Focus Groups, Written By B2B International
When, how and where to set them up. How to run them
FOCUS GROUPS have been refined over many years by consumer researchers, who in turn borrowed the tool from the social sciences. They comprise a small number of carefully selected people who, under the guidance of a leader, discuss a subject. In industrial markets, researchers may have to accept that it is logistically too difficult to bring together respondents who are widely scattered. For this reason depth interviews are and will remain (until multiple verbal/visual link ups become feasible) the most widely used qualitative research technique. There are, however, some occasions when focus groups are the preferred method.
Table 1: Factors which determine the choice of groups or depth interviews
Group discussions are favoured where there is a:
concept which needs exploring with the stimulus of other parties
problem requiring analysis from basics
complicated buying procedures
complex interaction of factors influencing motives
Depth interviews may be necessary if there is a:
widespread and small sample
sensitive subject (e.g. requiring disclosure of production techniques or identification of customers)
need to protect the respondent from possible bias introduced by others
need to formally control the sequence of questioning
need to obtain results quickly
the possibility the respondent will have to refer to files to answer questions
Group discussions are especially useful techniques for researching new products, testing new concepts or determining "what would happen if…" They work because of the interaction between the group members. Individuals are not under pressure to give spontaneous answers. They can digest the points raised by other members and, as they consider the implications of issues raised, ideas may be sparked off which would remain untapped in a personal interview.
The number and composition of groups. Between six and eight members normally constitute a group though there are no rules as to the ideal number. Ten or 12 must be considered the absolute maximum because anything larger would prevent each member from making a significant contribution to the discussion. Further, in a large group there is a greater chance that some participants will treat it as a stage to remonstrate and dramatise rather than be constructive in their arguments.
As few as three to five group members can still be effective since even with this small number there is sufficient scope for the cross fertilisation of ideas.
Nor are there hard and fast rules for deciding how many group discussions should be held. One may be enough though two is the preferred minimum, to obtain a better feel and counter the possibility of a rogue response. Even with two groups, a disparity of views would raise doubts as to which was correct. It is therefore advisable to conduct three or four groups, while the benefit of more than (say) eight is questionable. A large number of groups is justified only if it is thought necessary to hold separate discussions with different classes of respondent (eg respondents from large companies as opposed to small ones; Northern companies as against Southern; users and lapsed users, etc) .
Venues and timing of groups. Whereas consumer groups tend to be held in a viewing centre, industrial groups are usually convened in a hotel. A hotel venue is more businesslike and suited to the expectations of a person attending on behalf of his company. The venue must be easily accessible to all, preferably well known in the area and with good car parking facilities. The memory lingers of holding a group of electrical contractors at the Mitre at Hampton Court and losing valuable respondents because the car park was full and off-street parking was restricted for a good half mile all around.
The room in which the group is to be held should be inspected in advance. It needs to be small and intimate. It should also be free from traffic noise or piped music.
Lunch times are a favourite time for holding industrial groups. However, the combination of travelling time to the venue and the duration of the group itself takes at least two hours and puts pressure on busy schedules. Early evening (say 6 pm), straight from work, is an alternative but risks incursion into people's own time.
Getting people to attend. Because group discussions rely on such a small number of respondents, it is essential that care is taken to recruit the right profile. The ideal-profile must be decided at an early stage and might include any of the following criteria:
job function of status
awareness of a product or company
user or past user of a product or company
represent a company of a certain size (or buying volume)
A list is constructed of companies for interview, and recruitment begins by telephone. A recruitment questionnaire provides a screen to eliminate respondents that do not meet the criteria and collects advance data on basic questions.
THE SUCCESS of a group discussion is dependent on meeting the target attendance rate. The prospect of the group discussion should therefore be made to sound inviting, and barriers which could prevent attendance must be removed. Techniques which can be employed to achieve the maximum attendance level are listed below.
Creating positive reasons for attendance. Refer to the group as "a meeting of people like yourself from other companies".
Sound enthusiastic. Whip up the respondent's own enthusiasm so that he feels important. Make the group sound exciting - make the respondent want to attend.
Tantalise the respondent. Give a hook calculated to draw them in. "We will be asking you to comment on a new truck". "We will be discussing the way other people draw their customers in". (Of course the hook must be realistic otherwise dissonance will occur if the group fails to live up to expectations.)
Provide a material incentive. Drinks, a buffet, and £50 in notes will help draw them in.
Removing barriers preventing attendance. Fix the time and date to suit. If groups are being held on different dates and times of the day, slot the respondent into the one most convenient for them.
Arrange to pick them up. If transport to and from the group is difficult, arrange a taxi or pick up service. It makes the respondent feel important and the cost is insignificant.
Reassure. Respondents may believe that the exercise is a ploy to sell something; or may fear ignorance will be exposed and they could be made to look foolish.
After the initial approach has been made and a promise of attendance has been obtained, a letter should be sent confirming the arrangements. The letter acts formally to remind the respondent of the date, the time and the venue and verifies the purpose of the project. The content and construction of the letter must be businesslike, fully descriptive and emphasise the importance of attendance. A map showing the location of the venue should be enclosed.
One or two days before the group, respondents should be telephoned to remind them of the event and also to provide the researchers with view as to the attendance level. Even with all these precautions, only around a half to two thirds of respondents who promised to attend will actually do so . The researcher will not know until the time of the group whether there has been overkill with too many respondents or too few to generate a good discussion. The weather, football matches and television programmes cannot always be foreseen and can play havoc with response rates.
In the event of a very high turnout the researcher must decide whether it is possible to cope with a group of fourteen or fifteen respondents or if some should be sent away with their incentive. It may be worthwhile pre-emptying the possibility of a high response and arranging for two researchers to be present so that two groups can be run – accommodation permitting.
Hostessing the group. The hotel staff should be fully briefed to direct respondents to the room where the group is being held. A notice of welcome in the foyer and signs to guide respondents to the room should be arranged.
Some respondents always arrive early and others late. The researcher may be advised to keep away from the group during the period of assembly as natural curiosity on the part of those respondents who have arrived may trigger a premature discussion. An assistant or hostess should be available to welcome respondents, explain the procedure, serve food and drink and engage in small talk.
The assistant will also prove useful during the course of the group, changing tapes, passing around display products helping with seating arrangements.
After the group the assistant can give out the incentive (for which respondents should sign) and encourage those who appear to want to chit chat all night to retire to the bar rather than sit talking in the room.
Administering the group. Groups are led by a researcher whose role differs greatly from that of an interviewer. The group leader's role is:
To steer the discussion through a range of topics which are relevant to the problem. The order of these topics is determined as much by spontaneity within the group itself.
As a catalyst to provoke responses or introduce ideas. Sometimes the researcher should play devil's advocate or feign ignorance.
To draw a response from those who are quiet and curb those who attempt to monopolise.
A group leader does not question individuals as in the conventional interview. Empathy must be created with the members, relaxing them and getting a lively discussion underway. A brief introduction should be followed by an explanation that a tape recording will be made in view of the difficulties of note taking. It is then necessary to break the ice by asking each member to introduce themselves and their company.
Working from the list of topics, the researcher moves the discussion from the broad to the particular. Thus, if the objective of the group is to establish the opportunities for a new design of fastener, the discussion would begin with a debate of fastening problems in general followed by an exploration of attitudes to fasteners and finally homing in on the new fastener itself.
Within a small number of respondents there invariably is a dominant personality who may attempt to run the group or whose views colour those of the other members. Equally there may be slow thinkers, introverts, wits, drunks, compulsive talkers and the indifferent to deal with. The ability to be able to bring out the best from each without insulting or embarrassing anyone, requires both authority and tact.
Groups generally take between 60 and 90 minutes to administer, depending on the complexity of the subject and interruptions from films or product presentations.
B2B International Ltd
14 Ack Lane East
Tel: +44 (0) 161 440 6000
Fax: +44 (0) 161 440 6006