April 2018, San Francisco
Calling all data scientists: are you ready to be humbled by travel?
The role of the data scientist is now firmly embedded in the world of travel tech, and firms like Skyscanner and Booking.com are on the hunt for talent
Not all that long ago, in fact just three years ago, the role of ‘data scientist’ did not exist. How things have changed! Today data science is at the heart of every travel tech business and people with the right skills are very much in demand.
This we heard in the morning keynotes on day 2 of EyeforTravel’s Data Summit in Amsterdam last week. Both Onno Zoeter and Mark Shilton, principal data scientists at Booking.com and Skyscanner respectively, were very clear: they are hiring, and at all levels.
In terms of the standard skills required by data scientists there are three: mathematics/statistics, computer science and domain expertise (ie. knowledge of the business). However, according to Shilton, people with all three “mostly don’t exist”.
He was also quick to point out, that candidates lacking mathematics and statistics skills should raise a red alert. Why? “Because either through ignorance or malice this overlap of skills gives people the ability to create what appears to be a legitimate analysis without any understanding of how they got there or what they have created,” he says.
A humbling experience
Booking.com’s Zoeter, who has been in the data business since the 1990s, admits the job at a firm which processes 1.4 million hotel bookings each day, is anything but easy. “I’ve learnt the painful way,” he says, of his move into the travel industry two and a half years ago.
Previously, Zoeter's experience had included working on click-through rate prediction systems for a major search engine and designing an adaptive pricing solution for on street parking that is now in use in multiple cities. As a result, he was named one of Fortune’s 20 Big Data All Stars in 2014.
So, when Zoeter joined Booking.com one of his first thoughts was: how hard can it be to put hotels in the right order? What he has discovered, however, is that “it’s hard, much harder than you might think!” Indeed, working in the travel industry has been a “humbling experience”
How hard can it be to put hotels in the right order? Harder than you think!
So, what does Zoeter look for when hiring people? It’s tough because the term data scientist spans a wide range of skills.
“On the one hand you might have people who know databases and can ensure millisecond serve time all over the world. On the other, you have those who are really entrepreneurs and commercially aware,” he says.
In an ideal world, however, Booking.com, which today has a team of 50 data scientists, wants to hire people who are “technically autonomous and commercially aware”.
In hiring, Zoeter says they are actively trying to differentiate by giving people the opportunity to specialise.
Developing broad skills
When Shilton joined the Scottish metasearch firm (which was recently acquired by C-Trip) six years ago, there were fewer than a hundred employees. In those early days the Skyscanner team was essentially a couple of business intelligence (BI) analysts, of whom Shilton was one, and their role was essentially to produce BI reports.
When the firm made the move into so-called ‘data science’ it wanted data to be a skill that everyone at Skyscanner had, and one which would contribute the highest value to the business. Says Shilton: “We wanted everybody to be at a level that would allow them to do the basic analysis and to be able to ask and answer the questions ‘what happened?’ and in most cases ‘why?’”.
We wanted everybody to be at a level that would allow them to do the basic analysis
Today, however, Skyscanner employs 1,000 people and 23 data scientists, and the team is growing.
But Shilton warns that it’s very easy to get seduced by the flashy lights of innovative technology and the promise of new techniques. “None of it is worth anything unless you have a solid base to work from. And funnily enough, that base looks very much like the lower tiers of BI and analytics,” he says.
With this in mind, Skyscanner too is on the hunt for talent, and its goal is to develop people with a broad range of skills. This is part of a wider initiative to give people a more rounded skillset, “to become more ‘T’ shaped individuals,” he says.
Today data scientists at Skyscanner sit in teams where there is a specific need, but this is still coordinated by a central management function.
Data science is undoubtedly core to the business of Skyscanner and Booking.com.
Says Zoeter: “Aside from the hype, it is real. There is AI driven chess, we have self-driving cars, I can talk to my phone. There are things that things actually work.”
The big question, however, is this: are there enough skilled people to propel this innovation into the future?