How Data Modeling & Machine Learning Helped Reduce Infrastructure Costs by 70%
After consolidating several live events from major surfing brands onto its own video streaming platform, WSL needed better answers to questions about viewership. How many fans engage regularly? How many tune in to multiple events? Who are the superfans — the ones who drive the majority of consumption and engagement— and which new users are most likely to become superfans?
We combined and cleaned data from Analytics 360, Facebook and WSL’s internal system into a single data lake and modeled superfan attributes using machine learning algorithms. We then applied early engagement attributes of superfans so WSL could focus its marketing on new users.
The fan insights WSL gained using Google Analytics 360 with Google BigQuery allowed the organization to reduce cost per conversion by more than 50%, which more than doubled the efficiency of ad spend aimed at driving fan engagement and content viewership.
By focusing on higher ROI customers, WSL drives greater fan engagement with live events and digital content—without increasing its marketing budget. When weather conditions deliver the best waves and large numbers of fans are looking at content, WSL can easily scale its analytics on demand. And during the offseason, it can scale down, which has allowed it to reduce infrastructure costs by 70%.
Now its team can ask more sophisticated questions of its data: How do wave conditions affect viewership? How do athletes’ scores correlate with other factors such as wave height, time of day, or level of competition? And what exactly is a superfan worth to the business?
"Moving our infrastructure to Google Cloud has allowed us to scale our use of resources to match traffic to our website, mobile and connected device apps. During our offseason, we reduced infrastructure costs by 70% by scaling our resources down."