SOFTWARE IN PRACTICE
Congratulations on Your New Project
Congratulations. Your board of directors has handed you an exciting new systems project. It's large in dollars and scope, it's complex, it applies interesting new technologies and its critical to the company's viability.
Mmmm you wonder. Is this an opportunity to make your mark as a professional project manager ... or is it a chance to gamble your career on a risky enterprise that could easily fail?
The board has handed you a project charter. They want your strategy. Board members are eyeing you carefully; waiting for an answer.
You know that well planned projects seldom fail. So you brief them on the project planning process - like this:
Documenting the Deliverables
Thank you ladies and gentlemen. Now firstly I need to document what the project will deliver.
To achieve that my team will capture the customer's requirements, do some preliminary design and define the project's scope of works. For that I'll use a work breakdown structure (a WBS). The WBS will describe the scope of the end deliverable and any services needed to execute the project. I'll expand the WBS down to the work package level to give us a good handle on resourcing, scheduling and budgeting.
Who Will do the Work?
Next up I'll determine who'll do the work. I'll identify what's to be done in-house, what will be contracted out and what components will be purchased off-the-shelf. This process will allow me to allocate work packages to various performing organisations.
Finding the Milestones
As soon as possible we need to firm up our start and end dates and the intermediate milestones that will allow us to measure progress. We also need an overall framework to drive detailed planning. I'll achieve this with a master schedule. I'll develop the master schedule by identifying the major activities and milestones in our work packages. I'll then analyse work package interfaces to work out our major activity dependencies.
Assembling the Cost Estimate
Now I know you'd like a cost estimate right now but bare with me. When we've expanded our WBS to work package level we'll be in a position to generate a preliminary cost estimate that's plus or minus ten percent accurate. I'll generate the estimate by working through each work package using knowledge from previous projects and other top down estimating techniques.
Setting up Control Points
Next up we'll commence the real toil of detailed bottom-up scheduling and estimating. This work will be done by various planning teams with domain knowledge of the target technologies. I'll identify these teams using the WBS. By designating various WBS elements as management control points I can partition the plan into logical components. During project execution each control point becomes a control account where we'll accumulate project performance information.
For each control account I'll appoint a Control Account Manager (we call them CAMs). A CAM has responsibility for detailed planning and performance measurement.
Control Account Managers prepare detailed schedules for all work packages in the scope of their control accounts. The major milestones in each work package will trace to those of the master schedule. We call this vertical traceability.
If, on the basis of detailed planning, we discover that master schedule milestones can't be met... we'll negotiate.
Control Account Managers also prepare detailed bottom-up cost estimates for each control account. These figures will be reconciled with the preliminary cost estimates and any cost blowouts justified.
Establishing the Performance Measurement Baseline
To measure project performance accurately on a weekly basis, we need to distribute our lump sum estimates over time. Control account cost estimates will therefore be integrated with the schedule to produce time phased budgets. We need time phased budgets to implement earned value management. In the earned value domain the time phased budget becomes our planned value.
Once the control account budgets are complete I'll roll them up to produce an accurate time phased budget for the project. I'll then present this budget for your approval. On approval the budget will become the project's Performance Measurement Baseline.
Evolving the Performance Measurement Baseline
I must point out that developing the Performance Measurement Baseline is an iterative process. We'll go through more than one cycle of scheduling and estimating before we arrive at a baseline that's acceptable to everyone.
What Could Go Wrong?
And now I'd like to talk about what could go wrong; and what we're going to do about it.
About this talk
You're tasked with planning a large and complex project. Can you explain your planning strategy off the top of your head? In this talk Les Chambers works through the essentials of an effective project planning process that can be scaled for small, medium and large projects. The culmination of the process is a performance measurement baseline that provides clear visibility of progress throughout a project's life cycle.