Back to Gartner for Marketers Blog

The Big Game of Measuring Brand Advertising

By Lizzy Foo Kune | January 26, 2018 | 0 Comments

MarketingAdvertisingBranding and Value PropositionData and Analytics

Marketers have long debated how to effectively measure brand advertising. So much, in fact, that we’ve published a few research notes on the topic, and we’ve even got a presentation discussing it at Gartner’s Digital Marketing Conference (Does It Work? Measuring Brand Health in a Performance-Driven World).

These conversations become more frequent as the Super Bowl nears, and it’s no wonder why. With 2017 viewership hitting 111.9 million, and each second of advertising costing $168,333, marketers are on the hook to demonstrate the value of brand advertising.

The challenge? There are a multitude of methodologies available to you, and it’s hard to figure out which approach is best. What’s more, there are a number of biases that could lead to incorrect interpretation, resulting in differences in how marketers may judge the “success” of their campaigns.

I was thinking about this when reading about Skittles’ approach to this year’s Super Bowl. Instead of airing a spot during the game, like it has in the past, it’s going to show its ad to one person and livestream that individual’s reaction to it on the brand’s Facebook page. Measuring your brand is easy, if you’ve got a focus group of 1.

Of course, Skittles will likely seek to measure the organic reach of content related to the ad, changes in perception, and perhaps an increase in brand search volume. But think about that one person – are they really representative of the whole?

It reminded me of a notorious poll from 2016’s presidential election.  It overweighted one respondent in the survey, who happened to be entirely unrepresentative of his demographic group. The methodology behind the poll made it prone to extreme variance.

What, then, should marketers consider when measuring brand advertising?

  • Sample size and representation. N=1 may be extreme, but most of the time, you’ll need a large, representative sample from which to draw conclusions. Focusing too much on a specific audience segment – or drawing conclusions from too small an audience – may lead to incorrect measurement.
  • Data from past campaigns will likely give you a frame of reference when interpreting your current data. If you lack history, start with industry or peer benchmarks until you’ve shaped your own.
  • Think long-term. Consider lifetime value or brand equity measures as ways to estimate the indirect value that brand advertising brings to your program.
  • Estimate and correlate. We all know the adage correlation doesn’t equal causation, but sometimes, that’s all we’ve got. Correlate brand search volume with in-market brand activity to identify lifts in perception.

There’s not a single right way to measure brand advertising. You’ll likely use a combination of techniques and technologies to figure out its value, especially since it’s only one component of a complex customer journey.

But, whether dropping $5 million on a Super Bowl ad or showing an ad to a focus group of 1, brand measurement requires rigor. A sound methodology will at least get you to halftime.

Leave a Comment