The global travel industry is being re-built from the inside thanks to thousands of data analysts, computer scientists, and product engineers who are in their twenties and thirties.
But you could attend many trade conferences without hearing from these front-line soldiers. Panels tend to be reserved for people who are well established in their careers or who have executive-level perspectives.
Yet given the trade's big talk about Big Data, it may be refreshing to hear from a data analyst who is early in their career at a fast-rising startup.
Ewan Nicolson is senior analyst for Skyscanner, a travel search engine. He reports to the British company's newly hired chief data officer.
Nicolson is the opposite of a self-promoting, TED-talking Californian. Instead, he's a modest Scotsman. He isn't the type to toot his own horn at the expense of his company's 600 other employees.
That said, Nicolson has learned some lessons the hard way. These may be useful to technology professionals who are also early in their careers.
Culture matters
Choosing the right company to work for makes a difference, Nicolson said in a phone conversation with Tnooz. Culture can be difficult to gauge from the outside. You want to find out if a company's culture embraces an investment in data analytics.
"When researching a company, try to talk to people who are employed there or look at the products they've made. Try to determine if your potential bosses appreciate the value of data.
Do they give their employees the space to experiment and occasionally fail?"
Opting for a high-growth company also helps. He believes he's had more opportunities in his career working for a company that's one of the 50 fastest-growing tech companies in Europe than if he were elsewhere.
Finding inspiration
Once hired, many early-career analysts want to get assigned to important projects. Nicolson advises:
"The best thing you can do is Google for the phrase 'growth hacker' and read up as much you can on how people are using data for rapid growth. Check out sites like growthhackers.com and andrewchen.com."
But it may be difficult to come up with projects at your company that are suitable. When you take the records generated by your business and try to apply growth hacking techniques, you may get upset, says Nicolson.
The data -- be it Web server logs, Google Analytics details, or records of all the bookings your company has collected -- may not be quite as good as what growth hackers elsewhere have to work with.
Your company's records may be incomplete, so you may think you can't do as impressive things as the hackers elsewhere who are creating truly explosive growth.
For instance, you may not be able to join up activity records across devices and platforms to create a full picture of customer behavior.
But don't despair.
"No matter how simple or undetailed your data set is, you can still do something with it that helps advance your company and your own career.
You can build a prototype of where you could go in the future, and show that to your bosses to encourage them to invest in collecting data properly."
Productizing data
At Skyscanner, Nicolson's team found lots of valuable material in the data sets on hand, thanks to the company's 35 million monthly users.
Skyscanner has found that productizing data for internal and external business-to-business (B2B) can be useful and profitable. Nicolson has worked on several of these projects.
"For example, we collect a lot of data on customer behavior. We can find out that Italians tend to book their travel 14% closer to the travel date than Germans do. (Yes, that's an actual statistic we found.)
We can then optimise our marketing budgets and content accordingly.
But what we hadn't been doing is finding a way to offer these insights to third-parties, like destination marketing organisations, airlines, airports, and investment companies, who might also be interested in the trends.
So we created a product called Travel Insights.
For instance, a stakeholder in the company or an external client can access it to see which countries will send the most inbound visitors to their destination this year, separated by business and leisure travelers.
Before we created this product, we'd have to dedicate analyst hours to pulling reports and making spreadsheets and emailing them. That doesn't scale. As our startup expands, it isn't cost-effective to do that.
So now we're building data products. We ask ourselves, can we make something repeatable and self-service?
This model of making data products instead of one-off reports is something other companies could do as well.
My main recommendation, if you are thinking about doing that, is learn to code in Python, which is an elegant, powerful programming language.
Then you'll be able to automate the boring business intelligence work you may not otherwise want to do, meet people's needs faster, and leave your team with time to do more interesting work."
The Travel Insight product was launched in November 2014. But the idea was born in spring 2013.
"We had to prove that this would be valuable and something that people would be willing to pay for. Once we had done that, we spent time making it polished and tailored to client requests. We tested a user-friendly interface."
The goal was to make actionable information as a decision engine. But achieving that wasn't easy.
From time-to-time, the Skyscanner team would find information that seemed interesting. But from the perspective of clients, that information was "noise", not "signal" they would profit from.
Frequent product testing was key to make sure products passed the 'So what?' test.
Along the way, Nicolson learned that, at a travel company, you have to think of A/B testing in a broader way than you do in most other ecommerce industries.
"Travel isn't like other industries in that, as data analysts, we don't have the classic web analytics funnel that you often hear about. You don't have: Person comes in; person puts item in cart; person purchases item.
In travel, you have to understand more about the customer's motivations because travel is a much more complicated product than buying, say, golf clubs.
We have to track which pages customers view, which ones they spend a lot of time on, which ones convert better and which ones make people leave the site and not come back. We use anonymized data for this.
So, for us, A/B testing is not just about moving a button from one side of the screen to another based on customer feedback.
What we do is perform an experiment and try to understand what the results tell us about the model we're building of customer behavior. That's something others elsewhere can adapt for their own gains.
It's one of many things that make travel a uniquely rewarding vertical for data analysts."
MORE:
Skyscanner’s airfare data opened to airlines, airports, and marketers for B2B Big Data access
Poor data models are killing the travel loyalty star, says Booz Allen Hamilton
Dear travel industry, we desperately need more mentors for budding entrepreneurs. Thanks!
NB: Image of an eye looking at drop courtesy of Flickr/Laszlo-photo via Creative Commons.