 |
WebGroove is no longer supported. These docs are reference only. |
|
|
Overview 
Collecting User InformationThe personalization process begins as soon as a viewer enters a given web site. By cooperating with web servers and various specialized modules, our collector logs the activities of the online user. Activities such as page-views, queries, product purchases, banner ad click-throughs and participation in on-line communities are monitored and stored in a database as the user navigates the site. Building User Profiles A comprehensive user profile is created by analyzing user activities. To compile behavior patterns, both the action and the content in question are considered. Each action is assigned a numeric weight to determine how it effects the viewer interest scores. For example, reading about a recent baseball game would indicate an interest in sports, specifically baseball. However, purchasing a baseball glove or clicking on a baseball-related banner ad would indicate an even stronger interest and a higher score. In addition, viewer demographic data can be gathered through a non-intrusive survey. Only the most basic questions should be asked, such as age, gender, occupation, and zip code. More detailed information is inferred based on this simple foundation. For instance, income level can be easily determined based upon occupation and geographic location. This process is highly accurate compared to actual survey responses. If a viewer feels a survey question is too personal, he is likely to leave the site or give misleading information. Default behavioral profiles can also be defined for people who fall within various demographic areas. These standard profiles can be used to "seed" a behavior profile with information inferred from a basic demographic survey. All of this information is analyzed and compressed based on a variable set of parameters to create a compact viewer profile. Each profile represents an individual's personal interests as well as demographic and inferred psychographic information. These profiles are then used to make quick decisions about appropriate content, advertisements, and suggestions. Building Content and Ad Profiles Web site content, products, and banner ads must be segmented and profiled. This information is used to adjust user behavioral profiles as well as to deliver personalized targeted content. In many cases an online marketer or advertiser will specify targeting criteria for establishing a product or ad profile. The profile administrator allows approved people to directly manipulate profile information through a web-based front end. Content such as news, articles, and other areas of interest can be categorized and rated manually or through the use of automated text analysis. Our content analyzer acts as an automated editor, scanning your web pages, news feeds, and databases in order to make intelligent choices for categorization. Both the content itself and the way it is organized are taken into account. The rules and relations used by the analyzer can also be modified to improve accuracy and specialize for various business or content areas. Profile Analysis and Advanced Reporting The heart of the system is the profile analyzer. Through adaptive personalization, user profiles are compared with content, product and advertisement profiles. Defined targeting parameters are fine-tuned to behavioral patterns and viewer characteristics, generating an evolving predictive model. Based on these results, the system is able to quickly deliver real-time personalized choices for each individual. A wealth of valuable information can be extracted from the data held within this system. Our reporting tools empower administrators to query and analyze the databases to generate real-time reports, graphs, and charts. For instance, a report could compare the targeted audience profile for an advertisement with the average profile of the users who actually clicked on the banner. Through information of this type, marketing campaigns can be finely tuned to be much more effective. Targeted Content & Ad Delivery Relate to Advertising Traditional banner advertising takes a very broad "shotgun" approach. Throw enough ads online, and some of them are bound to reach interested people. WebGroove® Relate allows marketers to show ads only to those people who are likely to respond based on individual demographics and behavioral profiles. This can vastly reduce the number of wasted impressions and result in a much higher response rate. To focus even further, the profiles of individuals who actually responded to an advertisement can be examined to find commonalties. In this way, a true 1-to-1 marketing relationship is forged, allowing the advertiser to know his customers better than ever before. Relate to Commerce Imagine walking into a store, scanning the aisles, and finding one with a sign bearing your name. As you walk down that aisle, you see all of your favorite products lining the shelves. You also notice cool new products from your favorite companies along with other really interesting items you've never seen before. Would you keep shopping at that store? All of this is possible on the Internet by using WebGroove® Relate. Individual purchase history, demographics, and behavioral profiles are all considered and put to good use. Intelligent product recommendations and observed user preferences make the shopping experience much more personal and fulfilling for each visitor. Relate to Content Readers like a wide variety of online information, but finding what is really of interest can be a daunting task. A search engine can get closer to the mark, but it doesn't really know the reader or what they might like. A friend who knows you can recommend a magazine article and be pretty darned sure you're going to like it. These knowing recommendations are what WebGroove® Relate can provide. This system can be applied to many types of content, such as news feeds, web site directories, or recipe archives. Plus, the content that individuals access is used to further define their behavioral profiles and improve future recommendations. Relate to Community A feeling of community is a powerful way to build a loyal following for a web site and keep visitors coming back time and time again. Online chat, message boards, and opinion polls are just some of the interactive features that foster this type of devotion. Activities that engage a user can also tell us a lot about her interests and what type of person she is. This information is a valuable contribution toward building individual behavioral profiles. Knowledge of this type can be used for automatic user conveniences, such as prominently displaying chat rooms and messages boards that should be of interest. It can also be leveraged to make other areas of a web site much more personalized, effective, and profitable. For More Information Contact Floyd Bargy floyd@creativision.com Creativision Publishing Corporation 800.280.7747 or 602.785.7494 See Also  Online Demonstration
 Privacy Statement
Relationship Marketing
System Requirements
Value from Click-Streams
 White Paper
Up To WebGroove Relate
WebGroove® is a registered trademark of Creativision Publishing Corporation |
 |
 |
 |