Depending on your topic and the type of analysis you used, it might be that you wanted to focus on individual accounts or get the kind of detailed account you get from speaking to people individually? Or you might have thought the results would have been affected by using groups, eg people behave differently in groups and some people might have been less likely to speak up?