People always wonder if blondes or brunettes have more fun, but how often do you ask which color hair performs better for brands in their social media marketing? On top of that, where do other hair colors come into play?
We thought it would be fun to take a deeper dive into our own data to answer this question. To do this, we analyzed the content our platform collected on behalf of beauty brands. Specifically, we looked for assets that were tagged with blonde, brunette, or dyed hair (hair tagged as pink or purple) to see how each hair color impacted performance. We defined performance to be measured by click-through rate (CTR) and shopping clicks (see definition below).
Full disclosure before reading this, I have brown hair, but I can assure you that this study was conducted and analyzed by our business intelligence team to keep it as unbiased as possible.
Even before analyzing performance, most of the content that we collected corresponded to brunettes, which makes sense given that only 2% of the world’s population is naturally blonde.
In our research, we saw that content in the beauty vertical containing brunette hair tended to have better overall CTRs and shopping click rates, with dyed hair images coming in third place for performance. For the most part, this trend continued even when we broke beauty into smaller subsets like cosmetics or hair care. For example, when separating out the beauty category by cosmetics, we saw a similar result: Content with dyed hair tended to perform worse than both blonde and brunette hair. However, when observing the hair care category we didn’t see a significant change in performance between natural and unnatural hair colors.
While we can’t say we have all the answers from this study, we can use this data to make some educated inferences. For example, we may see a higher performance for brunettes versus blondes or dyed hair because people better connect to content that resonates with them, and dyed hair lends itself less to the authenticity factor. Or, the higher overall population of brunettes means that more of them relate to images that include brunettes versus other hair colors.
Similarly, dyed hair may not see a dip in performance in the hair care subset because consumers are expecting to see hair depicted in unique ways, whether that be different styles or colors.
Whichever way you slice the results, the key takeaway from this research is that it’s important to always keep your audience’s preferences in mind and to use internal data to understand what resonates best with them. Remember, too, that your customers can be broken out into several smaller groups who may react differently to various types of content.
By using the data you have at hand, paired with observing consumer behavior, your brand can better understand what type of user-generated content will have a more powerful impact on overall performance, and in turn your audience as well.
Methodology & Definitions
Definition of shopping clicks: The number of clicks through a widget’s shopping button leading to a product description or property page.
Definition of CTR: The number of clicks on an image to open the lightbox divided by the number of that image’s widget impressions.
Widgets: The pieces of code Olapic uses to display content on brands’ websites.
Methodology: Content used as part of this study was collected from March 8, 2017 – April 10, 2017, had a minimum of 50 widget views and 10 light box views, and are all in the Beauty vertical. Photos were classified as blond, brunette, or hair dyed pink or purple based on metadata from Google Vision.
Confidence Test: All of these findings were verified using Wilcoxon rank-sum and Kolmogorov-Smirnov tests.