 |
Clickstream Data Warehousing Companion website for the book by Mark Sweiger, Mark Madsen, Jimmy Langston, and Howard Lombard Published by John Wiley & Sons, January 2002 |
 |
 |
| Discussion Forum |
|
Approximately once every few weeks, the book authors will answer selected questions submitted by our site readership. These questions could be related to questions about book content, clickstream data warehousing at your particular site, or data warehousing in general. Hit the submit link below to send a question to us: Submit Latest Question: Q: I already use a log file analysis tool like WebTrends. Why would I need a clickstream data warehouse? Cindy H., Consultant, Michigan Mark Sweiger: Log file analysis tools are good candidates for collecting one category of web site statistics—site traffic. Examples of site traffic include site hits per hour, day, week, month, etc., referring URLs, referring host sites identification, page load failures and HTTP status codes. Most well run sites use one of these tools. For example, I use the Analog log file analysis tool to collect traffic statistics on this website. While site traffic statistics are certainly useful, they do not cover the second, more important category of web site statistics—user behavior. Understanding user behavior is the key to maximizing the business value of your website. For example, suppose you ran an online retail site. Suppose that site traffic statistics from your log file analysis tool told you that site hits are increasing by 25% each month. If traffic statistics from log file analysis tools were all you had for business intelligence information, you might be ready to uncork a bottle of champagne, and tell your underwriters to take you public. But suppose that clickstream analysis of user behavior showed that shopping cart abandonments increased at a rate of 69% each month, meaning that the number of successful shopping trips are actually going down over time, despite the 25% increase in site traffic month-to-month. Furthermore, suppose user behavior analysis showed that the “shipping cost page” was the site of 81% of the abandonments. And suppose that analysis of the user abandonment behavior showed that shopping baskets of over $100 were abandoned at a rate 53% less than those under $100. This scenario actually happened at one online retailer, and the answer was that the high fixed-rate shipping cost was making smaller purchases uneconomic compared to the competition. Once the cost of a basket of goods got high enough, the fixed shipping cost was no longer the primary factor in shopping cart abandonment, and more baskets went to transaction completion. To solve this problem, the e-tailer needed to tier its shipping costs to be related to the value of the goods being sold, and after it did so, the shopping cart abandonment picture improved significantly. The above shopping cart analysis could only be done using a clickstream data warehouse that tracked individual user behavior, including the value of the goods placed in the shopping cart. Although site traffic statistics from the log file analysis appeared to be good, the actual business trajectory was down once more sophisticated user behavior clickstream analysis was performed. While site traffic statistics are one set of indicators for e-business management, they are not sufficient to run a sophisticated online business. Building a clickstream data warehouse is the only way to gain the necessary understanding of user behavior required to propel the business forward over time. |
Book Endorsements Book Table of Contents Book Authors Referenced Material Related Articles Related Links Download a Project Plan Discussion Forum Links © Copyright 2001, 2002 Clickstream Consulting, All Rights Reserved |