Viralspace Plan drives 48% reduction in Cost Per Acquisition for Philips, a Fortune 500 consumer electronics company
At Viralspace, we help marketers drive make data-driven decisions about their ad creative using AI. In early 2020 we made our first module, Viralspace Plan, available to our current list of highly curated clients. Today, we’re excited to share some proven results.
In February 2020, the Philips Sonicare team prepared for a comprehensive quarterly creative refresh. Our friends at Ogilvy, the digital agency partner for Philips, had been looking for a way to bring AI-driven creative optimization to their client. They decided to use Viralspace Plan to cut through the guesswork of choosing between thousands of feature options and write a data-backed creative brief.
Benjamin Snyers, the Head of Social at Ogilvy USA and Managing Director of Social.Lab, is convinced that creative optimization is the final frontier in today’s digital advertising world where bidding and targeting are solved problems.
“We believe the place where we can have the most lift is on the creative side. That’s why we decided to invest in AI-driven creative optimization with Viralspace.”
— Benjamin Snyers, Managing Director, Head of Social at Ogilvy USA
We partnered with Ogilvy to build a custom machine learning model based on their data. From this model, we pulled the top factors that drive performance for Philips Sonicare on Facebook and Instagram ads, choosing the best out of tens of thousands of feature options. These insights were displayed in a self-serve format on Viralspace Plan, which allows users to filter to specific time ranges, target audiences, and product categories. Ogilvy quickly extracted insights on their best backgrounds, messaging, compositions, product colors, video pacing, and more, and brought this data into their creative brief.
The Ogilvy team then leveraged the insights to focus on creative elements that drove the lowest Cost Per Acquisition (Leads and Add to Carts) in their Q2 campaign. This not only drastically reduced the time spent on creative analysis but also drove a 48% reduction in cost and 20x increase in conversion rate (purchases / link clicks) compared to the previous quarter.
These results were made possible with our highly customized approach to machine learning model training. For every client, we build a totally custom model based on client’s brand data. We careful stick to their brand voice and tailor our recommendations to their unique audiences. This part of the process is extremely important as we believe that no two brands are exactly alike.