Creating True Communication With Customers

Essentially communications works as follows: Organizations emit information, and the information is carried to other organizations that sense and interpret it. Media is the thing outside of organizations, between them, that carries information. Books, movies, white papers, email, and purchase orders are all media. Media affects the information it carries. For example, communicating complex mathematics via music, or emotions via email, can cause information to be lost or incorrectly added. Anyone who has ever played the "telephone game" knows how the spoken word medium can transform information.

Historically mass media has been a one-way street. Movies, TV, and newspapers carry information from a small group of authors to millions of people. The Internet, however, is a marked change from that.

The Internet enables mass two-way communication. Email, Web forms, and online chat allow organizations to emit information and almost immediately sense feedback. In Understanding Media, Marshall McLuhan wrote, "... anybody who begins to examine the patterns of automation finds that perfecting the individual machine by making it automatic involves feedback. That means introducing an information loop or circuit, where before there had been merely a one-way loop or mechanical sequence. Feedback is the end of the lineality that came into the Western world with the Alphabet." And feedback is what makes the Internet fundamentally a new medium.

So what?
Organizations must now use Internet channels like email, Web forms, and online chat to gather feedback from their customers and interpret it to find out what they want. The process is simple and cyclic: emit information, sense responses, interpret them, act, and repeat. This is a big change from traditional monologue marketing, where control belongs to the marketer. In a systemic feedback system, marketers must relinquish some control. eCRM acts as the epidermal layer of an organization: a protective skin, loaded with sensors, that touches media to sense the market environment.

Most of today's eCRM systems are nothing like that. They focus on operational efficiencies like wait times and call center agent productivity. They sense and dispose input, ASAP, without interpreting it. If they do perform any interpretation it usually focuses on customer age, income, address, and past transactions. That's a far cry from figuring out what customers want. Quite frankly, today's eCRM systems are bad listeners.

The case for listening
In the early 1990s Ford, General Motors, and Dodge all sold similar pickup trucks. They all looked and drove like big cars. Dodge, which had a declining 7 percent market share, did market research on new body designs. One design was very macho, with extended fenders, a massive grill, and a powerful engine. Critics and the public hated it; 80 percent of the public didn't like the new shape. However, the remaining 20 percent had an immediate visceral reaction and loved it. Since 20 is greater than 7, Dodge went with the macho design and tripled sales in just one year.

Dodge listened to the market, identified a customer desire, and achieved a 300 percent increase in sales. Research has shown that good listening affects companies' bottom line: In Hearing the Voice of the Market: Competitive Advantage Through Creative Use of Market Information, authors Vincent P. Barabba and Gerald Zaltman write, "... the degree of market orientation, that is, the degree to which the voice of the market is or is not used throughout a business, is strongly related to profitability."

E-listening
On a busy day Borders (a $3 billion per year bookseller) receives 1,000 customer emails. The U.S. Senate receives 1 to 2 million per day. The volume is increasing and new technologies are coming online every year. For example, "chat bots" are software programs that automatically converse with customers using Internet instant messaging. The bots (short for robot) answer simple questions, dispense company information, and stimulate conversation with leading questions like, "So what do you think of the new extended fenders?"

Teasing customer wants out of millions of emails and chat sessions isn't easy. Luckily old techniques and new technologies address this problem. Social scientists have developed a technique called thematic analysis for identifying important themes within large volumes of noisy, multimedia information. Historically thematic analysis was very subjective and labor-intensive. However, advances in linguistic analysis, text mining, and data mining changed all that. We can now analyze gigabytes of text, extract key concepts, and discover key themes. Products in this area include Echo Mail, LexiQuest Mine from SPSS, WordStat from Provalis Research, and Text Analyst from Megaputer. These tools reduce gigabytes of text to concepts, categories, and simple relationships between them, allowing analysts to extract customer wants from the cacophony of incoming messages.

The Internet is fundamentally a new medium because it enables mass, two-way communication. The result is that organizations are being bombarded with millions of emails, Web forms, and chat sessions every month. Organizations can choose to view this as an operational annoyance to be dispatched as quickly and cheaply as possible, or they can view it as a new opportunity to listen to the market and dialog with customers in significant new ways. Thematic analysis, linguistic analysis, and text mining enable the opportunity.

About the Authors
Dr. McVey is a professor of marketing at Webster University and author of the book Value Proposition Marketing: How to Know, Get and Retain Customers. He is an international business consultant who has worked with American Century Investments, Sprint, Bank of Montreal, and PPG. Contact him at awmcvey@earthlink.net

Doug Bryan is a former lecturer in computer science at Stanford University and has done applied software R&D at Lockheed, Stanford University, Accenture, and SPSS. He is currently an independent consultant working in text mining and knowledge management. Contact him at DLBryan@acm.org