Asking e-Customers the Right Questions

Most Web merchants try to learn more about customers by analyzing purchasing trends, abandoned order forms or transaction details. But Kana Advisor 6, a "guided selling" program from Kana Communications of Redwood City, Calif., lets them actually ask consumers what they want.

By asking a site visitor questions and then comparing the answers to a database of customer profiles, Advisor can make product suggestions based on the customer's apparent needs. The product also helps with internal sales when suppliers ask vendors for special parts, best prices or other consultative pre-transaction details. It can be used as part of a comparison-shopping engine or can be set to pop up on screen at the time of purchase.

Customer service is still missing from most online transactions, according to Lou Agosta, Chicago-based director of data warehouse research for Giga Information Group. Without it, companies seeking to upsell or cross-sell at the time of purchase have to rely on previous transactions to see what other similar customers have purchased.

Most online merchants have been unable to move beyond rudimentary tracking and best-guess recommendations ("people who have bought this item have also bought...") to apply business logic to customer behavior and link purchase transactions to other possible sales. But Advisor can compare items from one or several manufacturers, according to stefania Nappi, Kana vice-president and the creator of Advisor.

Unlike systems from Ask Jeeves or Soliloquy that use natural language search technology and let users input questions, Advisor uses preset questions and then compiles answers and compares them to a stored database of profiles to infer details about users. Companies front-load the system with market research to identify customer types, then as the number of responses increases Advisor "learns" to propose appropriate products.

In release 6, Kana adds a browser-based advisor configuration tool that allows a business unit employee to enter the data needed to create an advisor to aid in product sales. Teams that combine marketing, financial, sales and advertising and product management can collaborate to set business rules that govern the advisor's behavior.

Several Kana customers use a personality to humanize their advisor, such as "Emily" at or "Jill" at CompUSA. The idea behind such personification, says Kana's Nappi, is to make the user more comfortable in sharing information.

Knowledge gathered from their advisor has unearthed details useful in marketing, according to Jonathan Sills,'s vice president for strategy and development in San Diego. Since many customers have described their ideal flowers as "bright and cheerful," the company has placed an emphasis on photography, Sills says. "We know vibrant colors appeal to the consumer."

Other features, such as a reminder list and a wish list, let a vendor anticipate demand or inspire customers to return more frequently, Sills adds. The addition of an online flower encyclopedia and advice on flower care goes beyond mere online ordering and offers a chance to build goodwill while learning about customer interests.

E-tailers need to find the reasons why certain products are examined and rejected by potential purchasers. With the help of a product like Kana, online retailers can use information captured to discover customer needs that are not being met.

Says Larry Hawes, senior analyst at the Delphi Group in Boston: "With more information available, a company may change its approach to the market."

CRM Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues