised for the first time in the middle of a difficult project, the same points sound like blame, and people defend the project instead of examining it.

Review at fixed points

Divide the project into stages, and at the end of each stage hold a short review against the original purpose and the stopping conditions. The project manager presents an honest forecast of the cost and time to finish, the risks that remain, and whether the expected benefit still holds. The sponsor then decides whether to continue, change the project or stop.

The value of fixed review points is that the question is asked whether or not anyone is worried. Projects rarely announce that they should stop. The signal builds slowly, in forecasts that creep and benefits that fade, and a scheduled review is often the only moment anyone looks at the whole picture.

Judge on what is left to spend

The money already spent on a project is gone, whatever happens next. It is natural to feel that stopping wastes it, and that feeling keeps many projects going too long. The useful comparison is between the cost still to come and the benefit still to be had.

Suppose a project has spent most of its budget, and the forecast to finish has grown. If the remaining cost is still worth the result, continue. If it is not, stop, however much has already been spent. The forecast has to be honest for this to work, which is one reason the person preparing it should not be the person most attached to the project.

Stop in an orderly way

A decision to stop should be carried out as carefully as a decision to finish. Make the site or the system safe. End or suspend contracts on their terms, and agree final accounts with each supplier. Keep what has value, such as designs, completed work, materials and data, and record where it is. Write a short account of why the project stopped and what was learned, so the next project starts better informed.

Pausing is sometimes better than stopping outright, when the need may come back. If so, record the state of the project well enough that someone else could restart it, and set a date to decide whether the pause becomes a stop.

A sponsor who stops a project well frees the remaining money for something that still serves the purpose.

We look before we commit.