It is a well recognized concept that the overall risk inherent in a project generally decreases as you reduce its scope or complexity. This leads to the common practice of splitting complex projects into many small “mini-projects” to divide the risk across them and hopefully improve the predictability of the overall initiative.
While this is a good approach, the challenge comes when putting it into practice – how do you decide on a method for slicing the larger project and how many projects should you created?
Here are a few ideas:
1. Split the initiative by value chains – the benefit of this approach is that the integration risk for any single customer or consumer is reduced, and if the need comes to reduce the scope within the overall initiative, you can terminate an individual project and still deliver some tangible value.
2. Split the initiative by the nature of the deliverables – For example, on a systems development or implementation project, technology deliverables might benefit from being created within one project to avoid cross-project integration challenges, and it may be possible to have operational readiness deliverables (e.g. SOPs, training documentation, or maintenance procedures) created in another.
3. Split the initiative based on resource skills or availability – One method of reducing risk is to match the skills of the project managers and team members to the work they are most comfortable with delivering. This might result in increased complexity from an integration perspective in which case it might make sense to remove the responsibility for integration from each mini-project’s team and centralize it in a separate integration project management function. Resource availability might be another driver – if there are multiple resource bottlenecks, is there a way to spread these bottlenecks across the projects as opposed to focusing them within a single one? The overall guidance here is to avoid splitting resources across multiple mini-projects if at all possible.
4. Split the initiative based on deliverables with maximum inter-dependencies – To reduce integration risks, review the deliverables list from your WBS and group the ones that are very tightly related into a single project.
Regardless of which method you use to split up complex projects, the key is to perform the rough cost-benefit analysis related to reduced complexity risk but increased lines of communication (don’t forget N*(N-1)/2) and integration risk.