Kill Your Darlings
By Paddy Miller and Thomas Wedell-Wedellsborg
A frequent problem in organizations is the difficulty of stopping ongoing projects. Some initiatives are dictated from above and can be hard to fight. But many other projects are self-initiated and could be stopped with a local decision, only somehow that doesn’t happen, causing an unnecessary and perpetual state of project overload. How do you kill projects?
The most important rule is similar to the advice given in first-aid courses: First, stop the bleeding. If you witness a traffic accident, before you rush to the aid of the wounded, you have to stop the oncoming traffic so the accident doesn’t get any worse. Similarly, if your organization is suffering from a plague of projects, stopping a few of them only fixes the symptoms of the plague. Your first step should be to change the way projects are started. One of the best tips is to make sure all projects are launched with a kill switch of some kind—that is, tangible metrics and milestones defining when the project should be out on probation and ultimately shut down. Reframing projects as time-limited experiments can also help, so the default decision is to stop the project. Also, being clear on the overall goals of the business can help determine whether a proposed project is off strategy before it starts.
Beyond this, a few pointers can make the process of killing projects easier and (slightly) less painful:
Evaluate multiple projects at once. It’s very hard to judge ongoing projects in isolation. Evaluate all ongoing projects together, if possible.
Bring your strategy razor. Projects are often allowed to live too long because they are labeled strategic projects, without necessarily being particularly strategic. Being clear on the exact strategic aims of the business—and starting the review meeting by sharing them—can help decide what projects to cut.
Regularly schedule the reviews. Set up a specific, fixed schedule for project reviews. If not, they tend to be postponed indefinitely.
Accept the pain. Nobody likes killing projects; accept that it’s painful and do it anyway. Showing unwarranted mercy won’t remove the pain—it just prolongs it and distributes it over more time and people.