Behavioral Analytics: The Why and How of E-Shopping
By determining the behavioral preferences of the customers coming to their store, e-marketers can align language and behavior of the Web site to provide those customers with the best shopping experience, one that feels tailored for them. Similarly, outbound marketing and traffic generation activities can be aligned.
03/12/08 4:00 AM PT
Think of this scene: A woman walks into a high-end department store on Fifth Avenue. As she enters, a salesperson rushes up to her, shouting: "Welcome back, Ms. Smith. Last time you were here, you bought some makeup. I bet you want to buy some perfume today, right?"
Web-based commerce has made massive strides in the past 10 years. The leaders in the space are using manifold technologies in an effort to attract qualified customers -- and to keep them returning. However, simple examples like the one above in which e-commerce best practices are compared with those employed in traditional stores lead some behavioral experts to wonder if something is missing.
Today's Analytics - the Who and the What
Over the last few years, there has been a significant shift in the focus of e-marketers. Initially, the promise of e-commerce was the sheer accessibility and volume potential of the Web site. The problem was marketing the Web site and driving customer traffic to the site. More recently, the focus of e-marketers has been how to close and retain customers, driving them directly back to their site based on their previous positive experiences at that store. In non-cyberspace terms, it's why we drive past the local supermarket to go to our preferred place to buy groceries.
As e-commerce technologies have evolved over the past decade, the understanding of the customer's behavior both online and offline has increased significantly. By integrating Web-tracking tools with other marketing systems, we determine the customer's demographic with relative ease. Understanding gender, location, income and other demographic factors provides important information on that customer's potential buying patterns. Similarly, past behavior on the Web site, including purchase history, browsing behaviors and other sites visited add to that important information.
Traditional Web analytics products provide an important view of Web site statistics. Many of these tools provide volumes (and volumes!) of information for the statistically biased e-marketer to consume, analyze and reanalyze. If visitors from the East Coast of the U.S. using Internet Explorer 6.0 are abandoning on the checkout page, your favorite analytics tool will point it out.
Extremely sophisticated product relevance solutions can determine what the customer might buy. Systems that analyze past purchase patterns can recommend related products (see the example above). Advertising placement solutions ensure that snowplows are not presented as "this week's special offer" to a Houston resident in November. Other solutions use "crowd behavior" as an indication of where the individual will click next.
The result of these very complex and sophisticated solutions is a detailed understanding of who the customer is and what they might want to buy.
Mr. Jones Goes Selling
Mr. Jones is a seasoned -- let's call him veteran -- salesperson. He's been doing it for 30 years and loves his job. He is the guy you meet that describes himself as a people person. Plus, he is very, very successful. He describes the selling process as a relationship. When he talks about his customers, he refers to them by their first name, describes the transaction as a journey and talks about their motivations and goals. His customers keep buying from him.
New behavioral analytics products are emerging that attempt to derive a new understanding of e-commerce customers beyond the demographic and historical tracking described above. These products are mapping Web-based actions -- keyword searches, navigation paths, click-patterns -- to traditional human communication patterns. Behavioral scientists and psychologists are using this input and applying compute power such as fast artificial neural networks to understand the more human factors of the customer's behavior as they go through the buying process.
Having a Mr. Jones-like understanding of the customer based on e-commerce site navigation has its challenges. After all, he has the advantage of using all of the subconscious behaviors that humans use in natural communication. Whether he knows it or not, Mr. Jones is constantly taking cues from his customers' body language (about 50 percent of the feedback), their tone (about 35 percent of the feedback) and their language -- the actual words that they are using (about 10 percent of the feedback). As a good salesperson, Mr. Jones understands that he can only make a sale (regardless of what he is selling) if he aligns his language and communication style with that of his customer.
Since body language and tone are not discernible from a Web dialog, these new behavioral analytics products are focusing on interpreting Web-based behavior as a proxy for the language used in a person-to-person communication. Independent research studies have shown that significant behavioral cues can be derived from Web-based dialogs to very high accuracies when compared to face-to-face interviews.
A New Customer View
Behavioral analytics products use this new customer information to derive the best way to communicate with a customer. Does that customer make decisions based on the product specifications and benefits or are they more likely to be influenced by customer success stories or reviews? (Mr. Jones knows that there is no better way to lose a sale than to say to a customer in the first group, "You are just like Bill Brown down the road, and he bought our product".) Is that customer motivated by the greater goals of using the product or by problem avoidance? How should the sale and the product be presented to the customer: as one with many choices going forward or as one step in a procedure? Will that customer be influenced by brand? Where on the adoption curve will that customer buy?
E-commerce marketers using these new products are experiencing a new view of their customers. By determining the behavioral preferences of the customers coming to their store, they can align language and behavior of the Web site to provide those customers with the best shopping experience, one that feels tailored for them. Similarly, outbound marketing and traffic generation activities can be aligned. This new awareness can have astounding results with some companies seeing 20 percent increases in revenue run-rates and 50 percent decreases in cost-per-conversion for pay-per-click campaigns.
So what does this mean for the future of e-commerce? Simply put, more personalization. E-commerce Web sites will drive to become differentiated by the customer experience and level of personalization on the site, resulting in a better shopping experience for all of us.
Mark Nagaitis is CEO of 7 Billion People, which provides software for e-retailers so they can personalize individuals' online shopping experiences to increase customer loyalty and closure rates.