Wading Beyond the Click StreamSuccessful, large-scale Internet marketing demands the best of both direct and traditional marketing efforts. Companies must use multiple communications channels, including e-mail, direct mail, inbound and outbound telemarketing and call centers, and print and broadcast media to acquire, retain and enhance customer relationships.
This is as true for pure dot-com marketers as it is for click-and-mortar marketers. To market effectively your analysis must move past traditional click-stream behavior to truly understand the key behavioral elements of customer relationship building. Marketers not only must understand their own customers, they also must understand them within the context of the entire market and competition.
Failure to fully understand and integrate data from multiple sources, including each customer's site behavior, will yield less than optimum marketing results. On-site activity captures key traffic, navigation and purchase behavior, but Web marketers often neglect the integration and analysis of data from other touch points.
Data collected from every communication and interaction with customers should be automatically incorporated into the marketing analysis process. Integrated database marketing is the major application of the customer relationship management technology that captures this data - and it must enable marketers to work across all channels to profitably keep up with their customers' total interactions with their businesses.
The consumer behavioral measures most important to this integration are those common to direct marketers. These measures view the customer relationship process sequentially:
• Prospect attraction (visitors).
• Conversion (click through to purchase or desired outcome for nonsales sites).
• Purchasing over time.
• Lifetime value.
Channel differences and influences on each of these must be determined to adjust a marketing campaign effectively. Questions include:
• Do visitors from different sites, banner ads and media convert at different rates?
• What Web page-viewing patterns distinguish conversions from nonconversions?
• Do they have different long-term/lifetime values?
• What percentage of my new customers from six months ago are still customers? From 12 months and 24 months ago?
• How do these patterns differ by channel?
• How have these patterns changed over time?
• What differences in these patterns exist for customers calling or e-mailing for service?
• How do my promotions and incentives influence these patterns?
• What purchase behavior distinguishes high lifetime value from low-value customers?
• Does the first product purchased make a difference in customer retention? In lifetime value?
• Which products are purchased together? By what types of customers?
Determining these behavioral measures and answering these questions is key to effective and efficient Internet marketing. Failure to do so in today's increasingly consolidating, bottom line-focused environment may spell doom. One company saw a dramatic example of this: Comprehensive analysis revealed that the entire lifetime value of the customer base was substantially less than a proposed multimedia advertising plan about to be launched.
In interpreting and evaluating these measures, Internet marketers historically suffered the bane of most DMers, as many had extremely detailed knowledge of their customers, but with no total market or competitive context as a base for comparison.
Fortunately, this type of information is now available to Internet marketers. As was the case with traditional marketers - especially those dealing with consumer packaged goods - information services using "projectable" panels or samples of the Internet user population have come to market, providing an ever-increasing array of improved measurements.
Companies such as Media Metrix, NetRatings, Nielsen, comScore, TeaLeaf and others offer services that track total surfing, site visits, stickiness and Web purchases. The value of these information sources, when properly applied, is immense. Not only do they enable the Web marketer to put its consumer franchise in perspective, but they also permit proper evaluation of the real opportunity for increases. This understanding of "How high is up?" is key to resource allocation.
More importantly, when this information is combined with analysis of the total customer-base behavioral measures, these services enable marketers to fully understand the total Web behavior and demographics of your best - and potentially best - customers, so that businesses can acquire more like them. They also enable you to identify other market segments likely to prove profitable.
Identifying the best customers is not an art form - it generally will involve some combination of the tried-and-true direct marketers' RFM. The exact combination that is right for a particular company or business depends on where it is in the process of relationship building and marketing strategy.
Obviously, from these external market-tracking panels, the best sites for banner advertisements can be seen. Through the use of media-linked segmentation systems such as LifePhase, Prizm and MicroVision, the most efficient print and broadcast media can be identified.
Additionally, predictive models can be built from this data to score and rank prospect and suspect lists for direct marketing and telemarketing follow-up efforts, as well as e-mail. The behavior and demographics collected from these sources, as well as from a company's own customer database, can be modeled and linked to sophisticated personalization technologies to provide integrated message targeting across e-mail, direct mail and telemarketing.
Rather than simply being based on a transaction, such personalization takes into account customers' potential lifetime value and cross-product purchase propensities. Finally, the availability of motivational elements from custom attitudinal studies as well as from syndicated studies, such as MindBase from Yankelovich, Norwalk, CT, can be modeled and scored into the personalization technology for e-mail, direct mail and telemarketing, as well as for copy development for print and broadcast media.
By combining analysis of customer behavior with the sophisticated tracking of total market behavior from outside sources, Internet marketers now have at their disposal the best of both worlds.
• Channing Stowell is senior analytics officer at Interelate, Eden Prairie, MN, a customer intelligence application service provider.