We call the image below a “declining work balance” graph. Part of its purpose is to predict when a project will be completed. Using appropriate estimating techniques (we’ll discuss those techniques another time) we’ve found we can be accurate to within about a week at predicting when a project will be done. This takes the guesswork out of when a long-term project will be completed and counteracts the schedule optimism that infects project participants.
Over the last few years we’ve noticed a trend in our project quality assurance work: being late (delivering a project later than planned) matters less to clients than it used to. We’re not talking a week or two late but often months or years late. With integrators and software developers mostly under fixed price contracts there’s little financial pressure to the buyer to be on time. Legislatures and the public don’t seem to care as much as they used to – we conjecture that they have suffered from so many late deliveries that late is now the norm. And being late is certainly better than failure. Does late matter? Well perhaps it doesn’t. An old project manager of ours used to say: “People don’t remember if a project was delivered late, they do remember if a project failed”.
However, your time is worth something. While you are tied up with late projects there are costs not only in salary but also in lost opportunity. “Opportunity Cost” is a measure of other, possibly more valuable, things you could be doing with your time. Further we’ll bet the “back fill” provided while you work on a project doesn’t do your job as well as you do so your agency is less effective. So perhaps being late does matter more than people recognize.