Mark Ross-Smith On The Future Of Big Data In Travel
We’re living in the greatest digital transformation of our time. Smartphones have given us unprecedented access to greater connection, commerce, and communication. As the power of mobile travel grows larger the world feels smaller. As travellers, we’re well informed and empowered to do and go further than every before — with just this single device.
With all this opportunity from technology comes equal challenge for companies to utilise data, or ‘big data’ as it’s known, taking this abounding stack of insights collected from consumers and increase business efficiencies. Travel providers are seeking new ways to utilise big data to disrupt, reach and delivery higher quality of customer care and services.
Mark Ross-Smith, founder of traveldatadaily.com is one of Australia’s (and Hong Kong’s) thought leaders in big data. We sat down to gather insights on how big data plays a crucial role in the future of travel for both consumers and businesses.
What’s the landscape of big data in travel today?
Frankly speaking, most travel companies (airlines, hotels, OTAs etc..) are moving slowly in the data science and big data space. Airlines are not known for their innovative qualities, and after enjoying low oil prices the last few years — using big data to solve problems and drive new revenue hasn’t been seen as a priority. Large hospitality chains, similarly, have been focusing largely on new properties and traditional advertising to entice customers to book directly instead of through OTA channels.
However, in saying this, there are innovations in the travel space we’ve come to light over the past year. More airlines have fast-tracked their integration to NDC, which offers the ability for airlines to manage their own offer system and move away from legacy IT systems which hinder their ability to leverage the true power of big data to drive revenue.
What opportunities are there for data-driven companies to disrupt the industry?
The opportunity I see is for smaller, more agile start-ups to take advantage of more traditional, slower moving organisations.
Uber is now beginning to sell transportation data to governments and city planners. The taxi industry has been sitting on the same data for decades, but never put the pieces together. So while Uber is busy taking market share of the traditional taxi business, they’re also building intelligent data products which serve a higher purpose. I also think we’ll see Uber begin to sell more personal data back to banks in the near future — as Uber is able to predict your credit rating.
Flyr raised $8 million last month, a US-based tech start-up which can predict whether to buy or hold out on buying an air ticket. Hopper is a similar business, and these start-ups will eventually take power away from airlines and GDS if left to their own devices.
PEX Portal, the US-based award flight search engine is leveraging data in a unique sense where they actually pay the airlines to be on their search engine. This creates a revenue stream for airlines while PEX can fill award inventory of airlines better than some airlines can themselves!
Hack Horizon is another innovative approach who aim to have the world’s first hack-a-thon at 35,000 ft, bringing together some of the smartest entrepreneurs in the travel industry, to create new solutions and products for the aviation industry.
2017 needs to be the year airlines step up, get involved and engage with start-ups and minds from outside the traditional aviation industry. Innovation rarely comes from within, and I think we’ll find this is the same for airlines.
Editors Note: Qantas has created their own accelerator program called AVRO
Where will consumers feel the effects of big data in travel?
There are both positive and adverse effects for consumers, but it’s important to remember that airlines make money from data, so if they lose your trust, or push the envelope too far — there is a very real risk their big data endeavours can have unintended adverse effects.
Cathay Pacific can predict what type of alcohol you will drink in first class without asking you, purely through the collection of small data. This has significant benefits for both the passenger and the airline — where you get a highly personalised experience, and the airline reduces wastage and the need to load excess bottles onto each flight, thus saving in weight and fuel burn.
On the flipside; If an airline knows you frequently fly with their competitor, it’s possible for the airline to offer dynamic pricing and a significantly better price for flights you may be searching for. No so much in Australia, but certainly internationally — airlines have large and granular data sets which help build a 360-degree view of each passenger, so they can understand what share of wallet they hold.
Airlines are then able to use data for price discrimination, whereby you may see different pricing for the same flight than your friends.
How should airlines be using big data to streamline and create efficiencies?
The most common issue I come across (not just in airlines), is management read a few articles on big data, go out and hire a bunch of data scientists and expect them to work all kinds of ‘data magic’, and the company will somehow benefit from the result. This is why most analytics projects fail.
It’s best to start with a top-down approach. Begin with a list of problems and work backwards to find data which can help solve, or improve the operational efficiency associated with the original problem. This is largely a creative process — so teams of data scientists won’t be adding a large amount of value until the final stages when validation is critical.
“Innovation rarely comes from within, and I think we’ll find this is the same for airlines”
Additionally, the intended goal to streamline efficiencies may not be possible, or a more important outcome may present itself.
For example, a Southeast-Asia based airline was looking at if they could predict when a customer would buy a flight ticket. Throughout the process, they began with segmenting the loyalty customer database and found all their predictive models using machine learning were coming in at 50% or less (indicating not enough data, or the wrong model is used). They used similar models and feature points (user data points) and found they could predict when a member would book an award flight with 81% accuracy. This lead to an internal shift in focus from revenue flights to the loyalty points currency and how their operational team treats award flights has affected their CRM and ultimately the profitability of the loyalty program.
Mark thanks for your insights! How can businesses best get ahold of you or follow you?
[Editor Note: Originally Published Feb 2017]