A good schedule doesn't tell what you want to hear, it tells you what you need to know, so you must ...
- A good schedule is predictive - it shows what is going to happen in the future and what the alternatives are if that doesn't actually happen
- A good schedule is reflective - it shows where the project stands in relations to the planned position against the actual work that has been accomplished
- A good schedule is dynamic - it can be adjusted when the reality of the project changes
- Build the requirements in a requirements management tool
Trace these requirements to the work activities in the schedule. Only do the work that produces the planned value to the customer. This is a "deliverables based" schedule.
- Build the PLAN before you build the SCHEDULE
Define a strategy for the successful completion of the project - the PLAN. Do this before defining the work steps needed to implement the PLAN.
- Manage the project with a project management tool
Excel and Power Point are worthless. All tasks need successors and predecessors. So don't use your favorite project management tool as a generator of Power Point presentations. You need a critical path, you need to know how much slack you have for each deliverable, you need to know what resources are assigned to work work. You need a credible schedule.
- Make task duration fit the duration guide
Answer the question: How long are you willing to wait before you find out you're late? Make the task duration match this risk.
- Use a RACI or RAM chart to assign accountability for all the deliverables
Single points of accountability are required for all deliverables. A single named person is accountable. None of this collective responsibility stuff. That only works in the agile training books.
- Every task has a deliverable
Do only the work that produces value to the customer. No level of effort in the plan or the schedule. If the work doesn't produce something of value, then why are you doing it?
- Have a Plan B and a Plan C
Bad things always happen to good projects. Plan for it. Have the Plan B built into the schedule.
- All cost and duration values are random variables
Know what the variances are and you'll have the basis of a credible schedule. Assume they are scalar values and you have a nice Power Point presentation
- In the end, it's all about the people
Work gets done through people. Have the best people possible, treat them right, reward them appropriately
The resulting schedule, derived from the Plan is actually many plans and schedules...
