By Bill Pepoon
Some of you may have noticed an article in the September 2002 edition of PMI Today regarding the formation of the College of Scheduling. This is only one of two PMI colleges (the other is the College of Performance Management) that have been proposed since PMI established guidelines for colleges in June of this year. According to PMI's website, a college is required to expand PMI's Project Management Body of Knowledge, whereas a SIG is a group of project management professionals who share similar interests. Of course, some SIG's (like D-P-C) are also interested in expanding the PMBOK, but it's not a requirement for maintaining our charter. Anyone interested in becoming a charter member of this college should click on the following link: http://www.pmi.org/collegeinfo
Judging by the board positions that have already been filled, the College of Scheduling already appears to be staffed by some individuals who know quite a lot about project management and scheduling. Nevertheless, they don't strike me as "hands-on" guys. Don't get me wrong. They know their stuff and they have used the software (one of them is co-founder of Primavera Systems!) but I can't help but wonder if they will spend too much time focusing on new "rules" for proper scheduling to somehow enforce better scheduling practices. The problem with this is that in many cases rules come down to personal preferences. Somebody thinks negative lags are a bad idea so he doesn't want anyone else using them, for example.
I see a lot of bad scheduling in my line of work, and most of it has very little to do with the existing rules not working. As a test, I ask people during the seminars that I teach a simple, but misleading, question: "How many open ends should a proper CPM schedule have?" By "open end" I mean of course activities that have no predecessors or successors. Someone will usually answer too quickly and say "none". Wrong. All proper CPM schedules have exactly two open ends. The very first activity in the schedule will obviously not have a predecessor, and the last activity will never have a successor.
On occasion I will find more open ends in a schedule that I'm reviewing, but most decent schedulers take the time to check their work before sending it out. Yes, the truly hapless ones will force dates using constraints and other tricks and leave out important activity relationships, but in many cases those individuals were in a hurry and/or did not have sufficient information to prepare the schedule. I’ve had a few clients who were far more interested in making sure that certain activities start on certain dates than discussing logic. In effect, they are trying to create a bar chart. My job is to determine whether these dates are feasible when all other factors are considered.
More often that not, bad schedules omit important work items. Hey, where's the roof? Believe it or not, I saw that once. This is a content issue and really has nothing to do with scheduling techniques. And I don't know why exactly, but bad schedules always seem to go from one extreme to the other: not enough activities to WAY too many. Tell them there's not enough detail in the schedule and next thing you know, they've got the lunch breaks in there. C'mon guys!
I recently took over a schedule from another consultant who was having trouble getting the initial schedule approved by the architect. While there were some problems with the logic and activity durations, the architect was really incensed by all the junk that didn't need to be in the schedule. For instance, every procurement item was broken down into a seven-step process:
The procurement phase of the schedule was actually bigger than the work phase, an area where the architect felt more detail was necessary. My philosophy has always been that the procurement phase is a summary of the contractor's submittal log, which clearly must list everything in the specifications that requires review and/or approval by the architect. Why the consultant had gone to the trouble of adding literally hundreds of activities that no one (including his own client) cared to see in the schedule, was a complete mystery. Then again, he was being paid by the hour!
When I started out as a scheduler I was taught that each procurement item should consist of three activities: (1) submit, (2) approve, and (3) fabrication/delivery. I've sometimes thought about inserting "resubmit" between steps 2 and 3, since re-submittals are a pretty common occurrence, but overall I've gotten by with just three activities for nearly twenty years. It may not be a rule but it is pretty good common sense. Besides which, I've never had a client ask for more detail in the procurement phase.
Common sense is actually an important part of scheduling. The level of detail, activity descriptions, codes, lag durations, etc. are driven by practical matters. A schedule with too much detail is not necessarily "wrong" but it becomes cumbersome to update and requires more adjustments as the work invariably gets done in a different manner than planned. Make the activity descriptions concise. Activity codes can be very useful for sorting activities, but why add codes that are redundant or superfluous?
The bigger problem is simply raising the awareness of the importance of scheduling. If more company executives were really concerned about scheduling the quality would improve. Most schedules today are given about as much consideration as the fax machine. When CPM was still relatively new and exciting, there were many people who were willing to believe that it could make a real difference. Then it seemed to stop working in the early 1990's and people started to give up. What went wrong?
Well, the advent of cheap PC's and off-the-shelf scheduling software allowed quite a few people who had no business calling themselves schedulers to muck up all kinds of projects. Unfortunately, the company executives (who knew even less about scheduling) blamed the process and not the people. How silly. I mean, has anyone been to a bad meeting? We didn't stop having meetings altogether as a result, did we?
I'm inclined to think that the reason alternative scheduling methods like Critical Chain Project Management are attracting so much attention is not because they are really any better than CPM but because they are not called "CPM". If something better truly came along I'd be the first in line, because it would make me look so cutting edge. So far CPM still looks better than the pretenders to the throne.
One of the goals of College of Scheduling (I suspect) is to put their imprimatur on scheduling, that if you follow certain procedures then you are a good scheduler. Whether that is a practical approach remains to be seen. Scheduling is still an art to a certain extent, something you learn through trial-and-error.