I decided it was time to explore how to predict the end of a project when I encountered two clients this year. One made a practice of yelling at the managers and project members, demanding they complete the projects when he wanted them completed. Big surprise, that was ineffective :-) The other client uses a more agile approach, dividing each project into month-long deliverables. This client doesn’t need to yell to “make” people complete the project on time. However, the client still wants higher output. Which is a different problem. I’ll try to address that tomorrow.
From my handout, here are the three things I think are critical to being able to predict project completion:
- Estimating what you have to do. Do you know what you want to accomplish and how well can you estimate that?
- Measuring along the way. What quantitative and qualitative measurements can you take during the project to refine your prediction?
- Knowing what done means. If you know how good the product has to be and how much content the product requires, you can know what done means. Otherwise you’re never done.
If you have other ideas about what is necessary for predicting project completion, let me know. I’ve devised several simulations to explore this. We’ll have fun!