Streamlining New York Commuting by refactoring the Coach USA App

The Challenge

Coach USA were experiencing high crash rates with their ticket scanning app which was drastically increasing the times their buses were stationary at bus stops. This had a knock on effect across the whole route, delaying commuters and making frustrated drivers miss their target journey times.

These issues were creating a frustrated work force and 50,000 unhappy commuters in and out of Manhattan everyday. With the technology being the blocker to the companies success either changes were required or a rethink of strategy.

Indiespring Action

The challenge was to make improvements quickly with the project to see how viable the app solution was for Coach USA in the long term. To this end did a lightning release with new analytical platforms to allow us to capture data and identify the issues within the app. We noticed that crash rates were soaring to around 90% of users experiencing one. We found issues in the code that created an infinite loop of recursion as every record was being sent separately, through the same process and this would only stop when the app crashed.

To fix this major issue in the code we created a system that all records for the last 30 seconds were sent by a single call, this allowed the app time to run and finish the process before another one started drastically improving the crash rate.

With crash rates reduced the feasibility of the app as a long term solution for the ticketing system of Coach USA was proved and further work to improve the app could be undertaken.

Alongside the data our new analytical tools were providing we spent time with end users, in this case the bus drivers, to identify their critical needs for the app. Speed and ease of use were identified leading to a future UI refresh to support them in speeding up scanning. The feedback from the users has been overwhelmingly positive after years of frustration and a large part of the opposition in adopting the app have now turned into some of its biggest advocates.

With the success of the initial releases and support for the app’s long term success growing we can now move our attention toward hitting the target of 3 seconds per scan. This key KPI would reduce onboarding time and save Manhattan several hundred hours in lost commuter time per day.

The Result

– 50% reduction in fatal app crashes in first 3 months
– Crash free users increased to 95% of user base