Search Branding: A Problem of Metrics

First published November 13, 2008 in Mediapost’s Search Insider

This whole question of branding in search came about because of a rather fuzzy definition: what exactly is brand lift? How do you measure it?

This was the problem we wrestled with when we did the first search brand lift study for Google and Honda. Failing anything better, we did the standard tests of aided and unaided message and brand recall. I’m not a huge fan of these metrics, because I don’t think they adequately capture the neural basis of brand. But given the nature of the study (which included a survey and some eye tracking) it was our only feasible alternative.

What is Brand Lift?

Before I talk more about the metrics, let’s talk more about the concept of brand lift, as it’s central to the argument. What is brand lift? As measured by brand recall metrics, it’s awareness of the brand. But actually, it’s to see if a concept of the brand is lodged in working memory immediately after exposure to the brand.

So, let’s walk through this a bit. We create different search results page scenarios with variations of exposure in our test brand, including a control with no exposure. We create scenarios that set up a reasonably natural interaction and scanning of the results page. And, at some point after this interaction (usually immediately following) we ask participants if they remember seeing a brand on the results page. We start with an open-ended question (unaided) and then provide a randomized list of brands to see if they choose the test brand (aided). We then measure variance against the control, correlated with the nature of the exposure.

Given this framework, we did find brand lift. Significant brand lift. And in one aspect, this is great news for marketers. The fact that a brand remained lodged in working memory is a very big win for search in capturing consumer attention. I’m going to be talking about why this is so in the next column. If we equate brand lift with engaged attention and carving a (temporary) niche in the prefrontal cortex, the study proves there’s a strong connection with search.

Brand Lift is in the Eyes of the Beholder

But this is where the metrics get fuzzy. Brand lift seems to be many things to many people. This is why ARF decided that we needed a new metric and started down the road of defining engagement. But the problems that soon plagued this endeavor highlighted the inherent challenge. Our engagement with brands is not a one-size-fits-all, measurable occurrence. Brand relationships, like all relationships, are complex and shifting. There are many degrees. In additional, there’s a question of modality, based on context and intent. My relationship with a brand, say Apple, can be significantly different if I’m shopping for a new laptop for work than if I’m helping my daughter with an iTunes download. Advertisers want a single, simple quantifiable number that defines our brand relationship. Such a beast doesn’t exist.

So, how do you measure lift? What is lift? Is it a temporary lodging in working memory? Is it a long-term strengthening of the synaptic connections in long-term memory? Is it firing up the limbic systems that are our emotional gatekeepers, getting a lump in our throat when we see a tearjerker ad? Is it digging out our deepest primal drives when we see attractive women hanging around guys on a golf course, implying the beer they’re carrying around (but, of course, not consuming) is the reason the women are volunteering for caddy duties (Yeah, like that takes place in the real world)?

We were asked to prove brand lift on the search page, and we did, by one metric. It’s an important metric — a vital one, in fact. But when advertisers hear brand lift, they all hear different things. The definition of brand lift seems to be in the mind of the beholder.  Ironically, I’ve been reading a book that’s taken on the impossible task of explaining quantum mechanics to me. The mind numbing problem with quantum mechanics is that the physical nature of something isn’t defined until it’s observed. Until then, it’s a probability wave. I’m suspecting the same thing might be happening with brand metrics. They don’t take shape until someone tries to define them. There are just too many variables.

A Call for Consilience

The problem with branding is that marketers don’t know what they don’t know. They have learned marketing and the art of pushing a message out, but very, very few marketers have taken the time to understand what happens on the receiving end. They know nothing about cognition, emotional tagging, the formation of memories or any other workings of the human brain. Only now, with the consilience that is beginning to happen in the academic world, are we applying neuroscience to marketing.  So, marketers are desperately trying to apply a universal metric to something that largely defies measurement, and the marketers don’t even know why it’s not working.  They’re getting numbers they can plug into the dashboards, but no one is sure if they’re indicative of what happens in the real world, or, more accurately, our neural representation of the real world.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.