While the hype may make "personalization" seem like a neatly organized subject, in truth the term and the practice vary considerably from situation to situation. The personalization practiced at Yahoo!, which focuses on merging stated preferences and browsing behavior in order to deliver requested content and targeted advertising differs considerably from the personalization employed at an online retail site, or in a corporate call center.
To make sense of the mess, Gartner Group defines five distinct categories of personalization:
Content personalization, usually a Web endeavor, fits the Yahoo! mold, and concerns itself with tailoring information (typically factual) either through expressed or implied user preferences.
Offer personalization is concerned with matching the right customer to the right product or deal, and also determines the optimal approach (both in terms of discount or promotion and channel of contact) to close the sale.
Product personalization is essentially synonymous with interactive product configuration-it allows customers to tailor components of their purchase to best suit their needs.
Price personalization is of prime importance for many B2B companies who want to display accurate, contractually linked pricing and availability information based on the complete known relationship between the buyer and seller. It is also of interest to consumer-facing firms who want to explore dynamic pricing approaches.
Service personalization focuses on delivering the customer to what the firm considers the optimal communication channel for support requests, then ensuring that the live or automated agent has a complete understanding of the business relationship before proceeding. This also belongs with the concept of establishing "platinum" customers and automatically routing them to appropriately deluxe customer support.
Because of the considerable expectations laid at the doorstep of offer personalization to boost revenues, profits, and repeat visits, many companies are greatly concerned with the problem of personalizing to the first-time Web visitor. In an attempt to personalize without asking questions, they often expend a great deal of time and energy trying to obtain third-party data on otherwise anonymous customers, or build complex models of the likely interests of fresh blood.
In a May 2001 Gartner Group research note, research director Walter Janowski challenges that strategy by reminding businesses that a large number of unidentified customers are clearly "passersby" who are likely to be unmoved by whatever guess the personalization engine returns. "Efforts expended to address the needs of these customers are likely to be wasted," he writes.
He goes on to suggest that targeting personalization efforts at unknown quantities also tends to alienate existing customers who might benefit more from a personalization strategy more heavily concerned with their needs. "Many enterprises target [passersby] by overengineering their Web sites to offer every possible piece of information a casual visitor might want. The end result is usually a confusing, burdensome design that serves instead to frustrate and drive away the more-likely customers."