How to Cut-Loose a Bad Research Respondent
They say that “One bad apple can spoil the bunch.” This is absolutely true in a focus group setting. A bad research respondent — I mean toxic or unsafe, not quiet — can ruin an entire group. This is another frustrating, but rare, limitation to focus groups. If you are new to market research, perhaps you have not yet met a “bad respondent.” You will. After reading this post you will have a better strategy for how to handle him or her. Let me share with you my worst experience with a bad research respondent in a focus group and how we resolved the issue successfully.
Chicago Focus Groups — Enter the Bad Respondent:
One of my fellow researchers asked me to join her for a focus group project she was doing in Chicago. I’m not a huge fan of focus groups, but I do like Chicago, so I obliged the request. The groups were fairly typical until our bad respondent arrived. It was obvious that she was totally inebriated. She had slurred speech and extremely blood-shot eyes. Anyone can be sympathetic to allergies, but this woman was totally high. If it was obvious to us in the backroom that this woman intoxicated, it was even more obvious to the other respondents. We had to extract her and do it in a way that was both polite and would not disrupt the whole session.
The Long Game — Working-up to Cutting the Bad Respondent:
Very quickly we devised a plan. As all the respondents were finishing introducing themselves, we sent a member of the team into the interview room. She announced that over the course of the group we might select certain members to join us for 1:1 conversations if they said something particularly interesting. She singled-out one of the more jovial men in the group to be our first 1:1 and she took him out for 5 minutes to a separate room in the facility for an interview.
When she returned, she took a second respondent and did the same thing. The moderator caught-on to what we were doing and tried to keep the bad respondent out of the discussion as best as possible. After the second “1:1 interview,” we asked the bad respondent to join us. A few of us on the research team escorted her to the door, paid her in full, and sent her on her way. She protested, but did not get violent — especially when she heard she was getting paid.
Why We Handled Our Bad Respondent This Way:
First, we didn’t want to cause a scene. We needed to get her out of there, but not wreck the mood of the rest of the group. Everyone in that room knew something was up with the bad respondent, but you never know how people might react to seeing someone escorted out of a room — peacefully or otherwise. It could have ruined the group. We wanted to tread lightly.
Second, we didn’t know how belligerent the bad respondent might become. By creating a natural pattern of removing people from the group the bad respondent was meant to feel less singled-out. When she was called to leave, she was happy to join. Paying her prevented any further retaliation.
Third, if for some reason the other people in the focus group didn’t realize how messed-up the bad respondent was, maybe it made them participate more hoping to get selected for a “special 1:1.”
The end result was actually a smooth focus group. The sessions were two hours in length, so in the scheme of things the time the two other respondents missed was not significant. They merged back into the group and perhaps thought themselves very special for being called-out for their unique perspectives.
If you ever find yourself in a situation where a respondent is not fit to participate, do not allow them to do so. If, somehow, they find their way into your research, treat them with dignity. Do not single them out or embarrass them. Find a way to extract them, compensate them, and move on. The $100 incentive is already a sunk cost. It’s much cheaper that causing a scene or risking someone’s safety.
So how about you? Have you ever had a bad respondent? Any good or bad stories to share? Advice for our younger readers on how you handled the situation? If so, feel free to leave it in the comments below.