Those of us who drive along highways in the Northeastern U.S., stay at Marriott Hotels worldwide or shop for consumer goods on the Web are experiencing the benefits of conjoint analysis, a research methodology used to measure consumer preferences that was first developed by Wharton marketing professor
Since then, Green’s career has tracked the growth years of advertising and marketing as the field has become ever more complex and sophisticated. Conjoint analysis is now used by almost all of the Fortune 500 companies and in a variety of other places as well – from book publishers and transportation authorities to government agencies, law firms, hospitals and, most recently, Internet companies. Green and colleague Abba M. Krieger continue to create and refine new research techniques that organizations use to market their products and services.
Consider bundling. “People have been bundling goods and services for years,” Green says, “but doing it mentally. It has only recently been amenable to analytics.” He points to the telecommunications field as an example. “When you arrange for telephone service, there are about 25 different types of service that the company may offer you, such as call waiting, caller ID, or call forwarding. Among these offerings there may be three or four fixed bundles, typically priced at a discount for the total bundle.
“From an analytical standpoint, when there are 25 components that could appear or not appear in various bundles, you are talking about zillions of possibilities. Analytics does two things: it helps you select the actual slate of offerings – which bundles and which separate items there should be – and it helps you price them. Manufacturers are clearly interested in those combinations that will maximize their returns, subject to respecting the preferences of the individual customers.”
The potential uses for bundling are enormous, says Green. Other businesses that can benefit from bundling information include computer hardware and software manufacturers; cable and satellite television companies that bundle various channel offerings; and even the local sports team or symphony orchestra, as they try to determine what particular series or package of tickets are most appealing.
Meanwhile, Green and Krieger are working with another industry – pharmaceuticals – using conjoint analysis to gather information on drug use and physician preferences. The issue concerns pharmaceutical manufacturers’ attempts to target advertising more and more at end-users, trying a “pull-through” strategy in which patients ask their physicians for specific medications. While drug manufacturers have spent heavily on television and print advertising aimed directly at consumers, “the results so far are hard to measure,” Green says. With data gathered through conjoint analysis, however, manufacturers may be able to change their messages quickly enough to respond to changing preferences articulated by end-users.
In conjoint analysis, researchers – using information gathered through surveys and questionnaires – measure trade-offs that consumers are willing to make, often unconsciously, when they buy a particular product or service. For example, a consumer might pay more for a product if it can be conveniently bought, or if it has a more attractive design than a competitor’s version. An employee might take a certain job, even if it pays less than a similar position, if the job has better hours, better benefits, or offers more opportunity for advancement.
Measuring these trade-offs requires sophisticated software programs and mathematical models to interpret consumer preferences and to predict how the market will react to new products or services. In the end, the company gathering the information has a stronger understanding of what consumers want and can tailor both its product and its marketing strategy far more precisely.
Whereas marketing theories typically tend to marinate in academia for a decade or so before “real world” practitioners take notice, conjoint analysis was immediately seized on by marketing research specialty firms as well as large corporations, all of whom saw its immediate value as a tool for developing products that consumers actually want.
Green’s first commercial application of conjoint analysis was with Bissell, a manufacturer of spot cleaning equipment interested in creating a new kind of product container. Two decades later, Marriott hired Green and his colleague, Jerry Wind, to conduct a large-scale consumer study among business and non-business travelers. The goal was to establish an “optimal” hotel that would attract new customers without cannibalizing existing business. The eventual result was the “Courtyard by Marriott” concept, which grew from three hotels in 1983 to approximately 450 hotels worldwide now with sales in the billions.
And in the 1990s, Green and Krieger helped devise a methodology for collecting data for the now highly-successful E-ZPass system for automated toll collections used on highways and bridges in the Northeastern U.S. “We looked at such things as how many lanes should be available, how many people were expected to switch from the traditional system to the E-Zpass lanes, how one should pay for the transponder tag, possible discounts for heavy usage, and so forth,” Green says. “The forecasts we made 10 years ago have turned out to be on the money.”
The advent of the Internet has, of course, had an enormous impact on the marketplace in recent years. But where many predict a future in which virtually all business will soon be done on the web, Green sees no imminent revolution: “Stores are not going to close their doors. Many people still want to see and touch a product before they buy it.” Instead, he believes that a retailer’s Internet site will complement physical locations.
Web sites can, he observes, play a vital role in helping purchasers make buying decisions and helping companies gather information about buyers’ preferences. Green notes that conjoint analysis is ideally suited for the Internet. Sites such as “activebuyersguide.com”, activeresearch.com”, “sawtoothsoftware.com”, and “personalogic.com” all use variations of conjoint analysis to lead a consumer through a series of questions to determine those attributes and features that are most important to the consumer as he or she makes her purchasing decision.
He also foresees high growth potential on the Internet for business-to-business commerce, especially for the purchase and sale of products that are well-specified in terms of features and price. “If a company is looking for sulfuric acid or a specific type of light fixture, the Web can be an easy way to search for a supplier and ask for a price quote.”
In the business-to-consumer area, Green makes note of the number of companies – from Priceline to Amazon – that have recently “overextended” themselves. He also wonders about regulatory issues raised by the Internet, including the possibility of a sales tax on goods bought and sold on the Web. “That issue is still up for grabs,” he says. “But I think taxes are bound to happen eventually because the lure for governments in being able to get those extra funds will be extremely enticing.”
Despite all the changes that the Internet and other economic developments have brought over the years, Green says the basic thrust of marketing strategy remains the same: “Finding a value proposition that works, that is unique, and then trying to find a market niche for which this value proposition will be attractive … It’s very rare that any one product can be all things to all people, which is why you are finding much more interest these days in market carving and the development of products or services that respond to a particular segment.”
