D02 – Start, end, or cancel projects

This is a draft of the manual for review. If you have any comments, please email them to info@omimo.org. Please check back for the final version.

What

Each program has multiple projects under it. Those projects start based on the target Scenario and hopefully end as expected. Rather than the normal ending, some projects may be canceled. These major events can happen any day during the cycle.

The cyclic go/no-go decisions of projects will be indirectly responded to by this activity.

Why

The program’s projects are defined to help achieve the expected outcomes of the program and, therefore, should be completely aligned with the end goal. This is key, especially because programs have a high level of uncertainty and cannot be run by a plan that’s fully defined and frozen upfront. Instead, they should be closely monitored and adjusted as needed.

Who

The decision to start the project should be taken by the project sponsor, program sponsor, and program manager. The other decisions will be made by the project sponsor and program manager, and the program sponsor will be involved only if there are doubts or if the impact of the decision is too high.

How

Each project should be initiated by creating a simple team and a common understanding (usually a high-level plan). Toward the end of project initiation, there should be a go/no-go decision. After that, there should be frequent go/no-go decisions as well (e.g., every month). Finally, the last decision is to end the project when it’s complete. All these decisions will be made in this program activity.

To make the decision, the Business Case of the project will be updated based on its latest forecasts of investment and benefit. The impact of the update will be evaluated according to the target Scenario. If the update is too different from the Scenario and difficult to evaluate, the monthly cycle will be run to replan everything.

Don’t keep projects that are nearly done open for long, but try to conclude and close them as soon as possible. In general, it’s best to limit the number of projects that are run simultaneously in order to simplify their management and increase predictability.