By Jeffrey Lee
Earlier this year, I was shopping for some workout gear on Nike.com/cn when I stumbled onto a familiar sight for me – my very own face. Nike had reached out to me after I had participated in one of their community-driven “Mile Dash” races in Shanghai. They snapped a photo of me during a sweaty moment in my 800m relay, and asked if they could use it. Lo and behold, and a few signatures later, I find myself (and my bright yellow short shorts) at the top of Nike’s homepage. I also visited Nike’s official social media accounts (e.g. WeChat, Weibo), where their marketing team had also promoted the race photo.
Me on the left, in yellow shorts.
Let me add a disclaimer, which goes without saying for anyone who knows me. I’m not a professional runner, nor a model. But I do happen to be a marketing professor who studies social media. And so I was especially curious about the marketing strategy, as I soon learned that my experience was not unique and that other everyday people were also being promoted on Nike’s owned media. What was Nike trying to accomplish through this strategy? A senior manager at Nike whom I spoke to suggested that “we like to highlight the community whenever we can.” But what exactly was the potential outcome of this strategy?
Nike’s Weibo account, and me again in yellow shorts!
Some new work from my data science team at NYU Shanghai (co-led with Professor Enric Junqué de Fortuny, and with Kehang Zhang ’19, Junhai Ma ’21, and Frank Zhou ‘21) provides a few answers. We asked nearly four thousand consumers online about the types of people who come to mind when they think of brands like Nike, Lululemon, etc. For some of these consumers, we also showed them social media posts featuring everyday people in their Nikes, etc. We subsequently allowed these consumers to write whatever nouns and adjectives came to mind for these brands. Using a variety of state-of-the-art Natural Language Processing techniques, we were able to analyze how brand perceptions subtly changed as consumers viewed these social media posts.
Figure of possible Nike descriptions in a semantic space.
What we learned was that these photos and posts of everyday people can indeed change brand perceptions, specifically about the stereotypical consumer of the brand. Normally, in marketing we tend to study the effect of celebrity ads on consumer brand perceptions. But through these social media posts of everyday people, we found it is quite possible to strengthen existing brand associations (e.g. believing that Nike is a brand for athletes) or to weaken them too (e.g. believing that Nike is for many different types of people). For a brand like Nike, a community-based strategy involving posts and re-posts of everyday people might broaden associations and allow the company to appear more ubiquitous relative to their competition. We show in our work that the broadening of such associations can be tracked using novel data science approaches.
Tracking the people who come to mind for a brand might be very important. For instance, take the NYU brand. What type of person comes to mind when we think about NYU? Do we think of specific types of professors (e.g. “energetic”, “quirky”) or certain types of students (e.g. “hardworking”, “creative”)? Does it matter whether the audience is internal (e.g. from the perspective of the NYU community) or external (e.g. from the perspective of rising high school seniors deciding where to apply)?
Our data science team is continuing to answer these types of questions as we seek to understand stereotypes relating to brands and groups of people, and how they might be reshaped due to social media. For now, one thing is certain: I will dress much better the next time I go to a Nike race… …and so should you!