This is an article I just came across in the recent edition of Quirk’s magazine:

“TRC Market Research, Philadelphia, has launched Bracket, a market research technique designed to aid marketers in new product development. Bracket aims to offer a method for prioritizing choices using a tournament-style structure that eliminates losing features to make the task more engaging and cognitively challenging. Bracket begins by grouping choices into sets. Respondents then choose their favorites from each set and these favorites progress onto the next round where they’re once again grouped (bracketed) and evaluated. Ultimately each person is given a final set of choices representing their unique best-of-the-best and is asked one last time which they most prefer. Bracket is designed to handle a large number (upward of 50) of features, benefits or positioning statements without inducing respondent fatigue or sacrificing data quality and analytic power.”

I thought it was an interesting take on market research, in an attempt to pare down features or benefits and compare the importance of each.  The bracket is designed to take two choices, similar to conjoint and pair each head-to-head.  The participant is forced to choose one over the other and move to the next round.  Although it could get very tedious if the choices are upward of 50 like the article says, it’s still a creative technique to study consumer choice.

For example, you can use the bracket technique to analyze features considered during a car purchase.  You can understand commonality in choices of ‘sunroof’ vs. ‘satellite radio’ or ‘heated seats’ vs. ‘DVD player’ and so on.  It reminds me of a similar technique I’ve used while moderating focus groups where I work with the group participants to force them into a choice of their top five features of a service or product.  Although they may have mentioned a list of 20, we attempt to come to a group consensus on the top five.  From there I used a food-chain exercise to rank order the top five from one to five, using labeled visuals.  This creates a nice discussion and gets participants speaking about what goes into their ranking system and why one feature is more important than another.  Visuals and ranking systems like this create good dialogue in market research.

As a result, I would gravitate towards using this bracket analysis in qualitative research more so than quantitative.  Other than asking participants to go through their round of 64 (or 68 now a’days), a focus group could eliminate those non-critical features and move the group directly into the Elite Eight or Final Four.  Think of it as the “Tom Izzo” jump.  Another issue with features would be the ranking or rating system itself to set up the bracket (1 vs. 16, 2 vs. 15, etc.)  If you randomize the features, you may end up with the two best features squaring off in the first round, with one being eliminated too early, which would severely compromise the results (and ruined Syracuse’s 1991 season).  So some type of ranking or weighting model would be needed to be built into the seedings or the researcher would almost have to set it up as a conjoint with a ton of different versions of brackets.  So when you have 64 features, you can see where that would get a little out-of-hand in quantitative studies.

Nonetheless, I do like creative thinking in market research where researchers try to reinvent old techniques. This bracket design is definitely something the RMS Analytics team will consider in our next focus group project as I think it be an engaging exercise for participants.  What do you think of this type of bracket structure in market research?

Research & Marketing Strategies (RMS) is a market research firm in Syracuse NY.