AMIC - Ad Info



Are Focus Groups REAL Research? Shouldn't we just do a survey?
Can't we just do a few groups? Do we really need the quantitative?

(Adapted from "FOCUS GROUPS: HOCUS POCUS?" published in MARKETING TOOLS, August 1994.
© 1994 by Bill Weylock)

Bill Weylock is President of Weylock Associates Inc,
specializing in business-to-business and interactive marketing research.

You hear a lot about focus groups these days - at cocktail parties, in business meetings, even in television sitcoms. You also read a lot about focus groups - in national magazines, in business journals, and even in college level textbooks. 

On the positive side, you hear that focus groups have helped political campaigns, increased sales of faltering products, raised customer satisfaction levels, and helped build award-winning advertising campaigns. Some of the accounts make it sound as if groups are an almost magical technique, working these marketing miracles with little help from human beings. 

On the negative side, you hear that groups are unscientific voodoo that tell people only what they want to hear - a crutch for lazy and irresponsible marketers. Pointing out, correctly, that groups are easily misused, irresponsible detractors go on to damn the entire focus group approach as pseudo-research. 

I make a lot of my living designing, running, and reporting findings from focus groups, so you won't be surprised that I come down on the positive side. If focus groups are used when appropriate, and if they are properly designed, conducted, and analyzed, they are extremely valuable. 

I agree, however, that if they are used for inappropriate tasks, if they are designed carelessly or run by unskilled moderators, and if they are not analyzed carefully by a professional, they can be dangerously misleading. 

Many people have the idea that focus groups are an alternative to surveys and other quantitative research. This is simply not true, and it causes a lot of trouble for both marketers and marketing researchers. Focus groups and surveys are ideal partners

Groups can generate questions to be asked in surveys. They generate hypotheses that can be tested quantitatively. They help phrase survey questions properly, in language that really speaks to respondents. 

  • Jerry is a political campaign manager in a senate race. 

    He needs to market his client to the broadest possible audience and needs to avoid offending as many people as possible. He needs research into the public's opinion on issues facing the state. 

    He can conduct a survey asking how many people approve of this or disapprove of that, but he wants something more useful. He wants to know what kinds of things his candidate should be saying in response to these problems. 

    He needs to know what different points of view there are on the key issues, how many people hold each opinion, how firm their positions are, what might change them, and what kinds of programs seem appealing in response. 

    His first problem is that he doesn't really know how to ask questions about some of the issues. They are complicated, and he doesn't want to use language that the public will not understand. He also wants to be careful not to bias the responses by inadvertently using inappropriate language to describe them.

Focus groups can come to the rescue.  
  • He convenes several focus groups of frequent voters and opinion leaders (important, because who want to hear from people who won't play a role in the election). The groups discuss the issues in general and provide many insights into how they talk about things to each other, how the various issues interrelate, what kinds of differences of opinion there are, and how their concerns might be met. 

    Jerry then does a survey to find out how many people in the state feel each issue is important and how they feel it should be best addressed. His candidate emerges with a lot of live ammunition to use in the campaign.

Groups can also be used to probe issues that emerge in quantitative studies. 
  • Let's suppose Jerry's candidate is being hammered in the polls. His credibility goes way down, and surveys reveal that voters "don't relate well" to him. 

    Focus groups can pull together people who say they "don't relate" to the candidate. Under careful moderation, they can discuss what there is about Jerry's candidate that rubs them the wrong way. They can watch commercials and television appearances and point to key moments: 

    "There! When he shifts his eyes like that. You can tell he doesn't believe what he's saying." 

    "It always seems as if he's angry at the reporters. They're just doing their job." 

    "I don't like the way he crosses his legs." 

    All of those things are changeable. Since the numbers show that the problem is real and not just the cranky opinion a few picky people, it is probably worth Jerry's while to coach the candidate on public behavior. 

    For another example, an advertising agency sends three test television commercials out for audience response analysis. The agency's favorite commercial scores low on persuasiveness. Of course the client has no interest in producing the commercial since it has tested poorly, but the agency believes in the approach. Worse, the agency does not understand what to do in order to make a more persuasive commercial. 

    Focus group respondents can look at the three commercials and discuss them among themselves, going deeply into the reactions they have to various elements and sharing their views. The moderator can suggest ways in which the commercial might be altered to make it more persuasive, and the panel can give feedback on which changes might be effective.

It is fairly easy to see why the match of focus groups and surveys works well. It may not be quite so easy to see why using only one method has risks
  • If Jerry does a survey on the issues without doing the focus groups, he may not ask the right questions. OR he may not ask the questions in the right way. If questions are not very carefully phrased, they can bias answers. If some people misunderstand the question, even slightly, they may be providing incorrect answers. Later, when Jerry and his team look to the survey results for guidance, they may be misled. 

    If Jerry merely accepted the results of the other survey and didn't do groups to follow it up, he might conclude that the candidate should smile more, kiss more babies, or trot out his family. None of those things would work because the key problems would not have been identified. 

    If Jerry does the focus groups without the survey, he runs an even greater risk. The campaign may respond perfectly to a concern that is shared by only a few people who happened to come to the focus groups. The may waste precious media dollars that could be spent on issues that concern a much broader segment of the public. 

    If the candidate responds to one or two focus groups without any polls to indicate that he's in any kind of trouble, he might stop crossing his legs, start chatting with reporters, and stare directly into the camera at all times. In fact, these may not have been problems at all, and the wider public might lose their identification with him as his "personality" changes in public appearances. 

    "Why has he started sucking up to the press?"

For some reason, when surveys are proved wrong, people question the skill of the pollsters. When focus groups are misused, by not supporting them with quantitative data, groups get blamed for being "unscientific" or "misleading." Focus groups are quite "scientific." They are simply not a substitute for statistical sampling techniques. 

Surveys, by the way, are not all "science." There is considerable art to constructing appropriate questions and putting them in the appropriate order. There is also a great deal of intuition required for constructing the questionnaire and performing the analysis. 

These general rules should be helpful, but knowing what kind of research to perform for what kind of marketing issue is an art in and of itself. It's one of the main things a research consultant, familiar with various techniques and options, can provide. At least, you should be better prepared to approach a researcher, and you should have a better sense of why you need to. 

The following chart lays out some important differences between qualitative and quantitative research. Even if you forget everything else, however, please remember that they are partners, and not competitors. 

  Qualitative Quantitative
Main techniques for gathering data Focus groups and depth interviews Surveys and scientific samples
Kinds of questions answered Why? Through what thought process? In what way? In connection with what other behavior or thoughts? How many? How much? Where? How often?
Role of interviewer Critical: interviewer must think on feet and frame questions and probes in response to whatever respondents say. A highly trained professional is advisable. Important, but interviewers need only to be able to read scripts. They should not improvise. Minimally trained, responsible employees are suitable.
Questions asked Questions vary in order and phrasing from group to group and interview to interview. New questions are added, old ones dropped Should be (ideally) exactly the same for each interview. Order and phasing of questions carefully controlled.
Number of interviews Fewer interviews tending to last a longer time Many interviews in order to give a projectable scientific sample
Kinds of findings Develop hypotheses, gain insights, explore language options, refine concepts, flesh out numerical data, provide diagnostics on advertising copy Test hypotheses, prioritize factors, provide data for mathematical modeling and projections


ReSearch Guru



About AMIC

     North America
     Asia / Pacific
     Hong Kong
     The Netherlands
     United Kingdom
     South Africa

Media Guru
Ad Info
Ad Jobs
Ad News
Site & Ad Info
Industry Sites
AMIC is designed, built, and maintained by AMIC.COM Inc. a member of the Telmar Group of Companies.
Copyright © 1996 - 2018 AMIC.COM Inc.; all rights reserved
Reuse of the contents of this site without the express written permission of the Telmar Group is strictly prohibited