Trainline: Data Insights Provide Platform For Seamless User Experiences

splunk data analytics

A culture of experimentation is enabling Trainline to streamline the web experience for customers through data-driven insights

The use of big data analytics in business is going through a wave of evolution at the moment, with the focus now very much based on providing real-time insights.

London-based Deliveroo, for example, is combining big data with machine learning to improve its delivery efficiency, while the Environment Agency is transforming how it responds to flooding incidents across the country through the use of real-time data intelligence.

Even Wimbledon is getting in on the act, using artificial intelligence to provide “insights in the moment” and increase engagement with fans at this year’s tournament.

And digital train ticket retailer Trainline is no exception. Speaking to Silicon, the company’s CTO Mark Holt explained how it is using insights provided by big data to differentiate itself and improve the customer experience.


Data driven

“I’m slightly obsessed with user experience because it’s everything now, it’s what differentiates us from everyone else,” said Holt. “And we can apply a lot of intelligence to do that. The most basic data product is just getting the data right, getting that baseline there and then laying smarts over the top.”

The Trainline platform is now run 100 percent on Amazon Web Services (AWS), which Holt claims has enabled it to funnel all of the customer data it collects into one place and generate useful insights from there, all while saving costs.

“If we were sitting in a big data centre we’d be spending fortunes on disks and servers and processing and it would be a horrible experience. So we use a massive amount of Amazon stuff and also we try not to run too much of that stuff ourselves, we let Amazon run it for us. That helps us to crunch a lot of data.”

In terms of the website specifically, Trainline is focused on “being able to create wonderful, frictionless experiences and making it as easy as possible” for visitors.

This is achieved through a combination of logging over 900 million train movement events every year to help predict train patterns, along with the collection of user behaviour data to gain an understanding of travelling trends.

And then there are the more innovative projects going on, such as a tool that advises customers where they are most likely to find a seat depending on the location and direction of their specific journey.

“We’ve built this thing called BusyBots, which is crowdsourcing seat information saying things like if you go to the back of the train you might get a seat,” Holt explained. “Very simple, but actually the models we create are quite complicated.

“For example, if you’re going from London Bridge to Brighton, go to the back of the train and you won’t get a seat but everyone’s going to get off at Gatwick Airport and they’ll all get off from the back of the train. So it’s that kind of stuff that is quite interesting.”

data centre, big data

Creating culture

One factor that Holt claims has been essential to the company’s innovation with data is the culture. Developers are encouraged to experiment, test out their ideas and solve problems, with the customers being the ones who reap the rewards.

 “What we try to do is give the teams the ability to make stuff happen,” he said. “It’s about creating a culture. You can’t teach creativity but you can create a culture which encourages it and I fundamentally believe in that.

“The only thing I can do as a CTO or CIO is create a culture where people can innovate. I don’t have time to write the code, I don’t have time to sit down and dive into the depths of the user experience, but what I can do is make sure the culture is there.”

And this has had the added benefit of making Trainline an attractive place for developers to work, which is invaluable when you consider the skills gap currently affecting the industry.

So, for the time being, things are looking good. Trainline will always have to compete with the likes of Facebook and Google for the top developers but, if the culture remains, there’s no reason why the attraction won’t either.

Quiz: What do you know about Big Data?