We Love You Brand, But I Love Rival More

As I was speeding out of town two weeks ago to visit the Buchanan&associates advertising agency in Dublin, Ohio, I ran some quick analyses on our brand physiology data.

In this study, we measured psychophysiological responses to brand logos. In previous studies, we had found that self-reported responses to brands look a lot like self-reported responses to other emotional stimuli.

Admittedly, the early August analyses were fast and incomplete. But some of the results looked weird to me.

This first round of analyses compared physiological responses to the most-loved brands and the least-loved brands.

Collectively, we had both. There were brands that seemingly everyone loved and those that seemingly everyone did not love.

When I first saw the "weird" data, I had a thought. Maybe, I thought, we have a bit of a groupthink problem. Sure, the aggregate data suggest that the brands are loved. But some of the people among our 54 participants might not have loved those brands.

Let's take a look at that idea. For Texas Tech students, our five most-loved brands were Disney, Starbucks, Google, Dr Pepper, and Target. They stood alone. We called them Lovemarks after Kevin Roberts' term.

But let's look at the results by person. Every brand had a complement. That is, there was another brand roughly in the product category. When comparing attitudes (self-reported answers to 3 scale items), how many people picked the Lovemark? (Ties are excluded).

Disney had 32 fans, but 16 people preferred 20th Century Fox.
Starbucks had 48 fans, but 5 people preferred Maxwell House.
Google had 32 fans, but 11 preferred Yahoo!.
Dr Pepper had 33 fans, but 19 people preferred Coca-Cola.
And finally, Target had 39 fans, but 11 preferred Wal-Mart.

The majority always wins. However, with the exception of no-real-equal Starbucks, there were no runaway landslides.

You see, the group does matter. The herd defines the trail. But the individual matters a lot, too. Just because most of you love something does not mean that I do. I might like it, but perhaps I love a rival more.

We already had planned to measure responses to individual Lovemarks. In fact, it was a driving motivation behind the entire study. These data just confirm that it is the correct strategy.

If data show that a group feels a certain way, then a (quasi) randomly selected individual from that group is likely to feel that way, too. But that participant has lived her or his own life and has unique attitudes. We'll do a lot better if we give that individual some agency in measuring her or his attitudes.