Editor’s note: Jerry Savage is a qualitative research associate at The Hartman Group, Inc., a Bellevue, Wash., research firm.

It seems reasonable - and, indeed, consistent with research and theory in the social sciences - that respondents will feel more comfortable, and thus will be more open, when they are interviewed in an environment that is familiar and comfortable to them. Holding discussion groups in boardrooms with one-way mirrors - as many firms currently do - not only effectively decontextualizes the behavior (and so removes the presence of important behavioral drivers) but also creates a research context that is not conducive to uncovering the cause of consumer behavior. It is within living rooms, kitchens and retail environments that consumers negotiate the meaning of brands and products and, therefore, it is within these environments that my colleagues and I conduct our research.

In essence, we have made things simpler and more personable so as to position ourselves to penetrate consumer segments that have significant influence on the sales and brand equity of our clients. This is altering the status quo of marketing research in the following ways. Recruiting, for instance, is not simply a matter of hiring a teleservices firm to administer a brief instrument to ensure the potential respondent is representative of the segment of interest. Rather this process takes the form of a brief conversation, which includes some essential questions as well as some probing into areas that are not traditionally examined by recruiters. We then arrange to conduct interviews and social network parties in consumers’ homes or retail environments - environments familiar to respondents and those in which people are more likely to talk comfortably and openly about things like value, brands and advertisements.

A study of social network parties and focus groups

To examine some of the differences between focus groups and social network parties, our firm conducted a series of groups between January and June of 2003. Most of these data were gathered as part of a study related to branding strategy and children’s nutrition.

During the first phase of the study we relied heavily upon social network parties, one-on-one in-home interviews, consumer collages and intellectual capital acquired through earlier research with the client. The second phase of the project consisted of a host of questions on several brands within the client’s portfolio. During this latter phase of the study we relied primarily on focus groups.

To kick the study off, we contacted a mother of two to host a social network party on children’s nutrition, shopping behavior and channel decisions. The initial contact was casual and focused on establishing rapport and ensuring the respondent was a good fit for the study’s objectives. This effort - and an appropriate cash incentive - let the respondent know that inviting friends to her house to talk with me about things related to the study’s objectives would be interesting and worthwhile. The party “host” was given some information about the study and told that we simply wanted to have a conversation with a group of people that share information about children’s nutrition. We were then able to work with the host to gather a group of respondents that are active in parent groups and get together to discuss their children as well as schools, local crime and a host of other issues. Some parents were more at the periphery of the social network and so provided an element to the group dynamic that served to ensure some diversity. As is typically the case with groups in which respondents know one another, some respondents assisted the recall of others and, occasionally, respondents corrected one another to make sure that they, to paraphrase, didn’t “mess up the research.” This group, and other similar groups in phase one of the study, contributed to a clearer understanding of the position of my client’s brand identity within specific communities and showed how that identity (and the brand equity tied to it) is negotiated within consumer networks.

A couple months later, my colleagues and I gathered again to conduct a series of groups to build on the findings of the first phase of the study. Part of the purpose of these groups was to explore similarities and differences between focus groups and social network parties and how each of these methods affect top-line research findings and bottom-line business decisions. My primary objective with this element of the study was to explore the group dynamic of each group type and come to a better understanding of what type of groups are most appropriate for specific types of questions. By the end of the project we completed 23 groups, roughly 15 one-on-one interviews, an analysis of 12 collages made by consumers and a cursory analysis of relevant survey research. The multi-method approach put us in a good position to examine the utility of various methods and, more specifically, the utility of focus groups and social network parties for specific types of marketing questions.

Negotiated within networks

This recent work - as well as previous studies we have conducted - has made it clear that the value of products, services and venues of purchase (and many other things) are negotiated within social networks (e.g., groups of family, friends and colleagues) and thus these groups are of particular interest to those who study marketing, economic and political questions. By contrast to the social network parties, the dynamic of the focus groups was contrived. It was simply not the sort of dynamic that average consumers experience on a day-to-day basis and was not representative of the kind of dynamic that is most likely to affect decisions about how to live, where to shop and what to buy. We feel for this reason alone it is indeed clear that the social network party is the superior method for most marketing questions.

Moreover, my colleagues and I have found that exploring group dynamics in a context within which groups of average folks are likely to congregate is critical to the development of accurate and actionable assessments of consumer behavior. It’s not surprising to most of us that consumers don’t really discuss or decide what to buy, where to shop or how to interpret a concept in conference rooms filled with other respondents who they don’t know. Rather, most of us, like our respondents, make such decisions among friends and in settings that we frequent on a day-to-day basis. I must say, it never ceases to amaze me how much money goes into renting focus group facilities with one-way mirrors. Such facilities are so artificial that I’m hard pressed to believe that one could arrive at a generalizable finding based on a study conducted in such a context.

Additionally, the nature of social network parties virtually ensures that the groups will not include “professional” focus group respondents. The problem with these individuals is that they think like the moderator and are quick to offer theories on things like brand identity and corporate positioning rather than simply describing their own behavior. Of course, the purpose of the groups is not so much to get the responses and insight of marketing professionals - or others who are attempting to do that work - but rather to get real consumer insight from average, run-of-the-mill consumers who are representative of the market and/or cultural category of interest. The recruiting method for social network parties allows us to avoid these individuals and also ensures that we don’t rely solely on databases of people who have expressed unique interest in participating in marketing research.

Finally, we have found that people are often quite comfortable speaking up in their own home or that of a friend and that the social network groups tend to be somewhat more productive than focus groups because people are accustomed to interacting with one another. The dynamic between respondents who know one another also serves to reduce response error. For instance, where the memory of one respondent may be a little hazy, other respondents knowledgeable about the behavior of that person are able to chime in so that we get an accurate rendition of the behavior of interest (e.g., what brand of vitamins they buy for their kids).