Portal Best Practices: It’s Time to Wake Up—Again!
“After nearly two years of dreamy infatuation with the corporate portal, the enterprise is waking up. Many companies that bought into portals as the killer application for knowledge management have found themselves instead with solutions that don’t come close to delivering the expected functions or performance.”
Many readers will recognize that quote—it appeared in print in October, 2000. So what has changed as we move into 2003? The answer, unfortunately, is “not much.” A Gartner Research study in October, 20021 offers the following profile of portal experience in Diagram 1 (Page 4, Special Supplement or refer to PDF version).
Clearly, many portal projects still end up as shelfware, underutilized, or have only modest success. The reasons, we believe, are twofold: (1) organizations have been slow to recognize the difference between simple “portals” and full-blown enterprise information portals (EIPs) or enterprise knowledge portals (EKPs), and (2) that the same best practices learned on large information technology projects can and should be applied to EIP/EKP projects.
Don’t Get Fooled
The physical task of setting up a portal is not very difficult or expensive. Many organizations have prepared Web sites that allow employees to review a database of company policies, or that allow purchasers to check a database listing products and stock on hand. However, an EIP or EKP is a different thing altogether, much in the way that a data warehouse differs from a database that supports order management. The task of inventorying and organizing a large body of structured, semi-structured, and unstructured content is daunting. Configuring the portal to provide effective access while enforcing needed security adds additional complexity, and the creation of a refresh and retire process that ensures timeliness and accuracy of the content is a major effort in its own right.
Start with an Appropriate Project Type
A good place to start is to identify the types of portal projects. This usually dictates the logical project phases. The P4 project model (Diagram 2) recognizes four project types: Prototype, Pilot, Pathfinder, and Production.
As shown, prototypes are good when the organization has little experience with the basic concepts and technologies involved, and needs the opportunity for basic experimentation. Pilot projects attempt to develop approaches that are candidates for production implementation. Typically, an “easy” knowledge domain is selected so that the project can focus on the mechanics of the processes and technologies. Pathfinder projects validate an approach using real-case scenarios that represent typical rather than worst case scenarios, and involve users from the operational environment. Production projects, obviously, are expected to roll out permanent new knowledge capabilities to the organization at large.
Recognize that Standard IT “Best Practices” Do Apply
In all phases, large EIP or EKP projects have a lot in common with other complex information technology projects. Yet all too often, organizations fail to recognize that lessons learned and “best practices” from information technology projects also apply to portal projects.
Information technology best practices are proven in many organizations across many industry sectors, including government and not-for-profit enterprises. Many of these same best practices, which fall into the following categories, should be applied to major portal projects:
- Alignment with organizational objectives
- Scope control
- Measurement
- Not forgetting the “how”
- Planning for maintenance
- Looking ahead.
While adopting best practices cannot guarantee success, ignoring best practices body of knowledge can significantly raise the odds of failure.
Alignment with Organizational Objectives
Before embarking on an EIP/EKP project, ask the obvious: What is our organization trying to accomplish and how will an EIP/EKP help to achieve those objectives? Many sophisticated organizations in North America, such as Coca-Cola and AT&T, are using portfolio management techniques2 to ensure that information technology resources and spending are aligned with overall organizational objectives.
Portfolio management techniques help define exactly what problems an EIP/EKP project will address, help to place EIP/EKP initiatives in the context of other strategic programs under consideration, and make sure that those problems are the ones important to the organization.
Several researchers have emphasized the need for executive support as a necessary condition for EIP/EKP success. Support at the executive level is important, but it is insufficient in itself. The sponsoring executive must be provided with the information she or he requires in order to make the investment case.
Scope Control
An EIP/EKP cannot be all things to all users. If your organization is in the middle of an effort that is trying to construct a “one portal per western hemisphere” environment, stop. Be like DuPont:
DuPont estimates up to $66 million in savings from the first phase of its portal for its 550-person sales and marketing unit.3
DuPont identified the scope of its portal effort and managed the features and information that would be made available in a series of phases. This approach to controlling the scope of the project provided a measured benefit and has led to similar initiatives in other areas.
A valuable technique that can be applied to determining the scope of the portal effort is the “Use Case” approach to understanding the system and its use. It defines what the portal needs to do in order to affect the tasks and how the tasks will change when the portal is in place, and findings are used to establish measurements for portal effectiveness.
Measurement
The importance of measurement is emphasized over and over in systems work. Six-Sigma, ISO 9002, and the SEI Maturity models all point to measurement as a mandatory feature of any successful organization. Yet when it comes to measurement, too many EIP/EKP efforts take the “Field of Dreams” approach: i.e., build it and they will come.
If organizations do not measure the factors that define success, how can anybody know if the effort was a success? The time to establish the measures of success is at the beginning, when organizational alignment is being considered. Success must be organizational success if it is going to be relevant.
Not Forgetting the “How”
In his study on Portal Pitfalls4, Craig Roth of META Group recommends: “Conduct an infrastructure impact assessment. This should be performed after features have been inventoried, but before selecting a portal product. The assessment should cover all infrastructure services that the portal will depend on but not provide itself: authentication and single sign-on, directory access, content/document management, workflow, enterprise application integration, collaboration, and search.”
An EIP or EKP is not a Philosopher’s Stone that magically turns the dross of data into the gold of valuable information without effort. The work to be done is the basic “blocking and tackling” of software development. It has been shown over and over in enterprise application integration, electronic commerce, and portal projects that as much as 80% of the cost and difficulty associated with implementation are due to two primary factors:
- The complexity of hooking up the existing infrastructure components, including applications and middleware
- Problems with data quality and consistency.
It is essential to employ best practices in planning efforts and spend sufficient time in examining the various relevant factors before choosing portal tools or announcing target dates.
Planning for Maintenance
Eighty percent of the cost of software over its useful life is incurred after the product is initially deployed. There will be an impact on the Help Desk when the EIP/EKP becomes available. Resources, technology and time must be budgeted to handle change requests, which will start appearing within hours of deployment. A real education program will need to be in place to ensure that organizational benefits are realized. Compared to many operational support systems, maintaining and managing the portal requires a significant expansion of the dimensions that must be tracked and controlled. The information and its sources in the EIP may be far more varied in format and content than the data typically handled by invoicing or inventory systems (which may themselves be sources for the portal). The impact on users may be greater as they do more self-definition of their job functions.
Best practices exist for the maintenance of complex information technology environments. These are fully applicable to EIP or EKP projects. Why not take advantage of that knowledge to fully harvest the expected value from a portal initiative?
Looking Ahead
Many of us look at portals as the tool that will deliver the promised benefits. We have a lot invested and committed to this proposition. However, there is a danger that we have put on some blinders. It is possible to see “Web services” on the horizon. If we consider, for example, IBM’s vision for computing in the future and its offering of services on demand, what does that do to our current thinking about portals? Will we need to develop them or “own” them?
The practice of keeping eyes open, of looking ahead, is also part of the knowledge and best practices that have been gained over years of information technology activity.
Why Not Benefit from Knowledge Management Practices?
Ironically, being in the business of knowledge management and knowing when to benefit from our own expertise do not seem to go hand in hand. We cannot afford to discount and ignore a proven body of knowledge—information technology best practices—that can provide valuable assistance to us. It may be time to just stop the rush towards getting a portal in place, and to apply some of the knowledge we have at our disposal. By recognizing the true nature of EIPs and EKPs as complex information technology projects, and applying information technology best practices, all of our portal projects can end up as “visible” or even “resounding successes”—exceeding even the highest expectations.
1. “Management Update: Six Ways That Portal Projects Can Fail or Succeed;” Gartner Group, IGG-10092002-03; R. Vales, D. Gootzit, G. Phifer; 9 October 2002.
2. For information on IT Portfolio Management, see “The Information Paradox” by John Thorp (McGraw-Hill, February 1999).
3. “Unifying the Extended Enterprise”, Chris Grejtak, E-Vision Report from Broadvision, Inc.
4. “Top 10 portal pitfalls—and how to avoid them,” Craig Roth, META Group, May 2002 Special Report on Portals. Keep in mind that these results span a wide range of portals.