In today's world, most consumers have begun to use the Internet to purchase products or have at least started looking up information on the Web. Many times, a customer will activate a search engine such as Yahoo or Lycos and look for a particular site that interests them. Other times, individuals will go directly to a site that they know. In both cases, a customer's interaction with the Internet is based on their personal likes, dislikes and interests. These interactions provide valuable input into understanding a customer's behavior and what they are most likely to buy or what sites they are most likely to visit. If a customer's interaction with the Internet was captured and fed into a data warehouse, the warehouse could become a very powerful mechanism to analyze customers and to produce personalized Web pages for each customer. Personalized Web pages are a very powerful concept. Imagine going to the ESPN SportsZone site and having the scores and news for your favorite teams automatically listed. I know that in the spring and summer I check out the losing Mets, and in the winter, the losing Nets. As you can see, I have an affinity for losing sports teams. However, each time I go to the ESPN site, I have to enter parameters for the teams and news that I need to look up. If my interactions over the last two years had been captured, ESPN could have set up a personalized site for me based upon my interests. This concept holds true for everything on the Web from search engines (where an individual's common searches can be listed) to travel (where common destinations can be identified). For example, I have traveled from my home in New Jersey to a client in San Antonio on Continental Airlines every single week for the past eight months. Yet each week, I enter the Continental site and specify where I want to go and other redundant parameters. If Continental had captured my on-line purchases effectively in their data warehouse and prepared a personalized Web page based on my past purchases, I could click one button, enter the new flight dates and get rapid pricing information. Additionally, Continental might consider offering me bulk discounts based on my desire to book a number of flights at the same time or other promotions to destinations in Texas. The possibilities for additional revenue generation are endless. Amazon.com is one site that has utilized personalization quite effectively. By studying my book purchase patterns, the amazon.com Web site always tells me about the newest data warehousing books when I enter the site. To customers, ease of use is the number one priority on the Web. This is why AOL is very successful. A customer's interaction with AOL is made easier by the simple point-and-click interface. If a customer enters a Web site that is personalized for them based on past interactions with that site or their preferences, they will be much more likely to visit that site again.
Products from vendors such as BroadVision have made it much easier to personalize Web content. Through intelligent matching agents, BroadVision is able to use interaction and preference information to customize a Web site for a particular user. A marketing manager for a Web site can specify rules for personalization, and BroadVision will create that specific Web site based on the rules and information about the user that is stored in particular databases. This, in itself, is a very powerful medium to target specifically to users. However, if user interaction information was fed into a data warehouse, combined with other user information from other systems and profiled using data mining tools, the results could be even more powerful. Additional patterns would be derived and could be used as the basis of generating additional matching and personalization rules. BroadVision provides an open architecture for supporting Web personalization. Due to its ability to support information from several different types of databases such as Oracle, Informix and SQL Server, it could potentially be extended to support analysis from data mining models or data warehouse-based profiles. With the proliferation of Web personalization tools and the increased capture of customer contact information, we will certainly see the warehouse play a much larger role in the personalization of messages to customers.
Microsoft sits in a very unique position to drive the market for Web personalization and data warehousing. While most vendors offer individual components for data warehousing or Web personalization, Microsoft offers a suite of tools that has the potential to do it all. With the release of SQL Server 7 and OLAP services, Microsoft has provided the analytical capabilities for an integrated e-commerce strategy. Architecturally, a user could enter a Web site powered by Microsoft Site Server and interact with the Web site in a number of ways. These interactions would be automatically captured as part of SQL Server 7's integration with Site Server. Immediately, that information could be fed into another SQL 7 data warehouse and an OLAP cube that is poised to analyze the customer's interaction information. Intelligent agents would need to be designed to analyze this information and would provide profiles regarding customers back to Site Server, which could then personalize the Web site. The idea behind this concept is that throughout the information gathering, analysis and presentation process, one never leaves the Microsoft framework. Integrating the components becomes relatively easy because Microsoft has already designed the SQL Server 7 capture mechanisms for Site Server. It has already provided data migration and transformation tools to a data warehouse on SQL Server 7 and OLAP cube using its Data Transformation Services. The analysis and scheduling components are inherent, as part of the data warehousing functionality, and content can be personalized using Site Server. With other solutions, organizations may have to spend a great deal of time integrating tools with different specifications and setting up complex processes to move data. Microsoft is not quite there, yet. They still need to provide mechanisms for intelligent matching and a stronger Web personalization engine as part of Site Server, but the potential for having an integrated framework for intelligent Web personalization using data warehouses is a powerful concept that Microsoft has the potential for controlling.
The market for Web personalization using data warehouses will continue to grow. Several years ago, data warehouses served mainly as strategic decision-making tools that were used to analyze a corporation's profits and sales. Today, data warehouses are both tactical and strategic as they drive a corporation's efforts to become closer to its customers. As the use of the Internet proliferates and users become more and more drawn to sites that they feel at ease with, the personalization of Web sites to individuals will become even more important. Corporations which seek the edge in bringing customers to their Web sites and keeping those customers happy should invest a great deal of time in setting up a comprehensive strategy for one-to-one personalization of Web sites using data warehousing technology.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access