How many is the ‘right’ number for a focus group?


How many in a group image

There seems to be some kind of time-honoured expectation or assumption that 8 respondents is the ‘standard’ number required to make up a focus group.  I wondered how and when this was established and after some asking around others in the industry, it seems that the magic number 8 has some foundation in logic, although there are others who claim it was decided as a purely arbitrary number that just somehow stuck.

One definition suggests that 8 is the largest number of people in a group which would allow it to function naturally as a single discussion without breaking into smaller groups, and allows the moderator to adopt a passive style and let the group run itself. Another suggestion is that the format, and the number 8, emanate from psychoanalytical practice (although no detail was available on this).

The accepted view among researchers tends to be that the optimum size of groups should be determined by the nature of the objectives, the people you are talking to and perhaps also the way in which the moderator is comfortable working. So some research groups may be best conducted with 8+ people making use of workshop-type techniques with breakout and regrouping sessions; others will be suited to smaller groups of 4-6 people where everyone can converse easily and inclusively.

The consensus however seems to settle at 6-7 as the optimum size for a focus group, allowing the best balance of breadth and depth of conversation. 8 thus perhaps refers to the optimum number of respondents to recruit, on the basis that there may be one or two no-shows resulting in a group of 6-7 as a norm.

Personally I have increasingly found that even smaller groups of 4-5 yield as much as anything larger, giving as they do the opportunity to explore issues in real depth without the need to ensure that everyone – including the quiet one in the corner – has had the chance to respond to any particular question. Obviously the sample size is decreased but I don’t ever feel the quality of the responses is decreased or the overall outcome affected, and clients should not be afraid to consider a less-than-8-per-group approach.