Getting Close to the Customer: Quantitative vs. Qualitative ApproachesPublished: May 05, 2004 in Knowledge@Wharton
New ways to gauge customer feelings and attitudes are increasingly complementing numbers-oriented methods, although companies should never take an either-or-approach, according to Wharton faculty. “Thanks to technology and other innovations we know far more about our customers than we ever did before, but I would contend that we don’t understand more about our customers than we did 40 years ago,” says Wharton marketing professor Peter Fader. “We can put each customer’s order on a microchip, but as far as having a sense of what’s inside making him tick,” the answers remain elusive.
New technology to gather and process data can be overwhelming for marketers, he adds, noting that some companies are tracking consumers with radio devices and installing video cameras in homes. “It sounds compelling [as a way] to understand how people are using and consuming the products, but in the end you have miles of videotape,” says Fader. “It’s not a complete waste of time, but it’s a matter of knowing when to say, ‘When.’”
Companies need to weigh the benefits of launching huge data projects against the cost. If the payoff isn’t there, Fader suggests, then marketers risk a backlash in which general managers give up trying to understand the customer at all and rely solely on quantitative information. The end result could be that “the marketing function shrivels up and goes away and gets replaced by order-takers or inventory control specialists.”
Fader points to Wal-Mart’s technological focus, which has been used to wring costs from the company’s supply chain, as one effective strategy. “Companies have to pick their position. Wal-Mart has chosen a position of being on the ‘cost’ side. Many other companies claim to be on the customer intimacy side, but they are really just giving it lip service.” Fierce competition, he adds, has led companies to build proprietary marketing models that replace broader data formerly collected by industry associations. “In the old days there used to be much more transparency. The models were simpler. Industry associations had a bigger role. Companies said, ‘If we can’t get data, then let’s share.’”
From Data Collection to Creative Insight
Wharton marketing professor George Day suggests that quantitative approaches such as data mining – in which data is processed to find trends – are valuable for companies with established products. However, companies searching for new organic growth – in markets that are evolving or where technologies are converging – may not even have data, so in those areas qualitative methods are useful. “Even when quantitative data is most preferred you still need the qualitative [approach] to interpret what’s going on, to give it depth,” says Day. “You can’t hand an ad copywriter behavioral data and expect him to do a lot. He needs a more refined, colorful, multi-dimensional picture of the target customer.”
The real issue for companies is how customer insights are used, no matter what method is chosen to gather them. One interesting application, says Day, is occurring at Whirlpool and Procter & Gamble where customer information goes into building “concept banks” that drive new product development. This information, which might include consumer reaction to certain product attributes, can be mixed and matched to develop new concepts. Then, Day says, the company can take an intermediate, more qualitative step by getting an in-house group to brainstorm ways to translate those concepts into products. “You still have to get from the data to a creative insight. Someone has to have an intuition, a sense that this is worth pursuing.”
Marketing professor Americus Reed has noticed a resurgence of techniques on the qualitative side of market research in which companies try to get customers to project their true opinions. “There are certain situations, topics and issues that consumers may not be willing to discuss candidly. [Perhaps] they don’t want to communicate a negative self-image to someone else, or it could be they may not even know the reason why they do certain things,” says Reed. To find the truth, marketers get subjects to “project” what they really think. For example, consumers might be asked to view a cartoon of people in a certain situation, then fill in the dialog bubble over the characters’ heads. Alternatively, they might simply be asked what they think about the scenario, or to describe what is going on. “It’s an open-ended elicitation. The whole idea of projective techniques is people will project their own opinions and attitudes.”
On the quantitative side, Reed says, market researchers are looking into methods that not only record the responses consumers have to questionnaires, but more importantly the response time, or the implicit reaction. Again, survey participants might be reluctant to say what they really think when it comes to certain issues. Think Viagra or McDonald’s. “What you’re interested in is how quickly they respond. You’re looking at reactions in milliseconds. The stronger the association, the faster they should be able to do it,” says Reed, adding that this type of research was pioneered in social sciences as a way to understand reactions to controversial issues related to race or gender. It is now increasingly being applied to commercial research.
The Harley-Davidson Approach
Wharton marketing professor David Reibstein says qualitative approaches should be used in the period preceding quantitative research to establish the reactions to be studied or the terminology that will be adopted. “There are still a lot of the old standard techniques, like focus groups, that people use, but it’s all sort of a preamble to a more formal, larger sample size to get better results.” The rise of the Internet has made quantitative data collection much easier and allowed companies to analyze survey results faster, Reibstein adds. “I think there is value in the qualitative information, but I don’t see a wholesale switching back to it.”
He points to an example of a quantitative approach that is delivering strong results – the use of so-called split cable systems. In certain communities cable television companies show different commercials to different households who agree to participate. Those households then monitor what they buy so that marketers can determine whether the commercials had an impact on purchasing. These findings affect such decisions as how much advertising to do and how the specific ads should be focused.
Another intriguing approach to market research, says Wharton marketing professor Patti Williams, is the development of brand communities, in which companies bring groups of customers together to build deeper relationships. She points to Harley-Davidson managers who regularly put on leather jackets to ride with their customers at company-sponsored events. “Management goes out and does the event and is there for the hog ride. That’s a real qualitative way to collect data. They are building and developing trust and intimacy in a unique way.”
In addition, pharmaceuticals have developed web-based communities for patients to interact with each other and the company. “I don’t think consumers always know what they want,” says Williams. “With these experiential brand communities you don’t have to ask them anything directly. You can just observe. It’s more of an anthropological approach.” At the same time, there are limits on the kind of products that lend themselves to a brand community, she says, pointing to toilet paper as an example. Also, building brand communities is expensive and sometimes verges on being fashionable. “A lot of people are jumping on this [even though] it’s hard to do it well. Saturn has [used this approach] from the beginning. Jeep does Jeep Jamborees but could Hyundai?”
Williams, too, stresses a need for both hard numbers and more qualitative insights. “The best brand managers are open to whatever way gets the best answers. I don’t think the qualitative approaches are necessarily better, but I do think they have largely been neglected and are increasingly popular.”
According to Wharton marketing professor Wes Hutchinson, the popularity of qualitative and quantitative techniques falls along a continuum. “With the quantitative approaches the big news is how much information there is to be mined,” he says, but “in industrial markets and in business-to-business, getting to know your customer still means a sales force.”
In a sense, “this all started with computer automation. Customer Relationship Marketing is a mass version of what good sales forces have always done.” The real question is whether the marketing efforts are cost effective. Reaching Six Sigma quality, in which a process generates no more than 3.4 defects per million, is an admirable goal, but at what price? “You can have high quality but if you can’t cut costs to achieve that, then you have lost money,” Hutchinson says. “There’s a limit to where it’s actually profitable to continue to delight your customer … There has even been a small backlash,” with some people questioning the idea that more is always preferable. “Your profits might have been better at Five Sigma.”
Customer Advisory Boards
Some companies are creating customer advisory boards to help them find out what their customers really want. “What we have learned is that qualitative information is very powerful to the executives,” says Sean Geehan, president of Geehan Advisory Boards in Dayton, Ohio, which builds and manages customer boards. “They want trends. They want to identify broad strokes from profitable customer segments that can help them establish a sustainable competitive advantage.” Geehan’s clients have told him that when they set out to use more quantitative measures in market research, the data is often manipulated by a third party that does not fully grasp the firm’s own needs and strategies.