How Trainline puts customer experience at the heart of its operations

Trainline CTO Mark Holt says his organisation is focused on delivering great customer experiences – and he’s using a mixture of big data and machine learning to help the transport specialist deliver on its objectives.

Holt, who spoke with Essential Retail before Trainline’s recent flotation on the London Stock Exchange, says the firm’s innovative use of data helps its customers make more than 170,000 smarter journeys daily. Trainline’s platform hosts more than 80 million customer visits each month, and more than 80% of visits are via mobile devices. The firm sells more than 200 tickets every minute.

At the core of Trainline’s eCommerce approach sits the firm’s app. The app includes a range of innovative features that exploit real-time data to keep passengers up to date. Holt, who was been with the firm since 2014, says iteration is the key to developing and delivering great customer experiences in a digital age.

Trainline uses a mix of big data and machine learning
Trainline uses a mix of big data and machine learning

“Get it out the door, try it, see if it works, try it again, see if it works again, and keep trying,” he says. “We work in a mission-oriented structure, so we put people together who are focused on a specific objective. That team will include people from technology, product specialists, testers and marketing professionals. And that team will all be focused on delivering a single initiative.”

Experience-led features

The firm’s app includes a range of innovative experience-led features that have been developed and iterated by Trainline’s staff. A price prediction tool, for example, analyses historical information and suggests the most economical time to book tickets, while a rail journey planner helps customers plan their next trip.

BusyBot crowdsources data from passengers about train crowding and provides real-time information on seat availability. Holt says more than 25,000 people interact with the app every day, sending information to help keep the service up to date and informing other passengers where seats are available on public transport.

“It’s incredible that we get that level of engagement from people,” he says. “We’ve made it super-easy and super-flexible. It’s very subtle and low key – it doesn’t get in the way of what people are actually using the app to do. It just allows customers to help other passengers. And it illustrates the level of engagement we get from people who use our technology.”

Data crunching and predictive analysis

Holt says Trainline’s price prediction app provides a good example of how the firm is already taking advantage of machine learning. For some routes, the app can predict to the nearest hour how long a particular price will be valid from several weeks in advance of a journey. The ability to make this prediction is based on the significant amount of data the business collects. “We crunch this information and figure out the trends,” he says.

Holt says the BusyBot app similarly makes heavy use of machine-learning technology. He expects further data-led advances in the near future. “We churn out about 900 million train movements every year, which is all the information associated to journeys, such as the time a train leaves and other route data,” he says.

Work in pioneering areas continues. The firm now offers a voice-activated app, so people who use Google Assistant and Siri can interact with the firm’s services. Voice services are being extended to a chat bot, so app users can ask questions about key issues, such as pricing and refunds, and the chatbot figures out the requirements of customers through machine-learning technology.

Holt says the key is to develop and iterate products on behalf of the customer, with the aim of producing higher quality experiences. “It’s important that we focus on getting stuff out of the door that people use,” he says. “We want to focus on products and services that make a difference for our customers. We will constantly be looking at things and asking, ‘who’s using this and why are they using it?’”

“Get it out the door, try it, see if it works, try it again, see if it works again, and keep trying"Mark Holt, CTO, Trainline

Trainline also uses customer feedback to help iterate around its products, not just via feedback from its apps but also in person. The firm brings customers into its headquarters in Holborn, London on a regular basis. Holt says user-testing sessions take place at least once a week, sometimes several times.

“We bring people in and we film them using the app,” he says. “We’re constantly analysing what our customers think. We’ve completed what’s probably the biggest rail survey that’s ever been done across Europe to look at customer pain points, interviewing thousands of people to see if they have the same or different problems.”