Personalization: Putting Data to Work

For the past several weeks, I have been providing an overview of tools that can be used not only to help marketers track and analyze activity on web sites but also to evaluate the success of e-business initiatives. Focused on effectively gathering, managing, and analyzing data about customer activity, this series should provide a foundation for building a web-centric business intelligence strategy.

Having established a framework for marketers to assess their specific needs and determine which tools will provide the best solutions, I thought it time to turn to one of the most important real-world applications of data: personalization.

Although personalized relationships are certainly not new to the B2B world, relying on a web site or software (as opposed to a sales team) to establish and drive personalized relationships is a new challenge. And it’s a challenge that will become increasingly acute as the Internet becomes a standard platform for B2B commerce.

I think of personalization in the following three major categories, though shades of gray will always exist among them:

• Customization
• Simple personalization
• Rules-based personalization

Customization. Perhaps the most common and popular form of personalization, customization is manifested by providing customers the opportunity to design web environments (extranets) with information pertinent to an individual user.

Easiest to think of as “many to many” personalization, customization allows organizations to use general information to drive apparently personalized relationships with customers. Think of how beneficial a customer would find having current pricing information on only the product he or she is interested in.

Simple personalization. In this approach, an organization draws information from its customer database and delivers it to a unique customer. Purchase history, billing information, and shipping preferences are all information that can be stored and provided to users to modify and manage. Although the coordination of data from multiple departments may be challenging, this simple approach is utilitarian and focused.

I often also think of Amazon.com’s “collaborative filtering” approach to product recommendations as simple personalization, though the decision process may be complex given the number of attributes, because the same recommendations are being made to multiple customers.

Rules-based personalization. I believe the most exciting approach to personalization is one that leverages all the data we have analyzed about customer relationships and uses that information to create unique relationships with individual customers. I cringe at using the “one to one” term, but it is an appropriate (if overused) expression to describe the goal of the relationship.

Rules-based personalization uses web tracking, online analytical processing (OLAP), and data mining solutions either to predict the needs/behavior of unique customers and provide the opportunity for interaction or to provide unique service and support based on the relationship history. Though in its infancy, rules-based personalization will come to dominate the B2B landscape as companies move closer to designing relationships based on enterprise profitability models.

Some Solutions

Implementing customization and simple personalization can be accomplished through a variety of custom solutions and are often intermingled. By leveraging existing data in customer databases, organizations can develop extranets to provide unique customer environments. Obviously, the challenge grows as the customer base increases, but there are specific business processes, such as cataloging, purchasing, and service, that can be designed for multiple customers and still appear to satisfy the needs of each one individually.

The more intriguing challenge is rules-based personalization. I will delve further into these solutions and their applicability to B2B marketing in my next article.

All Things Considered

Personalization is certainly not the only end result of tracking and analysis. Information can, and arguably should, be applied across the entire enterprise to assess and drive profitable customer and supplier relationships.

Personalization is a topic I previously addressed from a different vantage point. In that article, I was concerned with privacy issues. While a significant component of the personalization conversation, those concerns have not been re-addressed here. I encourage you to read that article, though, and tread cautiously as you choose how you will use information.