Churn Prediction for Gym Customer Retention
Overview
According to a Harvard Business Review study, acquiring a new customer is 5-25 times more expensive than retaining an existing one.
Many gyms, health clubs, and leisure centers adopt a simple strategy aimed at attracting new customers of which almost half of them will refuse their membership in the first year. Retaining members remains a huge challenge and this is probably the holy grail of this industry.
Challenge
Churn is a common problem in small and medium-sized gyms. One of the largest Portuguese companies in this business was losing many customers in the first months of activity. On average, people left the gym in the second month of enrollment.
This massive annual turnover jeopardized the implementation of other important emerging trends such as flexible, low-commitment offers, and pay-as-you-go models, which are only possible with ongoing active members
Lisbon,
Portugal
Opportunity
Some of the reasons for the churn are already well known in the industry (falling attendance rate, demographic characteristics, and overcrowding are just some of them), but the introduction of these new trends increases the complexity of the member’s universe resulting in more personalized data.
This data conceals many hidden features that translate emergent complex behaviors. Those features can only be analyzed and regularly monitored using involved algorithms, and that’s where AI can be game-changing building a 360º predictive churn model for your business.
Impact
In this project, we implemented a model for churn prediction based on machine learning algorithms considered state of the art. We were able to drastically reduce the dropout rate as well as track the main reasons associated with this phenomenon. This enabled the company to establish new opportunities for attracting customers, such as group classes and flexible schedules.