In the Know
In the Know
Novell’s knowledge management culture saves millions and improves service quality.
By Donna Fluss
Customer Interface Magazine
In 1997, Novell, a $1 billion leader in eBusiness solutions and Net Services” was struggling to provide consistent responses to callers. Like many companies providing technical support, answers to customer questions and incidents varied based on the expertise and knowledge of the support person or engineer handling each call. One customer could receive three or four different answers to the same question from speaking to multiple technical support engineers (TSEs). While each answer could be right, since there are often multiple approaches to resolving a problem, this inconsistency was inefficient, expensive and annoying to internal and external customers. Novell wanted to resolve the problem quickly and knew that a knowledge management (KM) system that seamlessly integrated with their call tracking system was the right answer. Still they had grave concerns about the success rates of KM implementations, since over 80 percent fail. Novell senior management was willing to invest millions if a technology investment would fix the problem, but didn’t want to become another failure statistic.
Novell’s Customer Service Organization
Novell Technical Services was using a Vantive call tracking system to manage the cases. The Vantive system was great at logging and tracking incidents but was not a KM tool designed to manage an active and rapidly changing knowledge base (KB). Within the Vantive system, TSEs were able to document customer solutions and recommendations in a Technical Information Document (TID) and could associate a TID with each case, but could not effectively manage and manipulate the database of TIDs.
Novell Technical Services handles over 40 products and receives over 300 incidents daily, of which 86 percent are calls to its two primary support centers in Provo, Utah and in The Netherlands. The call centers provide 24/7, follow-the-sun support, to all qualified callers. Some customers are serviced in person by dedicated field engineers but still have access to the call centers. Due to the variability of products supported, the only feasible way to be consistent with customer inquiries is to have a flexible knowledge base of predefined “solutions” for common problems. Enterprises that provide technical support are dependent on their ability to collect, store, recollect and communicate pre-established databases of knowledge to their customers. Most companies use informal databases of knowledge but Novell, having already gone down that path with less than total success, wanted to invest in a formal KM system that integrated with their call tracking system and gave them sophisticated tools to author, edit, publish manage and report.
After a careful market assessment of solutions and existing KM implementations, Novell decided the benefits of KM were worth the risks and selected Primus’ Solution Publisher (now called eSupport), Solution Explorer (now called eServerWeb) and Solution Builder (now called eServer Desktop). To mitigate its financial exposure and risk, however, Novell decided on a six-month trial involving only 25 support engineers. If the beta failed, they would lose their implementation costs, but nothing else.
Before beginning the beta, Novell carefully laid out their objectives, established a baseline and methods of measuring the results on an ongoing basis. Novell defined both productivity and quality objectives:
- Increase efficiency,
- Reduce escalations to level 2 and 3 support staff,
- Reduce learning curve of support staff, and
- Increase customers’ ability to service themselves.
- Improve consistency of customer responses,
- Improve accuracy of the KM of solutions, and
- Allow Novell’s extended community to contribute “solutions.”
Having made the decision to invest in KM, Novell moved quickly and by March of 1998, 25 support engineers were ceding and piloting the new Primus KB. Five months into the beta, the system test was a dismal failure. Technical support personnel using the new Primus KM system complained that it was difficult to navigate and took too much time to create new “solutions.” The new KB only exacerbated the problems the support staff had already encountered with the Vantive system.
Disappointed but not willing to quit, Novell took an in-depth look at their entire customer service operation and discovered that their motivation and reward systems were measuring and encouraging the wrong behavior. They were rewarding their support staff for the quantity of solutions created, rather than quality and appropriateness. In doing so they were creating a second dirty database with duplicate “solutions,” just as they had already done in the Vantive system. Up to this point, Novell had made the classic mistake of seeing their KM project as a pure systems effort and hadn’t modified their operating procedures, policies, or training programs, prior to implementing the new system. Once Novell realized that their KM implementation was really the beginning of an all-encompassing corporate business transformation, they changed their measurement and reward systems (see Figure 1). Only then did the effort immediately become a success.
According to Novell, “It’s not just a tool-set, it’s a cultural revolution! The major implementation hurdle in a knowledge-sharing system is not the technology. It is in getting people used to it. Sharing ideas freely is one thing were you at a physical meeting and can see reaction. It is another when you are working electronically and the reactions of others may or may not come.” (See Figure 2.)
Novell has invested $2.6 million in its three-year KM initiative. This cost includes all hardware, software, implementation costs, training, maintenance and internal support. Amazingly, within 11 months from the time of the full implementation of the new KB, the Primus KM system had a positive ROI and started to pay for itself (see Figure 3).
Novell has a classic three-tiered support structure. Tier 1 customer service representatives are the first line of defense and in 1997 they were no more than message takers, gathering each caller’s company information, documenting the problem and verifying the customer’s Service Level Agreement. Tier 1 representatives received, on average, one week of training before being put on the phone. Tier 1 message takers passed the messages onto the Tier 2 Technical Support Engineers (TSEs). TSEs are Certified Novell Engineers and typically solved 90 percent of the inquiries they received. The remaining 10 percent required the assistance of Tier 3 Resolution Engineers and 2 percent of these required action on the part of the development team.
For project justification purposes, Novell was looking for productivity savings from two major categories: engineering efficiency and incidence prevention. The outcome was much better than even Novell expected; they exceeded their original objectives and in addition, realized measurable savings from other unexpected categories including reduction in training time and an increase in the number of calls handled by less expensive Tier 1 customer service representatives. The new KB and associated training programs allowed Novell Tier 1 customer service representatives to answer 15 percent of incoming calls. This had the added benefit of enriching the jobs of the Tier 1 customer service representatives and reducing agent attrition. Additional benefits including shifting the calls to less expensive staff and speeding up resolution rates, thereby improving overall customer satisfaction (see Figure 4).
A major objective of the KM project was to decrease the average incident resolution time. In 1997, the average time to resolution was 26 days. By the end of the first quarter of fiscal 2001, Novell’s average time to resolution had decreased 46 percent to 14 days. This productivity savings was then applied to Novell’s fully loaded Full Time Equivalent (FTE) cost to realize savings of $1 million in fiscal 1999 and $3 million in fiscal 2000. The savings were significantly smaller in 1999 because only a portion of their support staff was using the new KM system.
Novell was even more conservative in its calculation of incidents prevention since it did not make the new KB available to customers on its Web until early 2001 and has not yet done any advertising or marketing of this new support feature. Novell estimated savings of $69,000 in 1999 and $795,000 in 2000. They expect these numbers to increase greatly beginning in 2001, as more customers become aware of their new advanced self-service KB.
Best Practices and Lessons Learned
Novell’s success is the envy of all companies investing in KM, but their success was not an accident. It was the direct result of the company’s flexibility, willingness to invest in new systems and openness to cultural change. Novell knew their customer service organization needed better tools in order to provide consistently outstanding customer service and decided the most immediate need was KM. They learned the hard way that systems alone were not enough. Novell succeeded, where so many other companies had failed, because they were willing to throw out everything they had known to be true and searched for new truths. The most important best practice and lesson to be learned from Novell’s KM success is that enterprises that are not willing or able to shift their operating philosophies and corporate culture should not invest in KM.
Novell’s second best practice goes against “standard” wisdom about KM. Novell did not force its staff, partners or customers to use the new KM system. Novell installed the Primus KM tool to assist their technical support staff in performing their daily jobs, not to replace them. Novell had invested a great deal in their support and engineering staffs and didn’t want to lose people because of the challenges in phasing in a new support system. Three years into the initiative, 15 percent of Novell’s support staff and engineers are still using the old Vantive tool for KM. Novell believes giving their staff a choice about which KB to use has helped them to keep their KB clean. While Novell is willing to maintain and support both the old and new KBs, all new technical support employees use the new KM tool.
In January 2001, Novell implemented SolutioNet, which gives external customers and partner’s direct access to the Primus KB (http://solutionet.novell.com/esupport/esupport.asp). Novell added the new Primus KB without removing access to the Vantive system. Again, Novell has been consistent in acting upon its operating philosophy of minimizing the impact of systems transitions on its customers, partners and employees.
Novell has avoided many of the common pitfalls that lead to the failure of customer relationship management (CRM) initiatives in general and KM efforts in particular. Prior to implementation, Novell defined specific and quantifiable objectives, established a baseline and measures to quantify the project’s benefits. Novell then divided their KM initiative into manageable and measurable phases, instead of trying to convert the entire support organization at one time. Each phase was given adequate time for implementation and review, prior to going ahead with the next one. Novell spent an entire year, from November 1998 to October 1999, creating new solutions and training its 10 U.S. support teams before beginning their European conversion in 2000.
Not only did Novell revamp its measurement, reward and promotion systems with input from all participating departments but also developed a specific KM training program and modified its existing training programs for all levels of support staff. Neglecting training is a common mistake that leads to the failure of many systems implementations. As KM is solely dependent on the input of its participants, poor training can kill the KB before the project can be rolled out.
Novell does not have a large KM staff to maintain its KB. Instead, this is the responsibility of the departments using the system, under the direction of a single Program Manager. The Program Manager is the only resource dedicated to the KM initiative on a full-time basis. Novell has four KM certification levels: Author, Editor, Publisher and Coach. (See Figure 5). Novell does not provide financial rewards for case creation. Instead, it motivates its staff to achieve promotions to higher KM certification levels, which include financial rewards.
“We’re proud of our KM achievements and hope that sharing our success will help others achieve similar results,” said Gig Griffith, Manager of Business Operations at Novell Technical Services. Gig emphasized that “Novell’s success was due to the willingness of the company’s management team and employees to embrace change and be flexible in accepting and maintaining existing systems for those uncomfortable with change.” Novell’s willingness to share their success and their ongoing commitment to sharing ideas and knowledge further demonstrates the innovativeness of their corporate culture.