Eggy Weggs

Remember the scene in "A Clockwork Orange" where the psychologist plays a word association game with Alex, showing him drawings and asking him to say whatever comes to mind? When I first saw that scene (many moons ago), I had two reactions: (1) cool technique! And (2) I'm no psychologist, but it sure sounds like Alex is screwed up again.

Which is how I tend to think about qualitative projective exercises. Namely: (1) they're cool to use and help me feel "creative," and (2) in and of themselves, they yield little more than broad generalities.

Projective exercises are supposed to tap into the underlying emotions that motivate consumer behavior. Quite frankly, I have my doubts about that--and I'll offer three techniques I've used in the past as examples:

1. Brand personification (if brand X were a person, what type of person would it be?): I've used this technique quite a bit, for many brands and product categories. But almost every time I use it, the results are so similar I could have written that section of the report in advance. One brand is old and stodgy and wears a business suit, another is young and hip and wears khakis. . . .

2. Photo sort (which photo best expresses how you feel using product X?): I've used this exercise a number of times, with results rarely rising above the level of cliché. "I picked the photo of the lion because I feel 'proud'. . . ."

3. Collage (before coming to the session, create a collage that expresses your feelings about a particular topic): I have less experience with this exercise, but in one example, where participants were asked to create a collage expressing what it's like to be a parent, one of the more noteworthy creations featured a picture of a "road winding through the mountains." Truly surprising and deep.

I realize there may have been something in the way I personally managed these exercises that led to such less-than-insightful results. However, the examples reasonably represent the two big concerns I have about projective techniques--namely:

1. They're by definition limiting. In a projective exercise, we're establishing the context and setting the ground rules for participants. You need to think of the brand as a person (not a car this time, or an animal, or a color, or a sound, etc.). You need to choose from this specific set of photos. You need to create the collage from magazines you happen to have available to you at home.

2. To come up with a narrow, limited description of staggeringly complex feelings and motivations, consumers often fall back into what's easy and expected: clichés and generalities that allow them to successfully complete this difficult exercise, in a way that meets the approval of the moderator and fellow participants.

To me, qualitative research is all about openness (talk to me about topic X in whatever terms and context make the most sense to you) and depth (I want to understand the nuances of how you personally feel). At best, projective exercises can serve as a bridge to get us there. But if they do, it's because the moderator is skilled in building rapport and asking the right probing questions after the exercise. If they don't, it's because the beguiling simplicity and consistency of the results (everyone thinks brand X is the "old guy") delude us into taking them at face value.

When it comes to using projective techniques, then, I like to put myself back in the psychologist's shoes from "A Clockwork Orange": Wow, this brand is really screwed up. Hopefully, I can use real moderator's skills to try and figure out why.