Diverse Representation in Ads Is Not Enough to Win The “New Wave” of Diverse, Young Americans
Share on facebook
Facebook
Share on twitter
Twitter
Share on linkedin
LinkedIn

Across 2019, we analyzed almost 150 ads, gathering almost 100,000 surveys and 20 million datapoints. Using this data, we developed the Cultural Fluency Quotient, a new metric to predict brand favorability and purchase intent, and ran machine learning on the data to derive powerful new insights into what matters for every demographic.  Read on for critical insights into the creative strategy you need to win the New Wave.

Access the Presentation 

Keys to Culturally Fluent Creative for the New Wave

In a climate of increasing tribalism exacerbated by social media polarization, advertisers must appeal to the most complex mix of demographics in American history while steering clear of unintended backlash.  Every quarter has its walk of shame for one or more brands, most recently Peloton for its widely reviled holiday commercial “The Gift that Gives Back” that tanked the stock by over 10% in early December 2019.

As many members know, we have been building a capability we call AdRate leveraging a database of consumer response to ads. Across the last 18 months, we have been conducting research based on a new way of looking at brand favorability called Groundswell and Backlash, and applied machine learning to reveal powerful insights into how people from different cultural backgrounds process ads.

As the database grows, our ability to derive deeper insights and develop more predictive metrics increases.  For this study we developed the Cultural Fluency Quotient (CFQ), a weighted combination of three factors that best predict post-view brand favorability and purchase intent, which is then indexed for each demographic.  We ranked ads on CFQ for each demographic and ran machine learning on the top and bottom performing ads to derive the factors that best predict both high Cultural Fluency and what to avoid.

One key insight here is to go beyond performance norms.  We therefore also look at how important a norm is to high CFQ.  After all, it makes no sense to focus overly on how well an ad’s visuals perform (for example), if visuals are not a driver of cultural fluency.  For this reason, we use our machine learning results to derive importance scores an dozens of attributes of ads.  We then plot the results on a 2×2, as shown below.  The winning ads do well (horizontal axis) on what matters (vertical axis).

When we run the numbers, the findings are similar for every demographic. The best ads tell a simple story using ONE multicultural perspective, with attention to authentic texture.  These ads avoid the trap of representing every demographic at once, and ensure the viewer is not confused by the relationship between the product and the story.

The top two insights from this analysis imply:

  • It’s Not Just Casting: Creating common ground is not just “representation.”  You see that in the chart below that People & Characters are not as important as Story and Message. Diverse representation is necessary but it’s only price of entry.

  • The Story is Everything: Storytelling is by far the most impactful way to build cultural relevance. No story, no cultural fluency.
  •  

Few ads better exemplify this point than US Banks “Hard Work Works: Flying Home.”