In Software Project Management, Time is Money
The nice thing about software development projects is that the project manager rarely has to worry directly about the budget. If the developers you need for the team are already on your staff, you just require a clear idea of how long you will need these people for. In other words, time is the triple-constraint element you will be most concerned with. Obviously Budget is always important but in this case, I think you know the schedule is what you will be fixated on!
So the development and tracking of the project schedule will be the central concern for you as the project manager. Given how important this is, how effective are you at tracking your schedule? Generally, schedule management revolves around meeting milestones. Have we produced everything we said we would at this point? If the answer is “yes”, then we are on schedule and all is well.
Or is it? I have heard of one company (which shall be nameless) where milestones are traditionally scheduled for Mondays, so the team can make up for any delays by working weekends. All the projects seem to track to schedule, but the completion of each project is accompanied by a wave of resignations. So tracking to milestones keeps poor estimates hidden, by working the team to death.
Is there some way we can make this phenomenon visible? Well, if you take the view that time is money, you have a clue: use Earned Value Analysis. If you have completed a PMP® exam preparation course, or taken the PMP® exam itself, you will remember that Earned Value is a means of tracking costs firstly but can also be very useful for SV & SPI measurement. So, if you replace currency as your measure with time, this technique can be an effective schedule tracking mechanism as well.
For example, suppose we have a task that we have estimated as 30-hours’ work. Historical records suggest that we are doing well if an individual can devote 30 hours a week to project work, so let us schedule a team member to start this on Monday morning, expecting completion by close-of-business Friday. In other words, our Planned Value (PV) is 30 hours.
Friday arrives and the deliverable is completed by close-of-business, so our Earned Value (EV) is also 30 hours. Our Schedule Performance Index (SPI = EV/PV) is 1, so we are bang on schedule, or are we?
Next week we study the actual times taken to complete the tasks. We find that the team member actually put 40 hours into reaching the milestone and spent an hour every day consulting with the software architect to resolve technical issues. This means that the Actual Cost (AC) of this deliverable is 45 hours, making the Cost Performance Index (CPI = EV/AC) 30/45 = 2/3. In other words, we underestimated this task by 50%!
Keeping track of actual values will improve the accuracy of your estimates over time – if only by amassing historical data. However, we must acknowledge that there are some organizations with “hero cultures” of long hours and 7-day weeks. Here, the project manager is under pressure to underestimate tasks deliberately. This is grossly unethical and should be avoided. Propose a technique like Critical Chaining instead. This is another PMP® certification topic; one that is covered in the Time Management knowledge area.