Business Impacts
80%
Model accuracy
Enhanced return on marketing spend
1000%
Conversion rate
Customer Key Facts
- Location : North America
- Industry : Advertising
Problem Context
Alphonso is a TV data company with the industry’s largest TV viewership footprint that focuses on advertising campaigns by targeting users through digital platforms. They have a Video On Demand platform for on-the-go digital generation that offers high quality, popular, localized video content in multiple languages for consumers across emerging markets.
Their Marketing team had a propensity model for lead scoring in which propensity scores are assigned to each user having higher tendency of watching a new TV show. However, the model performance was not satisfactory.
Challenges
- Classify each IP address into a single bucket group due to its multiple characteristics behavior
- Analyze voluminous data on real-time basis
- Multiple sources of data resulting in disparity in granularity
Technologies Used
Google Cloud Storage
Google Compute Engine
Google's BigQuery
VR Studio
Determining Propensity of Users to Watch a New TV Show Based on Historical Viewership
Solution
Quantiphi built a subscription propensity model which generates insights on users viewership behavior and accurately predicts the conversion rate of new users on their OTT platform. This helped the Marketing team better segment users and optimize the spend on advertisements by targeting only those users with a higher likelihood of converting into a premium user.
Result
- Targeting the right set of audience having higher propensity score to maximize campaign impact
- Optimizing the existing model led to an increase in accuracy from 8 percent to 80 percent
