Giving Rideshare Experiences a Lyft: Using Bayesian Network Modeling to Improve Rides for Drivers and Passengers
Praveen Sharma , Vice President, Marketing Science, Data & Analytics, Material
Abstract
What does it take to ensure rideshare experiences are positive, safe and worthy of five-star ratings for both passengers and drivers? We brought Bayesian network modeling to bear on the complex mix of factors that make up Lyft’s rideshare experiences. In two distinct studies – one focused on riders, the other on drivers – we mapped the attributes of ride experiences and uncovered relationships between them to reveal key areas for strategic intervention that traditional analytical methods might have missed.
Attendees will learn
- How to use Bayesian network modeling to structure and make sense of the complex, interconnected components of experiences
- How this data-guided approach helped Lyft prioritize improvements for both riders and drivers
- How to apply Bayesian methods to strategic roadmap development beyond traditional correlation or regression techniques
About the Presenter
Praveen is a dynamic modeler, consultant and senior leader within Material’s Data & Analytics group. With over 20 years of expertise in advanced analytics, including Bayesian Network analysis, choice modeling, segmentation, structural equation modeling and other statistical analysis for marketing research, he excels at tackling complex modeling challenges and delivering compelling recommendations to top executives. Praveen completed the Post Graduate Program from IIM Ahmedabad, a prestigious business school, and earned a Master’s in Marketing Research from UGA.