Community Transit Uses Data to Improve Operational Efficiency and Rider Experiences

Scenic beauty, temperate climates and vibrant industry are a few reasons why the United States’ Pacific Northwest attracts millions of visitors and career transplants annually. Whether they’re headed for the nearest hike or to work, residents and tourists alike rely on public transit to reach their destination and keep local communities connected.

Fueled by the mission “to help people get from where they are to where they want to be,” Community Transit operates bus, vanpool, shuttle and other transit services in the Seattle metropolitan area. To deliver on this mission, Community Transit turned to Snowflake to create a single source of truth for its data, help teams across its agency better understand their riders, gain data insights faster and ultimately deliver a better transit experience for customers. 

Operating more than 300 buses on 47 routes while ensuring on-time, reliable service is no small feat. Service of this magnitude requires a carefully crafted strategic vision, smart scheduling, effective recruiting and lots of data. “As a transit agency, we have data from a ton of sources, whether that be internal sources about employees or external sources about bus ridership,” says Ian Hafey, Senior Data Analyst at Community Transit. “Historically, we had all that information in separate systems, which led to data silos for our teams.”

Seeking to democratize insights for users across its organization, Community Transit began consolidating data into a SQL database. But managing hundreds of millions of rows of Community Transit’s data required complex indexing techniques, many database management tools and multiple extract, transform and load (ETL) data pipelines, which led to performance issues. Large data processing jobs took up to nine hours to run (assuming that nothing broke) during non-peak hours, which caused concerns about data freshness. Subsequent data processing delays forced Community Transit to limit access to historical data sets, such as the company’s Advanced Public Transportation Systems (APTS) data. According to Hafey, “We were forced to cut that data set off at two years because running it each time took 45 minutes.” 

Preparing for Community Transit’s massive route network restructuring initiative — one that calls for an annual increase of approximately 100,000 service hours — during a coach operator and mechanic shortage underscored data’s importance. Yet data reliability issues inhibited collaboration among scheduling, recruiting, training, finance and marketing teams. “We weren’t able to have meaningful conversations across the agency,” says Olivia Woods, Senior Manager of Research and Analytics at Community Transit. “We were all speaking a different language and had different numbers.”

Rearchitecting with Snowflake freed up Community Transit’s technical staff to focus on higher-impact work instead of database maintenance. Snowflake’s elastic performance engine scales to support virtually any number of users, data or jobs, allowing Community Transit to process more data in less time. Processing the entirety of Community Transit’s APTS data set — approximately 10 years’ worth of data — takes about 10 minutes with Snowflake. That’s five times more data processed, with 77% less wait time.

With fare and cost allocation models now in Snowflake, Community Transit better understands route productivity. “Previously, this modeling was done in a spreadsheet and required almost one FTE to keep it up to date,” Woods says. “Now it takes about one day of work to build another model, and we basically do two each year. It’s reduced a lot of overhead.”

Community Transit’s human resources department also relies on reports powered by Snowflake to understand recruitment pipelines without spreadsheets. According to Woods, “It’s helped them visualize their workflow, and that visibility allows them to achieve greater efficiency in their work.”

Using data to optimize routes and recruitment aligns with Community Transit’s vision of “making travel easy for all.” For example, as the coach operator shortage threatened to impact rider experience, Community Transit reconfigured its service to improve on-time performance and deliver a better experience for riders. “Now we barely have any missed trips,” Woods says.

Launching Community Transit’s analytics portal will further expand data accessibility and bolster efforts aimed at achieving 510,000 annual service hours by September 2026. Providing self-service access to cost per rider, cost per route and other metrics will reduce ad hoc reporting requests while extending data insights to even more teams for smarter decision-making. 

Already helping Community Transit save time and expedite collaboration with an external transit agency, Snowflake Secure Data Sharing could help the organization scale these efforts, making it easier to partner with additional agencies. Moving more data to Snowflake is also on the roadmap. 

“We’re still working on migrating our geospatial data into Snowflake,” Woods says. “That should deliver some big efficiency gains while reducing errors and helping us plan new routes to further improve service for our riders.”