Project Forecasting: More Lessons from Driving

A while ago, I entered a post about the importance of staying tuned in, drawing an analogy to driving. Well, another driving analogy had occured to me, this time about the need to focus on remaining time.

Let's put it this way. If you're driving from Philadelphia to New York City and you're at the entrance to the New Jersey Turnpike, what percent complete are you on your trip?

Some of you may guess certain percentages based on distance, but that's as foolish as basing project percent complete on the percent of budget or time that's been spent, without regard for work accomplished.

The quick answer is: Who cares what percent complete we are? What we really should be concerned with is how much time is left, assuming we care about what time we arrive to begin with.

But let's say that we DO care (i.e. schedule is a priority for us, as opposed to some other success factor). How can we measure whether we'll be there on time?

Simply using a percent complete tells us nothing. It's too subjective. What we need to know how much time is remaining. And that will depend on how fast you're going, how many miles are left, what barriers may arise (i.e. road closings, flat tires, etc.), how many stops you make, and a number of other variables. It's no different for projects.

For project schedule control, capturing percent complete is too theoretical, so that's not of much use to us. And capturing time spent tells us very little, except perhaps how long it took us to do prior work, which may not be an accurate indicator of future work. Besides, we can probably determine future work estimates more accurately through expert opinion and/or statistical sampling (combined with good planning).

Of course, there's no harm in entering time spent as long as people are disciplined to always include time remaining. Then a percent-complete can be calculated based on that. But the percent-complete itself is not a leading indicator, so is still of questionable value.

If we focus instead on time remaining at the task level, and combine that with barrier removal, risk planning, and regular reforecasts, we'd have much better control over whether we "arrive on time."

We can improve our ability to estimate in the future by capturing lessons learned, doing spot checks, and using the information to create project schedule templates and checklists, so future projects can avoid running over the same potholes.

Some may say, "Oh, we still need the percent-complete for Earned Value calculations."

Do we really? By putting a dollar amount to the time remaining, we can solve the same problem in a simpler fashion, answering the question: How much is it going to cost us to complete this project and what's our estimated time to arrival?

Just some food for thought. See my followup post on Project Forecasting and Uncertainty as well.