Percent complete problems
Posted by ElizabethHow many times have you heard in a status meeting that a particular task is 80% complete? It’s comforting, right? You know the task is nearly done and the person reporting has a good handle on what still needs to be completed for the final 20% is reached.
Unfortunately, you can’t deduce any of that from a percent complete measurement. The security that you might feel when someone tells you that a task is nearly done is misplaced. And 80% is a popular ‘almost done but not really’ number, which you will hear over and over from the members of your project team who don’t understand that this is not a good way to estimate completion and work remaining.
Measuring percent complete is common practice. Software like Microsoft Project even does it for you. It encourages you to enter a percent complete which then rolls up from sub-tasks to umbrella tasks to give you a final percent complete for the entire project. But while it might be common it is a dangerous way to measure project progress.
Meri Williams puts it like this in her book, The Principles of Project Management:
People are overoptimistic.
How often have you thought something would take half the amount [of] time that it actually took?
People lie.
They know when you want the job done and will reassure you that it’ll be ready, even if they haven’t actually started (often this comes after the overoptimism mentioned above).
Work expands to fill the available time.
Even if a task seems to be half-done after two days, if you’ve allocated six days to it, there’s a good chance that it will take six days to achieve.
Overoptimism is the most significant problem. Unless it is a task that has been done many times before, chances are that you don’t know exactly how long it will take, however good the estimating process. If you used to leave your college essays until the night before as you reckoned they wouldn’t take that long to do you will understand this estimating dilemma!
So, how can you get round estimating project progress by percent complete? There are a number of better ways to estimate progress. Earned Value Analysis is one of those, but I have always felt this is overkill for smaller projects. By the time you have set it up, the project will be completed. It is also only as good as the data you have to hand to plug in. However, on large projects EVA is a safe bet and its growing use in the UK is testament to the fact that some project managers find it incredibly useful.
Another approach is to track time to completion. Instead of looking at how much you have done on a particular task, look at how much there is still to do. Ask your team members to give you updates in terms of work remaining. Forget percentages, go for days, or hours, or another unit of time that makes sense to you and the project. It changes the mindset of the team to think about how much more effort is required instead of an arbitrary percent complete. You could measure effort to completion or elapsed time to completion, but make sure that everyone is clear what you choose: a task that has four hours remaining could be completed this afternoon, but if that is four hours of effort spread out over four weeks the project could be another month off seeing this particular task ticked off on the plan.
An easier, and completely non-scientific way, of measuring progress is to adopt a binary approach to completion. A task is either complete or it is not. Anything not complete requires further work. For an even clearer approach, Williams recommends measuring the number of deliverables completed against total deliverables instead of tracking task activity.
Whichever method you choose, try to think carefully about how it will impact the working practices of the team. Make it as easy as possible for them to give you meaningful updates – you’ll get better data and more accurate status reports. That will give you a much better chance of hitting your scheduled dates!
Related posts: