|
Traditionally, many companies have tried to make use of collected information to better understand customer wants, needs, and behaviors. When cashiers ask customers for their ZIP code or telephone number, you know that the company aggregates the collected information to generate demographic reports. Those reports include information such as the kinds of products purchased, the distance customers traveled to visit the store, and the amount of money spent each visit. At least one of the goals of this data collection is making the company able to more precisely focus its marketing campaigns.
Electronic commerce companies have an even better opportunity to collect this kind of information, especially because of one not-too-subtle difference: At a Web site, you don't have to ask the customer for information. Yet, when a company accumulates only Web statistics, page visit logs, timestamps, and the like, it may be squandering a significant opportunity. Many organizations attempt to make use of that data resource for business benefit, yet extracting actionable knowledge from server logs is certainly a challenging proposal.
It's therefore important to characterize customer behaviors from the start, instead of trying to re-create the user's thought processes from abstract clickstreams. By formulating views of customer actions and properly logging these actions, you can accumulate data with a value that can exceed the demographics on which traditional businesses rely. Within the Web-enabled environment, we refer to that knowledge as e-business intelligence (E-BI).
E-BI Benefits
What is the value of E-BI? Because the overall goal of it is improving the ability to capture and serve customers through microsegmentation, its attendant benefits include:
- Understanding who your users are and what they like to do
- Enabling the creation of a better customized user experience than could be created with a traditional brick-and-mortar site
- Increasing revenue through personalized one-to-one marketing
- Enhancing product branding and customer loyalty
- Decreasing the time to conversion
- Locating and drawing potential customers to your Web site.
What shapes E-BI strategies? Are they any different from business intelligence applications, which frankly, different kinds of companies have been using successfully for many years? The difference lies in the ability to transform the record of customer behavior into an analyzable format from which behavior patterns can be gleaned. In the world of the shopping mall, this would be akin to collecting thousands of store videotapes and being able to search through them all, looking for similar customer behaviors. The kind of knowledge that can be acquired through E-BI has value that is orders of magnitude greater than simple geographic and demographic breakdowns.
E-commerce vendors that can track the movements of every visitor to their Web sites have an opportunity to approach one-to-one marketing strategies through the combination of traditional demographic and psychographic information with online behavioral (clickstream) data. This information can be used in the creation of rich customer profiles, and the mining of these profiles for useful behavioral patterns and the application of the knowledge inherent in those patterns can help solve numerous business problems. Particularly exciting is the potential to convert Web visitors from browsers to purchasers. Profiling customers in the context of an E-BI strategy can assist in providing microsegmentation for targeting value-added products and services for cross-sell and up-sell opportunities.
A Note on Privacy
Although most Web surfers value the benefits that personalization and customization bring to the Web experience, they also harbor considerable reservations about who is collecting data about them and what that data is being used for. Privacy consistently ranks at the top of Internet users' concerns when online.
Clearly, you need to consider privacy issues seriously before jumping headlong into E-BI. At a minimum, e-commerce vendors need to clearly articulate their privacy policies to visitors, with clear explanations of what information is collected from visitors, what the information is being used for, how the collection of this information benefits the user experience at the site, how the information is being secured, and whether the information will ever be sold to third parties.
User Behaviors
What kinds of behaviors should you expect to find? For insight, let's rely on our experience at the shopping mall. As soon as you enter a store, you are bombarded with cues to influence your shopping behavior, whether you know it or not. Whether it is the strategic placement of specific products at eye-level, the way that the paths are laid out within the store, the hiding of the fitting rooms, the location of signage - all these details are carefully planned so as to encourage the browser to turn into a customer.
Store designers study the way people shop in a store to understand the best way to influence what they do. Either by watching videotapes, conducting surveys, or by surreptitiously watching the ebb and flow of strollers around the floor, marketing consultants look for patterns of behavior that lead toward increases in sales. While customer profiles describe what kind of people your users are, customer behavior describes what those users do. Behavior is defined as a recognizable pattern of actions, either exhibited by an individual user or by a set of users interacting within the system.
In the Web environment, we are treated to a microcosmic view of the way customers "stroll around the store." The entrance to the store is the company home page; the presented content provides the visual cues for influencing behavior. But because the interactions within this environment can be tracked electronically, we have an enhanced ability to analyze visitor behavior.
As a simple example, frequently a customer enters a store to learn about a certain product before making a decision to buy it. The kinds of actions this customer exhibits involve asking questions about the product, competing products, product features, and so on. A customer about to buy a product is more likely to ask about price, warranty, and return policies. You can see the same kinds of behaviors at a Web site; by looking at the kinds of content the user views during the visit, you may be able to determine if that customer is learning about the product or is ready to buy it.
User Actions
Actions are activities a user can perform while navigating a Web site. Tracking user behavior involves more than just collecting server log files. Instead of relying on the traditional server log data, you should incorporate a more meaningful characterization of user activity. First, it is necessary to specify the kinds of actions a user may perform while browsing at your site. Get more specific than just page views; rather, you want to superimpose business meaning on top of appropriate page views while ignoring meaningless ones. Behavior modeling then becomes a process of analyzing the sequence of actions that users perform, the context in which those actions are performed, and whether any particular behaviors can be generalized for later predictive purposes.
While each e-business's list of user actions may vary, here is a short list of some user actions that are interesting to log:
  
|
|
|
|
| |||||||||||||||||||||||||||||||
