Sprocket’s analytical chops and human-centered approach helped Ragnar Relay better predict future race attendance to better align resources.
ABOUT RAGNAR RELAY
The idea to run an overnight 24+ hour running relay across the mountains of Utah was a lifelong dream of Steve Hill. The dream became a reality in 2004 when Steve, his son Dan and Dan’s college roommate, Tanner Bell, organized the first Wasatch Back Relay, spanning 188 miles from Logan to Park City, UT. Starting small, the Ragnar Relay Series has grown from a single relay in Utah to 40 events in 6 different countries.
As Ragnar’s portfolio has continued to grow across the US and International markets, it has developed a portfolio of races with various formats, regional footprints and customer base. A key input into financial planning is how many teams will actually sign up for future races so that they can align operations, marketing, and capital properly. To help with this, Ragnar asked Sprocket for help in predicting future race attendance.
We’re a collaborative bunch at Sprocket, and began with several co-creation sessions with the Ragnar team to list out what team Ragnar thought contributed to race attendance. Said another way, our predictive modeling projects start off in a very analog and human-centered way.
Once we had gathered team hypotheses, we proceeded to build descriptive analytics as well as rapid prototype models which were used as alignment and inquiry moments in the project. These moments allowed us to surface initial findings and refine our hypotheses. Several iterations and multiple models later we arrived at an optimal model fit on Ragnar’s past race data.
The combination of tight Ragnar/Sprocket collaboration, strong predictive modeling chops, and never-ending curiosity allowed us to deliver a predictive model that significantly improved Ragnar’s race attendance forecasting ability and will allow them to align resources more efficiently going forward.