When traditional research trumps big data

Big data mining, online communities, predictive analytics, and other contemporary research tools and practices might dominate business news and workplace chatter, but a centuries-old, tried-and-true research method is also attracting its fair share of attention.

As companies seek out research methods that allow them to push beyond the “what” that’s now so readily accessible in digitized customer data to the “why” and the “how” that can best be determined by getting closer to customers’ behaviors, ethnography is experiencing a resurgence. Companies such as Procter & Gamble, Microsoft, and Intel now employ in-house anthropologists and consumer psychologists to study consumer behavior. Other companies, including Samsung, Microsoft, Adidas, and Lego, employ research firms that specialize in ethnography. One such firm, Red Associates (based in Copenhagen and New York), has shadowed consumers in London and Shanghai to help Intel measure social networking’s impact on the use of mobile devices and has visited house parties in Europe to help Pernod Ricard sell more vodka to baby boomers.

This kind of high-touch research doesn’t come cheap: Red charges $250,000 to $300,000 a month for its services. According to Maritz Research, a fully loaded ethnographic research study could range between $25,000 and $90,000, depending on the target audience and size of the program.

The Paris office of Landor, a global branding firm, spends anywhere from €5,000 for a concept or product name test in an ethnographic setting to €150,000 for in-depth, yearlong ethnographic studies. The cost is worth it, though, because ethnographic research offers insight that digital research or traditional focus groups don’t, says Luc Speisser, managing director of Landor’s Paris office. “People talking about packaging or ideas behind a hidden mirror is probably not the best context to make them comfortable and get them to tell the truth. It’s quite an artificial situation.”

In 2007, to provide richer consumer insights for clients such as Danone, Kraft Foods, Procter & Gamble, and Visa Europe, and to gather insights to win new business, Landor’s Paris office launched Landor Families, an ongoing ethnographic study of French families. For the past seven years, Landor has visited 11 families in their homes twice a year, peeking into their refrigerators and gathering information on their buying habits. The families also provide the agency with monthly updates on their day-to-day consumption, changes in behavior, and opinions on products, and Landor shares its findings with clients.

The multiyear study is a unique effort for Landor, but the firm has conducted shorter-term ethnographic projects, including an ethnographic study on sustainability habits for Procter & Gamble in Paris and London. Marketing News caught up with Speisser and Clément Chabert, strategy consultant at Landor’s Paris office, to discuss the project’s execution and resulting insights, and the future of ethnographic research.

Marketing News: How did you recruit the families for this study, and were there incentives involved?

Clément Chabert: We used an external recruiter to ensure a good level of quality in terms of recruitment. We started with 10 families and we have 11 families now because one of the couples broke up, and we decided to keep following them and the two families resulting from the breakup.

We used criteria like age, gender, children or no children, people living in cities, people living in remote areas, but we also used attitudinal and behavioral criteria. One group is called “reformers.” They’re early adopters, more liberal, more progressive, and more prone to adapt innovation. Another is “strugglers,” who are laggards in terms of innovation and adoption cycles. We have a young couple living in the center of Paris with young kids with a pretty high income, and their value system is liberal and progressive. On the other hand, we have a couple in their 60s with less disposable income.

They do not strictly represent all families in France, but our aim was to represent the diversity of contemporary families. It was a choice from the beginning. We didn’t want to apply a certain percentage of, say, in France, 30 percent of families are traditional families with a husband and wife and two kids, so let’s take three families out of 10 to protect that proportion. We didn’t want to recruit like that. We really wanted to recruit to get the broadest diversity of families.

Luc Speisser: The families have two types of incentives. The main one is financial. They have financial incentives every month, but beyond the financial part, they get to have occasions where they talk as a family. The feedback we have from many families is that the incentives aren’t just about getting a check at the end of the month, but also they’re interested in the observation itself.

CC:  The financial incentives are the same level of incentive you get when you’re participating in a focus group, so they’re mostly symbolic.

MN: What type of methodology did you use for the study?

LS:  There was a big Landor meeting in San Francisco in 2006 called One Landor. One of the outcomes of that meeting was that we had to get closer to the customers. We already had many qualitative and quantitative studies, but we wanted to get closer to the customers.

When we came back home to Paris we remembered when [Landor founder] Walter Landor created a mock supermarket in the 1960s to test the packaging for Kellogg’s Corn Flakes. At Landor headquarters, customers were asked to react to it in front of a cereal shelf that had been built there for that purpose. We told each other, “We should get back to that mentality,” so we gave ourselves five principles. The first one was that we would like to observe families over time, not just during one day, which is what you usually do when you do “in-homes.” I had done a lot of those. When you do an in-home with Unilever about detergent, you basically spend one hour talking about the life of the person you’re questioning and you spend eight hours looking at how the person does laundry. It’s quite restrictive. We wanted to go much further than that, to not only look at what was changing, but understand why, and we didn’t want to put a limit on time.

The second principle was that we wanted to meet families in the comfort of their own homes. We all had experience in marketing research, and looking at people talking about packaging or ideas behind a hidden mirror is probably not the best context to make them comfortable and get them to tell the truth. It’s quite an artificial situation.

The third principle was that we wanted to really understand family dynamics. If you’re looking at individuals, you don’t get the overall approach. It was important to look at the family dynamic and try to understand it.

The fourth principle was about trying to interact with family members as individuals. That was the counterpart of looking at families as a whole, and we really wanted all sorts of ways to look at each individual to see what they had in mind.

The fifth principle was to dig really deep into the contradictions between what they say and what they do.

Landor_people_work_104
Kelloggs cereal circa 1950's (Ref: Binder 2 00397_s_10ag6ctwzj0395)

CC:  In terms of how we observed them, we have families answer questions online every month, and we visit their houses twice a year. Once, instead of meeting them in their homes, we decided to go with them to the supermarket to observe them grocery shopping. Another time, we spent some time observing how they do online shopping.

LS:  Every month, the set of questions we send focuses on just one theme—for instance, digital—and some questions might be, “What is a brand that has disappointed you this month?” or “What is a brand that you discovered?” Then sometimes we have a product to test.

Contractually, we ask them 10 to 15 questions each month. They can answer via email or we have a blog where they answer, and they can send us their receipts from the supermarket or send us pictures of their fridge once they’ve done the shopping.

We can also call them back to ask them about contradictions. We started in 2007, so we’ve got more than seven years of tracking. Sometimes we can challenge them by asking questions like, “That’s interesting that you’re saying that now because two years ago, you said the contrary, so can you tell us more about that?” That’s what’s exciting about the study. We have that interplay with them. It’s a bias, definitely, but it’s also an opportunity for us to get the truth from them.

They know Clément very well, so they sometimes tell us details that are quite embarrassing because they’re really close. An important thing in the way we interact with them is that Clément sees them and interacts with them, and I’ve never met them. We tend to maintain that duality. I’ll never meet them because I don’t want to be emotional about them whatsoever. It keeps me very objective and, at the same time, leaves me to be free when I talk about them.

CC:  And talking about this relationship of intimacy and trust, we meet with them twice a year and we question them every month, but sometimes they get back to me spontaneously. For one family that I last met with in person in February, the mother gave me a call in March to talk about the customer experience she had when she bought a pair of shoes for her son.

MN: High-touch research like this is less common in the digital era. What insights can you get from this type of research that you can’t get from online research, or even from in-person focus groups?

LS:  We had this woman who is a single mother and she has limited revenue, and when we discussed with her in 2007 about whether she buys organic products, she almost laughed and said: “Are you kidding? It’s too expensive for me. I’ll never buy organic products. I can’t afford them.” Six months later, Clément went to her apartment and saw a catalog on the desk for organic products. He said, “Are you buying those?” She said, “Yes, definitely.” And the price was more expensive than organic products you’d see in the supermarket.

After discussing it with her, we discovered that she didn’t absolutely care about the price. She bought the products because they were organic, but most of all because the way they were sold was through parties like Tupperware parties. It enabled her to meet her new neighbors.

It makes you think that if you’re working with an organic brand or a company that’s selling organic products, you might want to reconsider the way you distribute them or explore new ways of distribution that add relationship or socialization benefits. You would never have discovered this if you had done a typical study because that woman would have told you, “Organic products: never,” in one survey and maybe 12 months later, she might have said she bought organic products. You would never understand why that changed. This shows the richness of the human being and that they might have, on paper, contradictory behaviors, but they’re always driven by emotional motivation that you wouldn’t have expected had you not discovered that catalog on the table.

MN: Consumers in this type of situation might change their behaviors if they know they’re being watched by researchers. How do you account for possible bias?

LS:  We were contacted about Landor Families very quickly by the media, by the equivalent of the Financial Times in France, Les Échos, and by national radio, and they loved the ethnographic principle. This created interest from the government and the French Ministry for Research and Higher Education, so they sent someone to us to audit the methodology. They assigned a PhD to work with us for three years, and she really worked with us on the bias and, yes, there are some biases, but they’ve been identified by a scientific and sociological approach, so they’re part of the study and we’re aware of them, but they’re also scientifically validated.

CC:  Also, we’re talking about families in the mix, and if one individual in the family tends to beat around the bush, someone else in the family might say, “You’re exaggerating.”

LS:  The PhD working on the methodology told us there’s always a bias when you ask questions about behaviors, but the biases from the families are easier to identify because of what Clément just said. And the bias is lower than when you have a focus group with 10 ladies around a package of laundry detergent and ask them in three hours what they like and understand about the packaging. In real life, they spend six seconds to make that choice in the supermarket.

MN: Tell me about the insights you gained from the project. How are you using the data?

LS:  We didn’t create the tool to sell it. We created it to improve our service and performance to our clients. That’s how I sold the project to our former CEO. I said: “Don’t ask me to do a profit and loss statement [P&L] about it. The P&L will be that this will enable us to win some business, to improve the quality of our work, to make it more insightful.”

We won a huge piece of business for a big brand in France, Lu, which is part of Kraft Foods, three or six months after starting the families project. We won the whole breakfast line because before we made the pitch, we asked our families, really quickly, how they took their breakfast. We came up with a very quick spectrum of different behaviors and attitudes toward breakfast. They told us, “Wow, you’ve gathered in one week as much information as we have in 10 years.” The objective of Landor Families isn’t to replace all of the other studies and research we have. It’s an interesting point of view, which enables us to be very agile and it’s very relevant to what is changing in customers’ minds.

MN: What is the cost of this type of research versus other research forms?

CC:  It’s equivalent, money-wise, to focus groups. On the breakfast example, if we had monetized it at the time, it was the equivalent of two good focus groups.

MN: Ethnographic research is becoming a huge trend in marketing, but simply tracking behavior isn’t enough. How do you interpret those behaviors and extrapolate on them to come up with other insights about consumers?

LS:  When we do research, getting insights is basically our job. It’s rare that you get an insight from a customer verbatim. We cross what the different families are telling us, and the different behaviors and attitudes, and we look at paradoxes and contradictions, and we have some soft signals that we use as hypotheses. Then we cross them with broader trend studies coming from WPP companies like the Futures Company and Millward Brown, and we look at those to see if they have the potential to transform into bigger trends.

MN: How will ethnographic research impact marketing research in the future? 

LS:  Landor Families raised a lot of interest in France on a national scale. Two years ago, there was a think tank event with Danone, L’Oreal, all of the big brands in France, called National Research Day. It covered all of the best practices on research—ethnographic, digital, whatever. I was asked to be one of the keynote speakers to present the first four to five years of results for Landor Families.

It was interesting because I realized I was in the middle of all of the big French research companies that were all aligning on the fact that research had to be much more balanced toward ethnographics. Everybody realized that typical quantitative research, even though it’s still quite important, wasn’t sufficient to give us everything we need to know about the customers, because a lot of research is looking at customers in the rearview mirror. It’s not predictive. It tells you what people think today, but our job is about knowing what people might think and do tomorrow, so you need to look at other ways and make assumptions. Your assumptions might be wrong, but you might also figure out future trends by looking at soft signals.

Ethnographics has a very big future because everybody realizes that traditional studies are not enough anymore. There are all of those companies that spend three years doing qualitative and quantitative research before launching a product, and they launch and, ultimately, there are still nine out of 10 innovations that are failing. Even though quantitative research is key to observe mass movement, Landor Families would be nothing if it couldn’t compare or integrate with bigger studies. Those two dimensions are necessary and complementary. It’s two different ways of looking at people, and they’re completing each other.

MN: How long will Landor Families continue?

LS:  We asked ourselves in the beginning how we would draw up contracts with the families. Do we ask them to commit for one year? For two years? We decided to set them free, and tell them they could join and leave whenever they want, and they’ve all stayed. We will stop when we feel that we’re not learning as much as we could learn other ways. So far, we’re still learning from them tremendously and have identified several major shifts in French consumer behavior.

Key takeaways

  • As companies seek out research methods that allow them to push beyond the “what” that’s now so readily accessible in digitized customer data to the “why” and the “how” that can best be determined by getting closer to customers’ behaviors, ethnography is experiencing a resurgence.
  • In 2007, to provide richer consumer insights for clients such as Danone, Kraft Foods, Procter & Gamble, and Visa Europe, and to gather insights to win new business, Landor’s Paris office launched Landor Families, an ongoing ethnographic study of French families.
  • For the past seven years, Landor has visited 11 families in their homes twice a year, peeking into their refrigerators and gathering information on their buying habits. The families also provide the agency with monthly updates on their day-to-day consumption, changes in behavior, and opinions on products, and Landor shares its findings with clients.

 

This article was first published as “C’est la vie” by Christine Birkner in Marketing News (June 2014), the flagship publication of the American Marketing Association.
© 2014 American Marketing Association. All rights reserved.