Applying KM to customer interaction
Most of the efforts in knowledge management to date have targeted internal work force productivity, leveraging the past experience of a company's various workgroups to accelerate the work of new project teams. Those benefits are great-cost reduction, faster time to market, better quality-but to really get the CEO's juices flowing, KM needs to translate directly into incremental revenue. Although they aren't calling it KM, a new group of software vendors are beginning to deliver tools for creating E-commerce Web sites that extract and process knowledge about the customer, and use that knowledge to drive sales.
One of the best examples is Vignette Software, which now bills its Story Server suite as "Internet Relationship Management." At first glance, Story Server looks like most Web content management software, a tool to coordinate management of complex, constantly changing Web sites, with content submitted by many authors throughout the company. What's different is its focus on understanding each customer through his or her "life cycle," which begins with a prospect's first visit to the site, progresses with repeat visits and ultimately results in a customer with a well-understood profile.
A key element is personalization. Story Server includes intelligent software agents that tailor the content to the user's expressed and implicit interests and preferences. For example, a Matching Agent tracks the user's navigation through the site and uses that behavior to infer content of interest. Presentation of Web page content and banner ads is adjusted dynamically as the site begins to "understand" the user. A Recommendation Agent, based on GroupLens technology from Net Perceptions, suggests specific products that should be of interest to the user, based on the choices of other users fitting a similar profile. Anyone who has experienced this feature, for example on Amazon.com, can attest to its marketing value. Finally, users further along in their life cycle, and who register with the site, have the opportunity to personalize the site explicitly. For example, users can track their personal investment portfolios on a financial services site, or monitor air fares and flight times for specific destinations on a travel services site, etc.
A second vital element is user profiling. Information about a user's navigation history, preferences and behavior visiting the site is recorded in a database. Companies can define data elements, unique to their business, that are tracked by the software. This data can be analyzed by both Vignette-supplied and third-party decision support and reporting tools to personalize content delivery and build statistical profiles.
With this kind of software, the link from knowledge to sales is direct, and the effects of content personalization become apparent in hard numbers. It may be a stretch to call this customer relationship management, but the value of the site to the customer does increase with additional visits and purchases, as the system learns. Knowledge-based E-commerce is a win for both customers and suppliers, and will become increasingly a central feature of commercial Web content management software.
Even without the revenue incentive, such a learning-based approach to personalization and user profiling would also be useful in ordinary collaborative KM software. And EDM vendors who see Web content management as nothing more than a low-end bolt-on to their core offering probably need to think again.