On behalf of a Top 3 U.S. health insurance brand, Yieldmo’s data science team analyzed a recent campaign in search of two audiences—likely and unlikely converters. After scrutinizing which engagement metrics and data points influenced conversions, they made an important breakthrough in cracking the ideal algorithmic code for our massive dataset. Comparing these two audiences side-by-side revealed that those who were highly engaged with the ads were significantly more likely to click than those who were not.
Here’s how it worked:
1 — Yieldmo captures dozens of proprietary exposure and engagement metrics, along with environmental and contextual data. For the first time, this provides a multi-dimensional view of customers’ interaction with ads.
2 — Machine learning mines our vast dataset for combinations of metrics and data points that reveal a propensity to click, play, convert, etc.
3 — Each customer is given a proprietary Yieldmo Engagement Score™ based on exposure, engagement, and other conditions of their interaction with the ad.
4 — Users with a high score are isolated for re-targeting, while users with a low score are dropped from the campaign to prevent wasted spend.
These results put us solidly on the path to proving the true power of engagement—all the nuanced micro-interactions between humans and their phones that Yieldmo alone measures. To be sure, this test was just one critical step toward tapping into the actionable insights within our platform, but it was as powerful as it was promising.