Gather More Insight, Not More Data

Online strategies involving a range of personalization technologies and associated solutions, such as e-marketing and customer relationship management, depend on relevant and accurate data collection. Marketers are inundated with data, not all of which are equal. An understanding of the data that can constitute the information profile of customers and site visitors helps guide its use.

Online data come in many forms and offer varying levels of visitor insight, as well as limitations, to understanding customers’ interests and needs. Understanding the respective strengths and weaknesses can help marketers hone in on the right tool for the job.

Web servers and e-business systems collect every click of anonymous and identified interactions. The result is an overwhelming amount of data. Anonymous clickstream data can be summarized into aggregate measurements such as hits, impressions and page views that alone offer limited utility. They give insight into traffic volumes and patterns – information that is most useful in managing site traffic loads but offers little in understanding the behaviors of visitors.

To identify individual user interactions, a unique identifier is needed. The simplest method, a cookie-based form of profiling, tracks movement across a site, providing information on where visitors go, how long they spend and what they click on. But cookie-based profiles tell you nothing once a visitor leaves a site. Though marketers dream of having a captive group of online visitors and customers, the reality is that their activity provides a myopic view of their interests.

The dependability of clickstream data relies on visited page meta keywords or generalized page descriptions. For example, it may be possible to detect that a visitor is interested in sports and golf specifically, but impossible to pick up a preference in the Senior PGA Tour for Jack Nicklaus.

Second-generation profiling, characterized by networked cookies, appends tracking information across a designated group of sites. This method measures behaviors across a network. For richer profiling it is still limited to a small percentage of the sites across the Web. It also suffers from the lack of detailed interest granularity and inaccuracy inherent in clickstream data.

Registration data and clickstream data sources provide only a piece of the picture of individuals’ interests and behaviors and, therefore, have limitations in understanding their online interests.

To develop an enterprisewide vision of customer behavior, e-businesses are consolidating customer data from multiple touch points. However, this company-centric viewpoint is still too limited. Customers disclose only a portion of themselves when interacting with a single enterprise; they display a much fuller spectrum of behavior when they interact with many Web sites. Therefore, if e-businesses want a full picture of their customers, they must augment their corporate knowledge with a Web-wide understanding.

Next-generation profiles track what users are doing in the browser and watch events such as mouse clicks, window scrolling and typing to understand what users are actively reading. A Web page with no affliated scrolling is probably not being read. Recognizing this nuance and wanting to track just “real” behavior, next-generation profiling captures subject context and pulls representative words and phrases only from read pages.

The key to next-generation profiling is the complete understanding of users’ online behavior. No other data source can identify consumer interests with the high level of understanding and the precise granularity of information. Interests can continually be ranked and measured by intensity.

Next-generation profiles also allow marketers to understand consumers’ online behavior and interests at a general and a granular level. The unique profiles are based on observed online behavior and the content that consumers interact with while online. The resulting data can be used to better understand and segment consumers and improve marketing strategies.

David Quan is co-founder and chief technology officer at neoButler Inc., Durham, NC, a provider of turnkey solutions for customer intelligence-driven marketing. Reach him at [email protected]

Related Posts