We were tasked by a major commuter rail operator to analyze ticket sales data to determine ridership with origin-destination and time-of-day detail. The operator had four distinct sales channels: mobile ticketing, paper ticket vending machines, monthly direct-mail ticket subscriptions, and on-board sales by train crew. The mobile ticket platform generated the most extensive data, including an activation record each time a periodic pass product or multi-ride carnet was activated. However, this data has a significant sync lag and only 40% of total market share. The other sales channels provide only point-of-sale data and not point-of-use. We developed a set of working assumptions and built a model that enabled the statistical utilization patterns implied by the mobile ticket users to be applied to the data generated by other sales channels, thereby providing a daily estimate of fifteen-minute resolution ridership demand by origin-destination pairs, a high level of granularity for service planning analysis.
Related Publications/Presentations:
Estimation of Pre-COVID19 Daily Ridership Patterns from Paper and Electronic Ticket Sales Data
Ticket Ridership Big Data Origin-Destination Presentation
