This article is intended for well-qualified PMs interested in new approaches. While it invokes statistical concepts, no mathematics beyond simple arithmetic is required. Even PMs who favor another approach will benefit from knowing this one, perhaps employing it as a crosscheck.
The article is self-contained. No collateral reading is required, although links are provided to interesting sources. The approach targets managing a single project, nothing broader.
This method works best for well-defined projects with clear success criteria. Such projects have crisp milestone dates and deliverables. Organizations that have consistently recorded their past project data will obtain superior results.
Our typical project size is roughly a year in length or more, with its first major milestone at 3 or 4 months. IT development projects are a good fit, as are manufacturing and construction projects, although a little less so, as they are better understood and subject to less variability. Healthcare projects, which include information processing, administration, and public services, are an emerging high-growth category. Because they are new, there is a much smaller database of historical project results, and that dearth discourages the use of this approach. However, this class will become more important over time, so we should monitor it as a potential candidate.
The more critical the project's result, the more important it is to use every tool in the bag to get it right. The method is more justified on important projects than on routine ones. If you and your project fit this description, please read on.