With its promise of letting companies better leverage everything they "know," knowledge management is now firmly established as a concept worth pursuing. Using groupware, databases, and other software tools, a growing number of businesses are trying to combine organizational data with the tacit information in employees' heads to create an enterprise repository of intellectual capital. It's an ambitious undertaking, and one that few companies have mastered.
In the pursuit of knowledge management, much of the work--and most of the payback--lies ahead. According to a new survey of 200 IT managers by InformationWeek Research, 94% of companies consider knowledge management strategic to their business or IT processes. But many of those companies apparently are in the early stages of their knowledge-management efforts. On average, companies capture only about 45% of their intellectual capital, according to the survey. Also, only 36% of companies have formal policies for sharing knowledge assets--and even fewer have formal policies for capturing such assets.
A variety of obstacles are preventing knowledge management from becoming ingrained in day-to-day business practices. Some of the biggest impediments involve people, not technology. Two-thirds of those surveyed say behavior modification on the part of employees is the biggest challenge moving forward (see sidebar story, "Mastering The Human Factor"). "Knowledge exists in people and in the interaction between people, work, and the problem to be solved," says Tom Brailsford, manager of knowledge leadership at Hallmark Cards Inc. in Kansas City, Mo. The challenge is to systematically capture information and share it across internal boundaries, he says.
Other barriers identified by the survey include getting buy-in from corporate management and the difficulty of classifying knowledge (see chart, below).
Still, businesses are forging ahead. This year, 62% of organizations will spend more on knowledge management than they did last year.
Companies as diverse as Hallmark; Platinum Technology; Schneider Automation; Sears, Roebuck; and Shell Oil are pursuing knowledge-management strategies as a way of going beyond conventional business objectives such as increasing revenue and cutting costs. Their goals are to apply the collective intelligence of their employees--and the institutional memory of past experience--toward innovation and gaining a unique position in current and future markets.
"Knowledge is what we're all about," says Susan O'Neill, deputy chief knowledge officer at consulting firm PricewaterhouseCoopers. "All of our profitability and viability is about how good we are at leveraging the intellectual assets of our people and making that available to our clients."
The Infrastructure Many technologies are being used to capture corporate knowledge. Because knowledge management has the potential to touch every employee and computer system within a business, companies are using a wide assortment of technologies to create their knowledge-management infrastructures. According to the InformationWeek Research survey, relational databases, text and document search engines, groupware, data warehouses, and data mining tools are the most popular components of a knowledge-management environment--in that order (see chart, below).
Sears, which is in the early stages of its knowledge-management initiative, is focusing its initial effort on data warehouses. The retail giant has more than 3 terabytes of customer data stored in 37 databases. That collective intelligence represents demographic data on 97 million households. An IT team at Sears is working with business staff to create an enterprise data model that, once in place, will let people in different divisions more easily share and reuse all of Sears' data. "One of our big issues is getting the data consolidated into a format that we can access in real time," says Marlene Gaidzik, a senior systems manager at Sears, in Hoffman Estates, Ill.
Sears' data model will eventually lead to marketing and sales efforts that will involve employees not traditionally involved in those areas. "Let's say a customer requests service for a dishwasher," says Gaidzik. "The repairman shows up and sees that the customer has a 25-year-old refrigerator. If the repairman could get that information to the right people at Sears, the next time the customer was billed, the bill would include a discount coupon for a refrigerator."
Sears' knowledge-management efforts will eventually extend beyond its data warehouse environment to include reporting systems, data-analysis tools, customer-relationship management and enterprise resource planning applications, and an intranet. "We need to share knowledge so that we are not only getting information out to the business divisions, but so they are also bringing information back in," says Gaidzik.
Like Sears, other businesses are discovering that if they are to achieve some of the more ambitious goals of a full-fledged knowledge-management strategy, they have to start with the basic IT plumbing. Scient Corp, a 1-year-old consulting firm in San Francisco, has also built a unified data model that helps it maintain consistent data across different applications and eliminate silos of information.
Creating the data model "was the most important thing we did," says DougKalish, Scient's chief knowledge officer. Kalish and his team defined a set of common objects that represent employees, clients, and units of work, among other metrics. The model provides a framework for representing data that resides in Scient's knowledge-management system, which runs on an intranet, using Microsoft's Internet Information Server and Site Server.
Scient has developed hooks from Site Server into its PeopleSoft human-resources and financial applications and into time and accounting applications based on Excel and Microsoft SQL Server. "PeopleSoft has its own data store and we can't change that," says Kalish, "but now we can read and write that data store and index the PeopleSoft data with our file system data and intranet data in a single repository. We can break down application boundaries, and that's essential."
The object framework also lets Scient append data that originates in other systems. "When a document is submitted to our knowledge-management system, we tag it with an employee object," Kalish says. "We can aggregate that with other information in the system and find out what other things that employee has contributed. We can find other characteristics, such as the clients they've worked with or what department or office they're in."
Scient is now working to add project-management functions to its knowledge-management system, so that its employees can learn from lessons discovered in completed projects as well as input new information on current projects. For instance, if an employee were assigned to do an evaluation of various Internet technologies, the system could tell that person how long the evaluation might take, based on previous experience. The system could also identify team members who worked on the earlier project in case the person wanted to contact those with previous experience in that subject area.
Data Framework Defining objects that tag or identify information about data--a twist on the metadata concept--is becoming increasingly popular in companies trying to leverage both the structured data in relational databases and unstructured data found on the Internet and in other sources. And it's where the up-and-coming Extensible Markup Language fits into the knowledge-management discussion.
XML is a set of rules for defining data structures. To date, nearly every software vendor has pledged support for XML, and a growing number of IT shops are using it to integrate disparate sources.
Platinum Technologies is using XML to build an index for a knowledge-management repository that holds documents on all the software vendor's products. Each piece of content gets an XML tag that identifies the source of the content. Other bits of information in the tag can identify who the product is sold to, which business unit within Platinum is responsible for the product, and which operating systems the product runs on.
The index is built using these XML tags. Scheduled scans of the repository look for any changes that have taken place in the documents that are managed there. If changes are found, a new index is created, says Glenn Shimkus, Platinum's director of worldwide knowledge management. The XML-based index can be easily searched and contains the most up-to-date information on Platinum's products, Shimkus says.
Platinum's foray into knowledge management began in the summer of 1997, chiefly because it had rapidly expanded through the acquisition of 74 companies in four years, boosting its product portfolio from about 40 products to 185. (Computer Associates last week agreed to acquire Platinum for $3.5 billion.)
The initial goal of Platinum's knowledge-management project was to give its global sales force of 1,500 people one place to go for information. Salespeople also wanted assurance the data was up-to-date.
The first system to come out of the project was Jaguar, a Web knowledge portal for sales and marketing that includes the XML-based index. Platinum has since built six portals around knowledge communities, or groups of individuals with common interests such as product development, sales and marketing, and business development.
The payoff, thus far, has been notable. Last year, Platinum saw a $6 million return on an investment of $750,000, says Shimkus. Much of the cost efficiencies came from the elimination of printing and mailing expenses.
More difficult to measure is what Shimkus refers to as "soft ROI." For example, using Jaguar, Platinum has increased its revenue per salesperson, or revenue productivity, by an estimated 4% year over year. "While there's no real way to measure that, we've based it on the idea that Jaguar saves a salesperson two hours per week in searching for information, and the average employee saves one hour per month," says Shimkus.
The success is spurring more knowledge-management projects. Platinum is now rolling out a knowledge portal for its consulting and education group so it can share best practices.
Intranets are also being used for knowledge management because they support easy access via Web browsers. On average, 52% of the InformationWeek Research survey respondents give their employees online access to their companies' knowledge assets.
Internet technologies give Schneider Automation, a manufacturer of industrial automation-control equipment, a platform for easy information access. Schneider built an extranet on top of Lotus Notes/Domino that is accessed via Web browser by the company's own employees and by the sales force it shares with its parent company, Schneider Electric in Paris.
Schneider Automation is in the process of pulling into a repository all the information needed for business development and technical support, says John McElfresh, director of electronic business with the North Andover, Mass., company. "It has strong workflow capabilities to guarantee that the information never goes out of date and is accurate," says McElfresh. Each piece of information, whether it's a PowerPoint presentation or technical document, has an expiration date and a responsible party that "owns" the data.
Schneider Automation has also built links between its SAP R/3 applications and Domino. Operational data such as product availability is fed into the knowledge-management system in batch transfers. Now the company is turning those links into data feeds on the extranet. "If someone wants to know the relative availability of a product, they can find that online and it's up-to-date," McElfresh says.
The company has also integrated help-desk software from Vantive Corp. with the Domino-based repository to manage technical support and marketing information. If an employee can't find the information he or she needs from the repository, the help-desk software lets the employee create a trouble ticket that is passed to an internal department called Customer Central. There, personnel are chartered with getting an answer to any query within two days--although the process often is completed within four hours, says McElfresh. Once the answer is found, it's put into the knowledge base.
Hallmark is about to launch what it calls a "knowledge-creation community" on a Web site that will be accessible to about 100 of the card company's retail stores. The site, based on a modified version of Pensare Inc.'s Knowledge Community software, will be a place where the retailers can communicate with Hallmark, exchange E-mail with each other, chat, share ideas on a bulletin board, and access a database of best practices.
Hallmark expects the site to be a fertile ground for information sharing among the retailers, as well as serving a supporting role in its own product development. "We know lots of our retailers have tried new things to increase sales, but there hasn't been a way to share those practices," says Brailsford.
First Steps Many other companies have their own aspirations for knowledge management, but, like Hallmark, Schneider, and Sears, they are in the early stages of entering this new world of information sharing.
A growing number of technology options promise to make the effort more manageable, but even vendors agree that not all of the pieces are in place. "This topic isn't fully baked," admits Scott Smith, managing principal of Global Knowledge Management Consulting Services at IBM Global Services. "We still have a lot of things to figure out."
Businesses still rely largely on their own ability to integrate different types of software together, Smith says. For instance, they may start with a collaborative engine such as Lotus Notes and then integrate text-mining tools, interactive meeting software, and portals to integrate external data. IBM and Lotus have aggressive plans to change this building-block approach to knowledge management. "Within the next 12 months, you will see these components share the same underlying technology so that companies can share information better," Smith says.
Those kinds of promises ring hollow among skeptics who feel the whole concept has been oversold. "Knowledge management is to the late 1990s what reengineering was to the early 1990s," says Ron Shevlin, an analyst with Forrester Research. "It's a management fad perpetrated by consultants and software vendors."
That's an extreme view, but its certainly true that most businesses have a long way to go in realizing the full benefits of a comprehensive knowledge-management strategy. Many companies lack formal policies for capturing and sharing tacit information, and the overlap of technical and cultural issues makes knowledge management a complex undertaking. "It isn't rocket science," says Platinum Technology's Shimkus, "but it is difficult."
Actually, an effective knowledge-management environment would theoretically help engineers build better rockets, faster. In that respect, the two concepts might be equally ambitious--and difficult.
Copyright 2009-2016, Douglas Kalish. All rights reserved.