Improving Targeted Advertising Through Ad Recommendation & Client Insights

Purpose

  • Create an ad recommendation system for a fast-growing email marketing startup that improves revenue.
  • Develop pathfinding analysis and advising on future data collection and advertiser engagement strategies.

Challenge

Improve revenue through:

  • Increased ad clicks on the platform.
  • Identifying underperforming advertisers and ways to improve.
  • Directing data collection and organization for future platform improvements.

Solution

  1. Analyze historical data on user behavior to augment client data with relevant third-party information that identifies engagement trends based on user groups.
  2. Identify underperforming user groups and ad content, and advise client on specific platform improvements to address these issues.
  3. Develop a predictive engine that provides a most-likely-to-be-clicked ad for each user based on current ad inventory and user profiles.
  4. Highlight future data needs for improving prediction.

Business Impact

PowerInbox was empowered to:

  • Attract more advertisers by providing guidelines for successful engagement.
  • Improve ad revenue by 40% through use of predictive recommendation model.
  • Focus efforts on specific ad content, user groups, and device types lagging in performance.
  • Cost-effectively leverage third-party data to improve ad recommendation.
  • Know what type and quantity of data to collect to improve prediction as the platform grew.

Feedback

“We worked with Sev on two projects for our ad recommendation platform–a predictive model for ad recommendation and an analysis of user conversion behavior. The work included highlighting ways we could improve our ad recommendation and also areas our clients could focus on to improve their revenue generation on the platform. Sev worked effectively with our data science staff. We were very pleased with the work and would recommend working with him.” – Matt. T. – Founder and CPO, PowerInbox