Look for the differences in project management software!
Harvey A. Levine Feature Editor
avoid the assumption that all project management software is similar, and … look for products that best support your individual needs. ”
Harvey A. Levine
PARADIGM … IT'S DEJA VU ALL OVER AGAIN
OK! So I'm a slow reader. But as I finally got around to reading my May issue of the PMNETwork, early in July, I couldn't help notice the use of one of today's more popular buzz-words: paradigm. There it was in Tony Welsh's excellent overview on the state of today's project management software. There it was again in the article by Bill Stinnett. With my curiosity piqued, I turned to the third article, by Sandra Rapps, and 10 and behold, there was paradigm, right in the first paragraph. Never having been one to pull in four-of-a-kind, I was prepared to fold with my threesome. But this was not to be. My good friend Lois Zells delivered the fourth paradigm, in the second page of the very next article.
Was this just a casual happening … an unusual alignment of the planets, perhaps, or maybe a gremlin in the editor's computer? (Do we still have gremlins? Or have they been replaced by viruses?) Looking for meaning in this most innocent of occurrences, I was moved to reflect on the trend toward sameness in so many things (such as automobile styling) and particularly in project management software.
Yes! There is, at least on the surface, the appearance of sameness in most of the available project management software products. This appearance is exacerbated by the tendency for software reviewers to comment on all of the like features of products in a comparative review. (Actually, I fear that the need to get a “good review in the media may be driving software developers to develop products to satisfy the media rather than the needs of the users. But that's another subject.)
LOOK FOR THE DIFFERENCES
This appearance of sameness is very deceiving. It is, in fact, an outright falsehood. The real truth is that there are considerable differences in the project management software that is available, and these differences can be of significance in meeting the needs of individual users and applications. Not only are there differences in the feature sets, but there are also differences in the actual functionality and performance.
For instance, almost all of the project management software programs have resource leveling features. But the results that they deliver can vary considerably, as shown by the following example.
Efficient Resource Leveling
In one of my resource leveling test models, I have a set of tasks that use 30 person-days of a single resource. If I make two units of that resource available, the obvious best result that I can get is 15 days. In my actual tests, with eight different programs, I obtained results ranging from 15 days to 20 days. I would think that this is significant, as two people for 20 days (40 person-days) could add 33 percent to the labor costs. In addition, the project gets unnecessarily stretched out, incurring potential time and overhead penalties. Obviously, a resource manager would never knowingly allow such inefficiency, and we should expect our project management software to give us a better result.
Getting back to the test model, in most instances, the best result attainable was 17 days. The only programs that could obtain the optimal 15 day result were those that had assignment splitting. These programs, such as Lucas' Artemis Prestige, ABT's Project Workbench, Computer Associate's SuperProject 2.0 and Symantec's TimeLine 5.0 (but not TimeLine for Windows) will allow a task to be split into two or more segments if a required resource is not available for the entire duration (this occurs when the program attempts to “reserve” the scarce resource for a higher priority task). This feature, often not recognized in the reviews, will give a result more like what you would expect a line manager to do if he or she were managing the tasks on a day-to-day basis. By the way, there are even more subtle variations within the programs noted above. SuperProject requires you to select assignment splitting for the entire project. The others allow you to select it on a task-by-task basis. Prestige also allows you to specify the maximum number of splits, the minimum split duration, and maximum split interval.
Discrete Resource Loading
Another area of differentiation is in how discrete the loading of resources can be. There are still a few programs left that offer very little flexibility in how you assign resources. In these most limited cases, you specify a resource for a task and quantify the assignment (either in units pertime period or total units for the task). The resource is assumed to be required for the entire duration of the task, in a linear application.
In the majority of programs, you may specify when the resource is working within the task duration. The resources need not work for the entire task duration and may work at different levels. ABT's Project Workbench has an interesting feature that provides the benefits of both discrete resource loading and inter-ruptible task scheduling. First, you may define an assignment as uniform which spreads the resource across the task duration. Or you may choose fixed where you may specify a desired loading profile or particular work days. This can be specified either in a table or a histogram that appears side-by-side in the “variably load resources” window. But the most interesting mode is the contour mode. When selected, the resource will be scheduled to work on the task only when it is available (after supporting locked or higher priority work). Such a task is likely to be scheduled in intermittent segments, around other tasks (acting as a “split assignment” mode). The result is to cause such resources to be fully loaded in the plan.
Resource vs. Effort Driven
Continuing on the subject of resource scheduling, we have seen a continual growth in the availability of programs featuring resource-driven task durations. All programs allow duration-driven task durations or fixed durations. In the resource-driven mode, one or more resources are assigned with a total units and units per time period loading. The system calculates the task duration based on the most limiting resource assignment.
A new, third task duration mode has become available, the effort-driven mode. Here, a total resource effort is designated for a task rather than for each resource assignment. As more resources are made available, the task duration is reduced. Symantec's TimeLine for Windows and TimeLine 5.0, and Computer Associate's SuperProject 2.0 (both Windows and non-Windows versions) offer effort-driven scheduling. TimeLine provides this feature in place of the resource-driven mode. SuperProject provides resource-driven and effort-driven, as well as fixed.
The typical use of the resource planning and leveling functions consists of assigning resources to tasks, specifying the quantity of resources available, and rescheduling tasks so as not to exceed the resource limits. An extension to this function, available in SuperProject 2.0, is the ability to specify tasks that produce resources, as well as consume them. Consider, for example, a series of tasks that involve pouring concrete. You may specify the total amount of concrete required for each task, or the quantity required for each time period. Then you may also create a task that represents the delivery of concrete to the job site, and specify the rate of delivery (such as cubic yards per hour). SuperProject 2.0 provides a predesigned view that shows the tasks at the top and the material use and inventory at the bottom. You can clearly see if the scheduled concrete use exceeds the inventory (rate of delivery). If it does, you can invoke the resource leveling function to reschedule the tasks to stay within the delivered quantities.
Feature and functionality differences are not limited to the resource scheduling area. Another important feature area for today's users is project summarization. Too many of the project management software programs key their summary capabilities to a single hierarchy, namely the outline structure (usually some type of Work Breakdown Structure). Some others have stretched this to two hierarchies (WBS and OBS). If task and resource summarization is one of your hot buttons, you will want to look at Scitor's Project Scheduler 5.
PS5 has two task hierarchies (WBS and OBS), plus a resource hierarchy (RBS). Each of the ten-character task codes or the five-character resource code can be summarized using any position of the coding structure. Wildcard entries can be used, and the codes can also be used for sorting and filtering. While the concept of using these codes for summarization, sorting and faltering is certainly not unique to PS5, that program's way of handling these functions is more powerful, direct, and clear than most of the other products in its class.
HAVING IT YOUR WAY
If customization is your passion, Microsoft Project for Windows 3.0 should be able to satisfy your appetite. Users have the option to select field size and titles, optional custom text fields and number fields, as well as text fonts and colors. The Gantt chart can be extensively customized. Tables and forms can be designed and saved to a list for instant access. Most programs are offering some customization features. WinProj goes the furthest in its class.
THE MAC IS BACK
Special features are not reserved for the PC platform. In addition to PS5 and WinProj (which are available in Macintosh as well as PC versions), a completely new project scheduler was announced on July 27 by Claris. The new MacProject PRO replaces MacProject II and provides over 100 new features. Most of these features have been available in competitive products, on both the Macintosh and PC platforms. However, this new Claris release for the Macintosh has a few new features that are worth noting.
The attached documents feature lets users attach memos, reports, diagrams and spreadsheets to tasks for supporting documentation. For instance, a task may refer to a work authorization form, in the user's word processor. Such applications can be launched, under either System 6 or 7, to review the attached item.
The new outline mode mimics today's traditional outliners, with an important exception. MacProject PRO users can use the outline to brainstorm as they develop the basic project outline. Then they can select which items in the outline are to be tasks in the project schedule. In other words, the outline can contain items in addition to the networked tasks, and each item can have notes (embedded text), attached documents, and be included in the Work Breakdown Structure.
These are just a few of the differentiating factors. You will find additional differences in the ways that these programs handle imposed dates, schedule and resource tracking, bar charts and network diagrams, and reporting and graphic outputs. If there is a message here, it is to avoid the assumption that all project management software is similar, and to look for products that best support your individual needs.
REDEFINING “HIGH-END” PMS
There was a time when we tended to divide PC-based project management software into two classes: high-end and low-end. And in those days—just a few years ago-we did not expect the low-end to be able to do as much as the high-end. During the past few years, the features of low-end programs have reached and, in some cases, exceeded that of the traditional high-end products. But a rich feature set does not always equate to better performance. You should not jump to conclusions about any project management software package without first examining the particular strengths and weaknesses of the product against your specific application and needs analysis.
What about the high-end products? They certainly are not standing still. Significant developments are emerging from the better known developers of high-end project management software, such as Lucas, PSDI, Computer Aided Management, Primavera, and Welcom. Many of these developments are aimed at supporting larger, multi-project environments, in what has come to be known as Enterprise-Wide Computing (we will see this expression at least as often as “paradigm.” Lucas' Prestige for Windows, PSDI's PROJECT/2 Series X, and CAM's PARISS Enterprise are new products that should be shipping by the time this column is published. We will have to address these developments in a future issue.
In the old days, hardly very long ago in the computer age, the cost of software was usually based on the processor upon which the software was installed. A more powerful machine usually indicated more users and more use of the software, and hence, a higher price was justified for the same software. As we got further away from application-dedicated machines, this use assumption lost validity and all too often, light users were penalized for placing the software on a large computer that was being shared by other applications and users.
As the popular platform shifted from mainframes and minis to PCs and individual workstations, the modal pricing strategy shifted to “per copy” pricing. Each user workstation was expected to have its own copy of the software, regardless of the size of the CPU, or the level of use by the users. LAN pricing, where available, followed a similar strategy, being based primarily on the number “of units in the system.
In some cases, pricing was geared to “the number of simultaneous users. ” With the emergence of today's client/server configurations for enterprise-wide computing, this “simultaneous user” pricing scheme would appear to be the most equitable to both the software vendors and the users. Today, there is a trend away from the scheduling guru, toward a system where all contributors to the planning and control process (including at the executive level) either want or are encouraged to access the software directly. Charging the limited user the same fee as the full-time user is not only unfair, it seems, but also serves to discourage the purchase and use of such software.
Because of the difficulty in administering software use control, many vendors and companies are reaching agreements on corporate-wide, unlimited use. However, with the proliferation of project management software use by small companies, the corporate pricing structure is often out of reach. Again, this all points to the “number -of- simultaneous-users” approach as a potentially fairest solution. We would be pleased to hear from our vendor and user readers on this subject, and will attempt to share such replies with the readership.
Harvey A. Levine, president, Project Knowledge Group (21 Pine-Ridge, Saratogo Springs, NY 12866) has been a practitioner of project management for over thirty years, primarily with General Electric Company, and is a past chairman of PMI. Mr. Levine has been adjunct professor of project management at Rensselaer Polytechnic Institute, Troy, NY and is the author of the book Project Management Using Microcomputers as well as several articles.
OCTOBER 1992 pm network
PMI research shows project teams that draw from an array of perspectives and skillsets deliver powerful outcomes.