NBA Team Data Engineering

Our client has an in-house business analytics department and partners with a third-party vendor who manages online ticket sales and provides their mobile ticketing app.

Challenge

The Team’s management staff makes frequent and critical decisions regarding ticket pricing, promotions, and availability based on analysis of internal and online ticket sales.  

Their online ticketing vendor’s data pipeline was slow and prone to breakdown, especially when the vendor made changes to its data structure, which happened frequently.  These failures impacted the inability of senior leadership to make strategic decisions about price adjustments and promotional campaigns prior to games.

While the organization is well-funded and capable in many ways, their analytics team is focused on business intelligence and executive-level decision-making. They lacked the internal capacity to build and manage the pipeline they needed, and they could not justify the business need to hire a full-time data engineer who could execute against this specific goal.

Solution

The CorrDyn Team:

• Rewrote the data pipeline between the vendor and the Team’s data warehouse to gracefully handle schema changes from the vendor.
• Developed the new pipeline to keep CorrDyn developers and team stakeholders apprised of changes in the pipeline while automating repairs caused by vendor changes so that breakdowns did not result in the loss of business-critical data.
• Automated testing against historical data to ensure data validity.
• Launched an automated business intelligence suite in Tableau that provides visibility to senior leadership and allows the analytics team to spend time driving decisions rather than producing reports.
• Scheduled all processes to run in sequence based on completion of events instead of time-based triggers. This led to decreased processing time and eliminated errors that would previously occur due to starting a downstream process early.

Results

Decision-makers are now able to make near real-time decisions because of the pipeline’s speed and reliability. They have shifted their focus toward new capabilities and away from questioning the accuracy or validity of their data. Since the project, they have incurred only a single failure in 4 months (without the loss of any data). Prior to this work, the Team experienced multiple pipeline failures per week.  

‍