Background
- A mid-sized casino the mid-west was experiencing poor repeat visits of first time visitors coupled with increased churn of their repeat customers
Approach
- Score all new guests following their first visit - What is their likelihood of returning 2+ times within 12 months and spending at least $100?
- Each new customer has a scores of their likelihood to repeat visits
- Developed another model to score customer churn probability
- Each repeat customer has churn probability score
- Developed both models based on logistic regression
- Used model drivers to recommend media, offer and creative
Results
- First time Visitors’ retrial increased by 20%
- Reduced customer churn by 15%
- The above two aspects contribute to $750K/year in incremental revenue