- Article by Ramanathan V, SetProve Consultancy
“Ram, most of our projects are going be-yond timelines and the estimated efforts. I don't think people are to be blamed as I see this problem even for projects executed by my best project manager” - confessed Surya, CEO of Startsoft Technology.
The first round of meeting with the project leads led me to a large list of excuses like “unforeseeable difficulties”. “took longer than expected”, “unexpected changes from client”. Even the most prestigious projects where everyone’s attention was focused and every expert was ready to assist - even those had slipped on schedules or effort estimates. They believed their estimations were wrong and tried to implement Function Point Analysis (FPA) with pain but no gain.
I noticed a common thread in almost all excuses that they are all expressions of uncertainty. Estimation cannot be certain because we talking about work to be done in future. Extrapolations are not always correct! But every project manager knows about it! That is why he adds a “safety factor‟ to all his estimates. Yet this has not given any comfort or increased the level of confidence for completing the project on time.
Even in FPA there are estimates, and most project manager don't understand why and how they estimate truly. Let use take an example. If we ask how long it will take (make an estimate) to reach home from office, you may say about 25 minutes. If we ask what you mean by “about‟, then explanations could be like “it may take 15 minutes or some times 1 hour - you see if there is some unforeseen problem like a flat tire it may take even 1 hour and 20 minutes. So “about‟ means “more often”. You can substitute a good driver to a good programmer, flat tire to technical glitch etc. So in this example if you call an experienced project manger to commit to an estimate, you will be surprised to see he may estimate it at around 45 minutes. But often he cannot explain his adding such a large safety factor moving estimate from 25 minutes to 45, except he may say it is based on “experience‟.
It is true that often project managers quote arithmetic mean as estimate. Then they will add some amount to it as safety. People who understand statistics immediately imagine a normal bell shaped curve but in reality if they can draw a graph on occurrences it will look like the one given below. In this distribution (unlike normal distribution) the number that covers most occurrences (say 90%@) is much farther than arithmetic mean or even median value. So project mangers must understand statistics in its practical application scenarios to provide good estimates.
But it still does not answer the question that if there is a large safety factor why do the project schedules slip!
Senior management too are guilty of their lack of understanding of statistics. Very often in the negotiation they try to bring the estimate number closer to averages (mean, median or mode) as they argue that average represent most likely scenario. So even the large safety factors are significantly trimmed without understanding underlying distribution.
By the way how many in senior management or project management know when to use “mean‟, “median‟ or “mode‟ or know differences between normal and skewed distribution? So FPA or not, project managers start with wrong understanding of distribution.
But the worst part is that after the estimation and as soon as the work commences they forget about the purpose behind the safety factor. You will very often see that “Late start date‟ is used more often than “Early start date‟. Don't you know most school students do their homework only at the last minute?
But it is not always the psychology behind using “late start‟. If you are to get material from vendors to start that task, delaying from “early start‟ to “late start‟ means delaying material procurement, that means delaying investment till the last minute. Every prudent finance manager would love that! Given every project manager works with resource constraints - more time (meaning late start) gives the buffer to ensure resource availability.
So mere introduction of sophisticated tools like FPA may not guarantee success unless the project managers have an in depth understanding of what they are using, why they are using and more critically how it impacts their project work.
Mere project schedules on MS project software or knowing about PERT/CPM will not help. When I read the book “Critical Chain‟$ I realized why often CPM (critical path method) did not help me! We require project managers who continue to learn new ideas in project management. They have to understand and improve their knowledge on practical application aspects of methods. If estimation is about future and thus brings uncertainty, we must credit project managers who do anticipate other risks that bring more uncertainty. Risk identification and management is critical act every project manager is supposed to do.
In this specific case I noticed that the team is not clear on a few aspects of specification and the project manager has written is his project plan that they must obtain clarity quickly. But yet in the midst of the project, this requirement is buried, but to resurface when they were in the midst of a crisis. Then the project manager is forced to accept what the customer says even if it will take more time. Now I see why there is a reason for delay like ‟unexpected changes from client‟.
Most risk management has apart from “contingency planning#‟ a key element namely “mitigation planning‟. If contingency plan is re-active strategy, mitigation plan is pro-active. It details what the project manager should do to mitigate (or reduce chances of) risk occurrence. Often mitigation plans are not prepared, but the equally sad part is that even when prepared they are forgotten (like safety factor!).
The reason ‟unexpected changes from client‟ should read as “not implementing mitigation plan‟. Project management more often is not about living with uncertainty but ability to understand what causes uncertainty and managing it. Understanding uncertainty begins with learning new subjects, techniques and continues with exploring the difficulty when those learning fail in practice.
$ - Critical Chain by Eliyahu M Goldratt
# - Contingency plans include specific strategies and actions to deal with specific variances to assumptions resulting in a particular problem
