Use Case Transportation
Purchase Propension model
In the airline industry, targeted marketing is essential for increasing ticket sales and enhancing customer engagement. The effectiveness of these marketing efforts relies on accurately predicting customer behavior and purchase intent.
The system uses advanced machine learning to analyze digital interactions from Google Analytics and comprehensive CRM data, improving the accuracy of purchase propensity predictions. Personalized marketing campaigns based on these predictions aim to convert high-propensity customers, optimizing marketing spend and boosting revenue.
Challenges
The project faces challenges in data integration due to inconsistencies in email usage and device sharing, and in capturing potential buyers, aiming to identify prospects even with false positives to target those near purchase.
Data integration
Merging Google Analytics with CRM data is complex due to varied email usage and device sharing, complicating user identification.
Capturing potential buyers
Balancing the need to identify all potential buyers, even at the cost of including many false positives. The objective is to target customers who are close to making a purchase, ensuring that marketing efforts can effectively influence their decision.
Solution
A CatBoost model has been implemented to predict the purchase probability of Iberia Airlines customers. This system integrates data from Google Analytics and CRM, providing a comprehensive view of customer behavior.
The CatBoost model predicts the likelihood of a purchase within the next 60 days using advanced machine learning techniques. This approach prioritizes recall over precision to ensure all potential buyers are captured, optimizing marketing efforts to increase conversions and revenue.
Tech stack
Results
High Recall: Enables effective targeting of customers with high purchase propensity, significantly improving marketing efforts by providing the final push needed for these customers to complete their purchases.
70%
recall
