Recreation Customer Segmentation: An advanced data clustering AI solution for customer segmentation

The Problem

To effectively tailor products and marketing campaigns, the City of Edmonton needed to overcome a lack of clarity regarding the demographic and behavioral characteristics of its active recreation members. The reliance on traditional, demographic-only segmentation prevented a clear understanding of whether existing membership programs and price structures aligned with actual customer needs. Without a data-driven, behavioral segmentation framework, recreation managers cannot identify distinct, high-potential user groups or optimize the allocation of municipal marketing and operational resources.

The Solution

To optimize recreation services, the City of Edmonton implemented a customer segmentation analysis that replaces traditional demographics with advanced data clustering. Utilizing hierarchical density-based clustering algorithms, the system analyzed facility attendance, visit frequency and amenity preferences. This methodology successfully categorized the user base into six distinct behavioral personas. By providing a deep understanding of shared customer traits, this analytical framework allows the City to align membership pricing with actual user needs, execute highly targeted marketing campaigns, and strategically deploy municipal resources toward the most impactful and profitable user groups. Built using open-source tools and existing budget.

The Impact

By mapping precise facility attendance patterns, visit frequencies and amenity preferences, the analysis successfully shifted the City away from generic, demographic-only marketing toward high-precision, cost-effective campaigns. Operationally, this empowers recreation managers to align facility programming and staffing with actual usage trends, modernize membership pricing to match customer needs and strategically deploy City resources toward the highest-potential user groups.

Lessons Learned

Methodologically, the project proved that advanced behavioral clustering outperforms traditional demographic grouping, though it requires a single source of truth for data integrity and continuous updates to remain accurate.

Launch year: 2022