CBR2 Success Story: Intergraph Corporation

Leading Computer Graphics Vendor Sees CBR2^TM as Wise Choice for Project Solomon

How does the company ranked number one for customer satisfaction in the CAD-graphics industry maintain its enviable reputation for quality service while experiencing explosive product growth? In the case of Intergraph, the world's leading supplier of computer graphics solutions, it will take the wisdom of Solomon, an aggressive project to upgrade its customer support services by integrating the inefficient hodgepodge of tools they are currently using with the unique CBR2 automated knowledge base products from Inference.

By capturing "perishable knowledge" in a single problem-solution repository and giving all its customer support analysts access to it, Intergraph intends to speed up first help response, increase the first call closure rate, lower support costs and even improve employee retention. But most importantly, Intergraph sees the case base retrieval technology provided by Inference as the key to maintaining its quality of service and leadership in customer satisfaction.

Founded in 1969, Intergraph today is a Fortune 500 company with over 10,000 employees in 50 countries worldwide and annual revenues for 1994 of approximately $1.2 billion. A significant portion of the company's revenue is generated by its global support organization, which in recent years has been rated number one in customer satisfaction and service by every formal customer satisfaction survey conducted in the CAD-graphics industry.

The company's existing three-tier customer support system relied on highly experienced help desk personnel supported by an inconsistent mixture of call tracking, problem tracking and information searching tools. Intergraph has been able to resolve a very large percentage of calls at the help desk level by staffing it with engineers and architects with four to five years experience in their product areas. Most of the company's support groups use internal Worldwide Web servers to collect and share information among the various product groups, while a customized version of WAIS is used by some groups to search archives of known problem-solution pairs.

But customer support managers at Intergraph could foresee that maintaining their quality of service would be difficult in the face of the 80 to 100 percent annual product growth rate the company was experiencing. For one thing, the existing system was highly compartmentalized, comprising 20 to 30 agent groups with specific areas of product expertise. When customers called in, they could be routed to any one of these groups depending on their specific problems. If a problem wasn't correctly identified initially, however, a customer could be bounced two or three times between various groups before the call got resolved, adversely impacting customer satisfaction.

With such tremendous product growth, training was becoming more of an issue, and it was becoming more difficult to give help desk personnel access to the information they needed. Additionally, in light of a projected service call growth rate of 40 to 50 percent using such highly skilled employees to staff the help desks was no longer going to be practical.

"We realized these experienced people were essentially coasting on the knowledge they had gained while working in the product areas, and that unless we provided a way to pump product information back into the help desk level we would quickly drain those resources." explained Paul Brownell, executive director of customer support. "We could take lots of experienced people and keep cycling them between product areas and support, a very expensive solution, or we could put in place a knowledge base tool to allow us to use generalists at the help desks and move our technical experts away from the phones."

To address the needs for their next-generation customer support system, Intergraph embarked upon project Solomon, a plan for automating knowledge engineering. Intergraph's vision was to develop a corporate system architecture with a problem-solution repository at its core for capturing and storing "perishable knowledge," and integrating it with their existing customer service systems. Intergraph saw project Solomon as a way to improve customer satisfaction and remain competitive by improving the efficiency and productivity of their support staff through better access to information, reducing their training costs and dependency on front-line technical experts and offering their customers proactive support tools.

"We see more support automation tools becoming important to the viability of our support business," Brownell said. "Each analyst should be able to support a wide range of products with less training, and provide consistent solutions to problems that have to be solved only once."

According to Brownell, with project Solomon in place most of the information help desk personnel would need to answer most of their calls would be at their fingertips without requiring years of experience. Intergraph began a search for a vendor by first establishing their selection criteria, including cost, ease of use, the ease of capturing existing problem-solution pairs and integrating the knowledge base with the current system. Intergraph also wanted a knowledge base that could evolve gracefully, capturing unresolved problems without having to be redesigned.

After investigating several competing vendors, Intergraph chose Inference to provide project Solomon's core knowledge base technology and tools and integrate them with the existing customer service system. They selected Inference's CasePoint 2.0 for problem resolution, CBR Express 2.0 for knowledge authoring and the optional CBR Express Generator for access to existing documents.

Their selection was influenced to a great extent by Inference's leadership position in the marketplace.

"We were attracted by Inference's strength in the marketplace. When we looked at who the help desk vendors were partnering with, we discovered that nearly everyone had already established strategic partnerships with Inference," said Brownell. "This spoke very strongly for the strength and longevity of their solution and their company."

Brownell also like the security of using a technology that although advanced is proven and mature.

"Inference was the clear front-runner when we compared the cost and technical strength of the product," said Brownell. "While some vendors claimed they had more advanced technology, once we dug into them we found they were more 'bleeding-edge' and potentially dangerous for a production environment."

With the help of Inference staff, Brownell and other Intergraph managers created a return on investment model to justify the system by estimating the productivity and training benefits they could gain using the CBR2 tools. Intergraph then conducted a small pilot project to test out its numbers as well as some of the new technologies such as bulk loading of existing problem-solution pairs.

Intergraph chose two product areas to test that, although quite different from each other, together represented Intergraph's overall support business. The company spent one month developing case bases for each product area after some author training, and established a minimal level of integration to allow CasePoint to automatically capture problem descriptions from the existing trouble ticket software. This way, support analysts would be presented with potential solutions and additional questions from CasePoint on their screens, along with the customer information, even before they speak to the customer.

After running the pilot test for one month, Intergraph noticed significant improvements in productivity even with the very limited case bases they were using. Not only were novice analysts able to solve problems quickly, but productivity increased by 20 percent for one product area while maintaining the first call closure rate.

"The numbers were much better than what we had hoped for," said Brownell. "We've been extremely pleased with the support we've received from Inference.

Now Inference is helping Intergraph to implement project Solomon. Inference's consulting staff is helping to direct the development process, providing invaluable assistance in establishing operating procedures and standards for developing case bases, as well as setting goals. Inference is training case base authors, enabling them to participate in shaping the direction of the Solomon system to make it more responsive to their needs. And as the more than 20 case bases for various product centers are being developed, Inference is helping to maintain quality by ensuring that Intergraph authors are staying on the right track.

After each case base goes live and gets more refined through use, Intergraph intends to distribute them to its regional offices for local use. After further review and refinement, Inference's CasePoint WebServer will enable Intergraph to put its case bases on-line so its customers can access them directly through the Internet's Worldwide Web, mitigating the support staff's workload even more. Eventually, Intergraph will create its case bases as part of the normal product development cycle, and may even embed the case bases into the products themselves.

"We expect the Solomon project will be a dramatic change for us, affecting not only the performance of our groups, but their organization and structure as well," said Brownell. "Ideally, if we do our jobs here perfectly, customers will buy our support services, we'll send them the knowledge bases, and they'd never call us."

# # #