This paper focuses on the research industry's adherence to the group d
iscussion in qualitative research, its validity for consumer research
in the late 1990s and its optimal implementation and practice. The aut
hors examine the thesis that the responses obtained from the consumer
are not as accurate as they could be and suggest that the input into,
and methodology of, conventional moderated groups may, in some instanc
es, be inappropriate and out of date. A number of consumer-driven, DIY
groups were carried out, without a moderator, in the IT and financial
service sectors. The unmoderated groups were given clear instructions
regarding the process and content of their tasks. Key issues are comp
ared for moderated and DIY groups. It is concluded that unmoderated gr
oups are likely to work best in mature sophisticated, post-modernist s
ocieties which are ready to, and experienced in, challenging the statu
s quo. Unmoderated groups allow a much clearer understanding of the ag
enda as set by respondents. The unmoderated approach has particular va
lue in exploratory and explanatory research, as well as the initial ph
ases of NPD work, whereas well-moderated groups are more appropriate f
or diagnostic, tactical and secondary stage NPD work.