By Gilad Barash, vice president of analytics, Dstillery
Data can show us what we expect but, more importantly, it can uncover surprisingly valuable information. When it comes to audience data used in programmatic advertising, it’s the unexpected data that has the potential to uncover untapped audiences for brands.
At first glance, a recent analysis of the visitors to nbcolympics.com leading up to the 2022 Beijing Winter Olympics revealed what would be expected: Olympic spectators are interested in exercise, health and fitness.
But that doesn’t tell the whole story.
Further analysis revealed surprising digital pathways that audiences took before visiting the NBC Winter Olympics website. In the weeks leading up to the world’s greatest skiers, snowboarders and bobsledders competing on the global stage, there were the expected visits to websites on winter sports, gear and apparel. But in the days before, many sought information on a decidedly summertime item: luxury sneakers. The data also showed an unusual overlap in sports interests, such as a significant crossover between Winter Olympics fans and fans of sports such as the NFL and WWE.
While interesting, why do these unexpected details matter in digital advertising?
Audience analysis is able to provide valuable insights into target audiences out of seemingly unrelated audience segments. The goal of these analyses starts with getting a look into actionable insights to ensure focused targeting. In digital advertising, the more accurate the targeting, the greater the engagement with ads and the higher the chances of conversions as brands access their most relevant audiences.
Unearthing surprising audiences for more effective targeting
You might expect fans of winter sports to dominate the viewership for the Winter Games. However, analysis of nbcolympics.com data showed a surprisingly different story — interest from winter sports enthusiasts didn’t outrank the interest from fans of other sports.
In fact, the top interests were football, pro wrestling, golf, swimming and softball. There were no winter sports in the top 20 behavioral audiences at all.
Programmatic advertisers may have assumed winter sports enthusiasts comprise their best audiences, unknowingly overlooking the most relevant audiences altogether. Data is the antidote to assumption.
Tracing paths to the final data point
Analyzing the pathways that site visitors took to reach a website like nbcolympics.com also can tell a lot about an audience’s interests and behaviors.
Looking at these web behaviors and pathways, intriguing data shows that in the four weeks leading up to the Olympics, a common pathway to the NBC Olympics site began with luxury sneaker researchers and shoppers. While sneakers aren’t a common association with winter sports, data analysis shows that younger audiences who follow athletes on social media may be interested in emulating athlete’s fashion trends, underlining the influence of social media and celebrity-athlete culture.
Allowing the data to tell the story rather than assuming what the data will say can reveal online origin points that brought users to a site that may have been otherwise overlooked through data assumption.
Finding common interests for greater precision
Finally, finding common interests within a target audience can reveal unexpected overlaps.
Analysis of audience overlaps surprisingly shows that trip planners dominate the behavioral audiences, making up three of the top five audiences on nbcolympics.com. As the Olympics allows viewers to travel vicariously through the athletes competing, those interested in travel can immerse themselves in other cultures from the comfort of their couch. This insight is valuable to programmatic advertising as travel enthusiasts otherwise might not be equated with the Winter Olympics.
Other audiences with strong overlap to Winter Olympics enthusiasts included fantasy sports fans and stream services researchers, in the latter case suggesting that Olympics fans were looking for new ways to engage with the games.
Using data science to uncover and understand audience behaviors is what sets actionable, data-driven audiences apart from generalized targeting. As programmatic advertisers look to identify and reach their known, and unknown, audiences, they’ll need to continually look at how to use the best data science at their disposal to find the overlaps, unexpected correlations and audience pathways.
About the Author
Gilad Barash has spent the past 10 years extracting stories from data and translating them into data-driven decision intelligence tools to meet business goals. He has led teams of analysts in support of these tasks and gives talks about how to extract actionable insights from data and communicate them to stakeholders inside and outside the organization. Gilad hosts a monthly podcast called “Who’s Your Data?” in which he interviews thought leaders from different industries about how data affects life and how life affects data. Prior to working at Dstillery, Gilad held various roles at HP, including Research Engineer at HP Labs Israel.