Skittles is Mars Wrigley’s largest Fruity Confections brand, with a storied history of creative excellence. In 2020, the brand set out to create a new modern suite of evergreen content – built not only to deliver the trademark Skittles storytelling the brand had become famous for, but to do so in ways that were more meaningfully personal to its audience. The goal was for this library of new Skittles assets to be dynamic and responsive, tailored for specific contexts and individual audience behaviors, to maximize both relevance and impact.

As a brand, Skittles is all about a “welcome disruption to the predictable” – providing an enjoyable interruption to the mundane hamster wheel of everyday life.

As we approached the assignment for more personalized, data-driven content, we realized that it was ‘predictable’ in its own right.

Ad-targeting had become ubiquitous. But as it's ability to address people on an increasingly "personal" level grew, so too did its intrusiveness. Ads following us. Apps listening to us. Devices tracking us. It had all just become so… creepy.

To take the piss out of all this “creepvertising” in classic Skittles fashion, we couldn’t just create a bunch of targeted ads like everyone else. Instead, we would create the most self-aware targeted ads in the history of targeted ads. Calling attention to the creep-factor, even as we amplified it further.

To build our suite of targeted assets, we identified the top content genres to design for, based on the preferences and digital behaviors of our audience.

Across multiple key global markets, we aggregated platform data (YouTube, Facebook, TikTok) with audience understanding from Mars Wrigley consumer profiles and our own proprietary Omni audience graph, to generate a ranking of reach and relevance.

Armed with those specific contexts, we got our creepiness on. Creating content that would not only be relevant to but actually make reference of the digital behaviors that would be used to target it.

The result was a suite of 168 unique digital-first brand assets, each engineered to be completely self-aware of how it was targeting the viewer. We used ad sequencing to our advantage as well, creating additional assets that would “follow” previous viewers and amplify the self-aware storytelling even further.

As of April, 2021, the campaign is launching gradually around the world – recently going live in Russia and UK, with other key markets across North America, Europe, and Latin America to follow.

Initial assessment of the content shows it driving nearly 3x the average rate of emotional engagement (as measured via passive facial coding) vs. market benchmarks.