Ensuring the Success of a Knowledge Management System
Many KM projects fail before they've even begun because the company has failed to consider the complex interplay of technological, strategic and cultural issues involved.
Posted Apr 20, 2001
  

Many companies have invested in strategic reengineering efforts, only to see them crash and burn when a flawed deployment across the enterprise discourages employees from using the systems. Companies committed to implementing large-scale knowledge management initiatives must work to ensure that pilot projects grow into successful enterprise-wide deployments.

All KM initiatives have aspects in common. Regardless of type, experts say that many of them fail because their implementers try to do too much in too little time. Companies benefit from beginning with controlled pilot projects to build a foundation upon which they can raise a KM initiative throughout the enterprise, according to Dave Andrew Anderson, a knowledge management specialist with Alexandria, Va.-based project management and consulting company Robbins-Gioia. Even when it is done this way, however, going from pilot to full deployment encounters a host of organizational, planning and cultural issues.

start With Success

Before rolling out a pilot, the project leader should make sure the planned deployment has the potential to deliver enterprise-wide benefits. Numerous knowledge management projects are begun at the departmental level without looking at this bigger picture, according to French Caldwell, research director for knowledge management at the Gartner Group in Bethesda, Md.

"People confuse information management with knowledge management," he says. "Often it's an efficiency exercise, where they try to trim a few people but provide no new economic value." Rather than simply use KM to manage information overload and eliminate positions, Caldwell says, project leaders should seek to boost revenues by methods such as providing better information for proposals that ultimately result in increased business.

A pilot is about more than testing the validity of a project and fine-tuning it. It should also be a small success story you can use as a persuasion tool for the rest of the company, says Robbins-Gioia's Anderson.

In 1998 management at Verizon Wireless in Walnut Creek, Calif., pressed Carl Eberling, executive director of enterprise resource planning (ERP) systems, to forego a pilot and roll out an enterprise-wide management system to 500 people right away. The application, built on Vista Plus, a report-based information management system from Quest Software Inc. of Irvine, Calif., allows Verizon to transform data from an ERP or custom format into a universal format such as HTML or PDF, so it can be sent to whomever needs it.

But Eberling resisted that pressure and began instead with a small-scale deployment. By conducting a pilot with 18 users for several months, Eberling was able to send a technical support person to spend time with individuals who were having problems; that would have been impossible if the project had started with 500 users. This handholding eliminated early snafus that could have killed the project soon after it started, and the application now runs across the entire enterprise.

Buy-in by All

Successful knowledge management requires many people to embrace an idea, ranging from the executive suites to the cubicles of users. The process starts at the pilot stage and continues through the rollout. Ryder System, the Miami-based transportation and logistics firm, developed a knowledge center using Lotus Notes, QuickPlace and Sametime from Lotus Development Corp. of Cambridge, Mass., and Cincinnati-based Synergistics' Prevail Solutions Suite. The project was designed to support various internal communities and provide a way to produce better and quicker business proposals. Although that concrete objective should yield immediate benefits, Ryder executives didn't take success of the knowledge center for granted.

"We knew we couldn't make this project happen unless we looked at how it... provided value to the different communities of practice," says Dave Baildon, Ryder's group director for knowledge management. So each time the system went through another iteration in the 20-person pilot, Baildon invited people from other departments to look at it, in an effort to make sure the new design would meet their needs as well when it rolled out company-wide.

Similarly, when C-bridge Internet Solutions Inc., a 900-person e-services consultancy in Cambridge, Mass., launched a project to collect both its tacit and explicit knowledge, Ralph Rodriguez, the company's vice president and chief information officer, knew it was important for the pilot to include a diverse set of participants.

"The nontechnical people are dreamy about knowledge management," he says. "The technical people think they know everything. You have to meet in the middle." Rodriguez says an enterprise-wide KM project can succeed only when it is founded on a range of perspectives that reflect the whole organization.

To ensure that the company's intellectual property would be managed properly, Rodriguez hired someone with a master's degree in library science to identify key attributes C-bridge needed in a knowledge management system that was to span the enterprise rather than just address departmental needs.

One of the most common errors is mistaking a knowledge management project for a technology initiative. Even a seemingly straightforward IT deployment must take people-related issues into account. For example, as part of its KM system, C-bridge built its own Lightweight Directory Access Protocol (LDAP) directory to populate passwords into the database so users can access all systems through a single password.

Although the single log-on is technological in nature, it is also a practical way to effect cultural change, says Rodriguez. Technology is most useful when it minimizes change to processes that people are used to doing. "A knowledge management project can fail if you try to do too drastic a cultural change with a couple of hundred older users," he says.

Articulate the Vision

Experts agree that an enterprise-wide KM initiative stands scant chance of success unless upper management takes ownership of the project and gets everyone behind it. Executives need a full understanding of what worked and didn't in the pilot so they can determine and articulate the potential value of moving the implementation across the enterprise.

Executive support can be expressed first in a series of e-mails sent to all employees before and after the enterprise-wide rollout. C-bridge, for example, sent out a regular stream of e-mails before the company-wide launch. "We let people know who was on the team--a who's who in the company--so the company realized this wasn't a pet project of some interns," Rodriguez says. Project acceptance spread through the company as employees began to understand that harvesting knowledge was key to the future of the company.

Look at the large companies that have made knowledge management a priority (whether they call it KM or not), say the experts, and you'll often find a visionary leader with a bird's-eye view who sets the strategic direction.

Ryder, for example, tabbed for the role of "missionary" Gene Tyndall, a senior vice president who had spent 20 years building up Ernst & Young Management Consulting's global supply-chain consulting practice. Tyndall's track record with Ernst & Young and his ability to persuade even skeptical employees was vital to the success of the project.

The missionary should "reach out to individuals who are feeling left out of the success story," those who rarely participate in threaded discussions or e-mails about the project, Anderson says. "The missionary can sit down and explain the importance of it, which gives it a lot of credibility."

Train by Example

On the heels of high-level advocacy comes training, which is vital to engaging individuals in knowledge sharing, particularly when it represents a new way of thinking about the value of one's work.

The most challenging aspect of ramping up Ryder's knowledge management system to 500 people in sales, marketing and operations was communicating the benefits, according to Baildon. Information that explained what knowledge management was and why it was important was not immediately welcomed by the pilot group. "People wanted hands-on information about using the technology more than they wanted a lecture on the KM process," says Baildon.

Baildon found that the most effective approach to engaging employees in Lotus Notes-based collaboration was to deliver it to groups of 10 to 12 users at a time and make the training interactive and case-based. Ryder also offers video-based e-learning so employees can fit it into their schedule.

The Daily Context

Other practitioners also see the value in associating KM projects with people's ongoing personal job concerns. Gartner Group's Caldwell says that KM projects must occur within the context of employees' normal work duties if the organization hopes to get them to participate.

"One of the biggest barriers that we face in working with organizations is that of time: 'I don't have time to learn this,'" says Brad Jackson, founder of Cogos Consulting in Cambridge, Mass., who has worked on a number of high-level KM projects. Participation increases, he notes, when users understand how the project will help them to achieve a goal they've set or to do something specific. Therefore, Jackson argues, "Finding areas of the greatest need and being successful with them is one of the best ways to start the process and scale it to the rest of the organization."

Widespread participation isn't by itself a measure of success. But because so much of knowledge management involves knowledge transfer and collaboration, it can provide a metric for evaluation. Ryder rates employees' "knowledge-sharing proficiency" at one of three levels: occasional contributors, leaders and mentors. Every employee is expected to perform at least at the level of contributor.

In most organizations where KM has been a success, participation is encouraged but voluntary. Incentives are useful, but most companies hesitate to provide them in the form of cash compensation. Instead, Anderson recommends recognition throw the Web site, newsletters and e-mail. "People really want to be recognized for the knowledge they bring to the organizations," he says.

Know What You Want

Company-wide KM deployment requires a commitment to ongoing funding. Project leaders should identify strategic objectives and associated metrics that demonstrate the economic value of the initiative.

Caldwell advises against developing new metrics specifically for the knowledge management program. Instead it's important to use business metrics that are already being tracked within the organization such as job satisfaction or whether it results in improved business processes or innovations in products and services. A second type of metric is job effectiveness. Gartner Group categorizes this as the number of times an employee uses a particular application or the number of times a particular page on the corporate intranet is accessed.

The most relevant metric to senior managers is usually strategic effectiveness, which includes the creation of winning proposals, the attraction of important new customers, a rapid rate of innovation and growth in high-value orders, says Caldwell. "These types of measures can contribute to the basis for calculating ROI [return on investment] as well as ensuring the alignment of the enterprise KM program to business objectives and strategic direction," he says.

When Ryder put together its knowledge center proposal, the ROI was based on increasing revenue. By collecting knowledge throughout the company, Baildon says, management expected that better proposals for clients would be prepared more quickly. This in fact is what happened, he adds.

Caldwell says that putting a KM program into effect will impact business processes only gradually as people associated with those processes learn how to incorporate the new capabilities into their work. Enterprise programs shouldn't be expected to show a ROI until at least 12 months after deployment, he says.

Many KM projects fail before they've even begun because the company has failed to consider the complex interplay of technological, strategic and cultural issues involved. The payoff for the organization that can successfully balance these concerns is big: transforming itself into a corporation whose chief asset is knowledge.