What does the consumer want? Why do individuals prefer one product or service over another? And how, precisely, do most consumers make their purchasing decisions?
These questions, which have baffled marketers since the first mass-produced product was placed on a shelf for sale, ultimately determine the success or failure of virtually any business venture. And much to the chagrin of many corporate executives, consumer attitudes today are, if anything, harder to read than ever as people freely rummage through an abundance of choices for everything from ordering a cup of coffee to buying a mobile phone to choosing a retirement plan.
But there is help on the way for marketers. Recent work on the art and science of consumer behavior has refined, updated, and strengthened an analytical tool known as consumer choice modeling, initially developed in the 1960s by Daniel McFadden, a winner of the 2000 Nobel Prize in economics. Simply put, this model examines the personal reasons for individual choices and provides techniques researchers can use to measure and predict those choices. By exploring why individuals make specific trade-offs among various product options, consumer choice modeling can determine the features that people in different economic and demographic strata are looking for and how much they are willing to pay.
Originally, this technique suffered from a lack of sophistication. A typical implementation involved asking respondents to react to lengthy paper-and-pencil surveys offering a series of preconfigured and static product or service possibilities. Although some insight about consumer preferences was typically evinced, it was often shallow, limited by researchers’ inability to dynamically change the direction of the questioning on the basis of the responses. However, advances in experimental designs and information technology — including broadband Internet access, digital imaging, video, and faster computing speeds — now allow researchers to better approximate the shopping experience when asking questions by adjusting product choices in reaction to a person’s answers. By analyzing the responses from a representative sample of consumers (or potential future customers), researchers can produce econometric models that depict the relative weighting of specific product features and price points.
Early in 2007, Booz & Company applied consumer choice modeling to identify and measure the drivers of demand for mobile phones. One of the more fascinating conclusions of this study: Although Apple Inc.’s iPhone was still months away from release and its price tag would be higher than that of most other phones, the Booz & Company model correctly predicted that it would be the most attractive overall offering to consumers.
In all, Booz & Company surveyed more than 1,800 consumers in the United States by simulating the actual mobile phone purchasing process and asking people to compare their existing package — device and mobile service contract — with alternatives. For example, owners of low-cost Sharp handsets running on pay-as-you-go carriers such as Virgin Mobile or Boost Mobile were offered a US$100-plus Samsung phone with Nextel service and a $250-plus LG phone with Verizon’s network. Respondents were asked, “If these two packages were your only alternatives, which one would you choose: Samsung/Nextel, LG/ Verizon, or neither?” and “If Samsung/Nextel were your only option, would you purchase it or continue to use your current package?”
The majority of the low- end and midlevel consumers we analyzed were highly commodity driven. Other than by offering an attractive handset price, it is almost impossible to convince an individual to change his or her current mobile phone package. In fact, further analyses revealed that one-third of U.S. consumers are unwilling to change their wireless package, no matter how much the handset price is lowered. Such reluctance to switch is unusually high and shows that the wireless device and service industry has largely failed to provide attractive new products with features that consumers attach real value to — at least since Research in Motion Ltd., the maker of the BlackBerry, combined e-mail and voice in one machine in the 1990s.
Of all phone users, owners of low-end handsets made by the Nokia Corporation value their phone package the least. Consequently, these consumers are the most willing to switch to another carrier and handset — an opportunity for competitors to attack Nokia’s base by, for example, producing a low-cost package with a function or two that outpaces the relatively plain Nokia product.
The consumer choice model also revealed that owners of handsets made by Sony Ericsson Mobile Communications AB, which tend to be highly designed, full-featured products, care much more than Nokia users about functionality, usage range, and purchase location (they prefer to buy their packages at stores that offer personal attention, rather than at Costco or Circuit City, for example). And although these customers, too, are price conscious, they’re willing to pay a premium to have their preferences met. A service provider could use these findings to target Sony Ericsson owners with a slightly less expensive offering that in all other ways matches their current package.
Consumer choice modeling also has the ability to predict the impact of future products and services on the market. To illustrate this, Booz & Company used the data collected from the mobile industry surveys to simulate the characteristics of “the ideal high-end phone” as consumers viewed it. From this, the survey gleaned that three primary factors — feature, design, and brand — are of paramount value to consumers considering a higher-priced model. These factors, of course, were exactly what Apple focused on in developing its blockbuster iPhone, launched in July 2007.
Significantly, as the model predicted, Apple stumbled when it came to price, which the survey showed matters at all levels of cell phone purchases. At a price point of $599 for an eight-gigabyte phone, the research forecasted that Apple would have difficulty reaching a significant portion of the high-end market. But the same research suggested that performance would improve quickly as soon as Apple cut prices. In fact, that is precisely what happened: In September 2007, Apple discounted the phone by $200, and sales rose well over 1,000 percent in the succeeding quarter from sales in the prior three-month period. And in June 2008, CEO Steve Jobs announced a much faster eight-gigabyte iPhone — using AT&T’s state-of-the-art 3G network — for only $199, a move that further aligned Apple’s pricing with that of its peers and that will almost certainly improve the product’s market share.
A perfect research topic for consumer choice modeling would be hybrid-electric vehicles, which have tripled in sales since 2004, though on an admittedly small base. As the first realistic alternative car that addressed environmental and fuel cost concerns, the hybrid was a novel idea that intrigued early adopters. Today, these cars are attractive to consumers put off by higher gasoline prices because they offer improved fuel economy (particularly to urban and suburban drivers) and, since they use the electrical power of the vehicle’s cordless battery when they can, because they do not require a new recharging infrastructure. These obvious benefits notwithstanding, sales of hybrids have also risen because of government tax incentives.
But hybrids are not the only possible response to environmental concerns and high gas prices. Advances in diesel and biofuel technology suggest that there may be more palatable choices to power the traditional automobile engine in the near future. Meanwhile, all-electric and hydrogen-powered vehicles are also in development and show some early promise. In short, the only trend certain in the auto industry these days is uncertainty. A great deal will depend on future environmental and tax policies, but at present, auto companies can focus on one factor they can understand and address: consumer demand. What do consumers really want, as opposed to what they say they want?
Each consumer makes his or her car purchase decision by simultaneously weighing diverse criteria, including brand, cost, performance, fuel economy, comfort, styling, service, environmental friendliness, and more. But if you asked individuals how they weigh these criteria — and many a carmaker has tried — they would be hard-pressed to articulate their decision-making process. Consumers’ choices in today’s complex marketplace are beyond the ability of even the consumers themselves to describe.
A consumer choice modeling project focused on hybrids would offer people different vehicle options and allow them to think like car buyers as they compared their typical car preferences (in other words, past purchases) with various hybrid possibilities. This study would focus on the reasons individuals make specific trade-offs among various options, such as fuel usage, CO2 emissions, battery range, performance, vehicle design, and price. With this data, auto companies could then deduce whether various consumer segments really want an environmentally responsible car, what features they are looking for, and, most important, how much they would be willing to pay. The efficacy of consumer choice modeling is that it allows manufacturers to isolate and identify customer preferences among an array of realistic product offerings without having to ask the open-ended question, “What do you want?”
We believe that consumer choice modeling is ideally suited for analysis of the most complex consumer decision processes and that it yields valuable insights for demand-driven strategy development by providing customer value segmentation maps, measuring market share impact of new product–service combinations, and assessing overall brand equity. Perhaps most important, choice modeling can reveal salient differences between managers’ beliefs about customers’ needs and preferences and customers’ actual needs and preferences. For managers seeking reliable feedback on how customers view their products and services, consumer choice modeling provides a rigorous way to turn customer-driven feedback into profitable and sustainable tactics for retaining or capturing market share.
John Jullens is a principal with Booz & Company in Cleveland. He specializes in demand-side transformation issues, including revenue growth strategies, brand management, customer retention, and retail channel effectiveness.
Gregor Harter is a partner with Booz & Company based in Munich. He focuses on the telecommunications and technology sector, developing market strategies and improving operational performance.