Local InsightEBizinsights tries to take the clickstream bull by the hornsBy Jack Hakim and Tom Spitzer Continued from Page 1 The other operations that you perform within the Visual Portal are to select specific data points by clicking on the bars that represent them, and either include only the selected values or exclude the selected values from the current analysis. When you're happy with the configuration of the selected data set, you can use a palette of tools that allow you to optimize the display of the overview chart. These tools let you pan, rotate, turn, and magnify the chart.
Playing with these features, it becomes obvious that you're working with quite a powerful data-driven graphics engine; it provides smooth, responsive performance with large amounts of data. Perhaps more useful, however, is the ability to copy an Insight out to an Excel worksheet, where you can add externally sourced cost and revenue data. The Visual Path Analysis application provides a collection of views of visitor paths to and through a Web site. (See Figure 2.) The application offers a selection from three predefined path displays and five predefined flow displays. The path displays depict a visitor's tour of the Web site as a series of rectangular boxes, each representing a page with the time on each page represented by the size of the box. Each of the flow analysis displays resembles a context map, with the graphic element (a circle or square) that depicts each data point sized to communicate the percentage of the data set that the particular data point represents. Our favorites were the Popular Referrals chart, which displays traffic from the leading sites that directly generated visitor traffic, and the Page Flow, which allows the user to select a page within the site and depicts the sources and destinations of visitors passing through that page. Overall, we found the Visual Path Analysis application to be well suited to understanding visitor behavior on a Web site. Our experience with eBizinsights XL underscored the challenges facing companies attempting to integrate Web-business efforts into a coherent marketing, merchandising, and delivery system. To optimize their communications with their markets, marketers need to make rational decisions about allocating marketing resources between internal and external Web initiatives and optimize those activities in the context of more traditional campaigns that may include trade shows, PR, print, and broadcast messaging, among other vehicles. At the other end of the spectrum lies the vexing challenge of assigning value to a wide range of outcomes that may include customers buying direct on the Web, customers buying your product from a Web retailer, and customers using the Web to research products and make purchasing decisions before consummating their purchases at a store or dealer. Analyzing the behavior of visitors to your Web site is only part of the story. Unfortunately, products such as eBizinsights XL that point a very powerful analytical gun at that limited part of the story are bound to raise as many questions as they answer. In our review, we immediately started asking questions such as how well did the marketing campaigns that yielded these customer outcomes correlate to cost? What was the value of the goods purchased by customers visiting as a result of a particular campaign? Are there differences in the value of relevant demographic characteristics of visitors coming from different domains? EBizinsights XL can help answer these questions but can't answer them on its own, and it doesn't let you add and correlate your own significant factors without significant cost and development effort. However, if your primary interest is to understand visitor behavior within your site, the following product attributes make this a fairly effective end-user tool:
Jack Hakim [jhakim@ecwise.com] and Tom Spitzer [tspitzer@ecwise.com] are principals at EC Wise Inc. With offices in the San Francisco Bay Area and Seattle, the firm specializes in creating user interfaces for analytical and other information systems that involve large, complex databases.
|
 |
Most Popular This Week
IE Weekly Newsletter
Subscribe to the newsletter
| |
| |||||||||||||||||||||||||||||||
