Shane Corstorphine

The Rise of the Machines

From bots to chat, A.I. to conversational UIs, cognitive computing, machine learning, deep learning... this session explains all.

Episode Summary

AI in the travel technology industry will still take a long time to change. Those who have been in travel a long time know that things take a long time to change. A major challenge in the travel technology industry, and I think is the fundamental issue, is the range of queries and choice in travel. It is astronomical. Massive queries, huge data. Because if you think about it, you got 10,000 airports so that gets you to 100 million different combinations of flights. Thousands of itineraries, hundreds of different partners flying those, different times, different schedules, different prices. And then you have different providers selling that same itinerary. So, the choice is absolutely vast. The challenge is to help the consumer to actually articulate and narrow down their searches to get the right solution for them.

What makes Skyscanner unique is their ability to add thousands of partners in a direct connect way. A lot of travel companies rely heavily on GDS’s to get the live pricing, whereas they have 1,200 direct connects with partners. When you move into fragmented markets where there’s a lot of low-cost carriers who do not have pricing available through a GDS, that allows them to have much better coverage of flights and prices than others may have who are solely operating with GDS reliance. It’s a huge advantage but also a massive challenge.

Skyscanner has started to experiment with bots and voice. It’s interesting to see the engagement increase from when they first went out earlier in the year. Month-to-month seeing 200% growth rates as more people are starting to engage this way, and what’s interesting is the queries are getting shorter as people start to figure out how to use it. Kind of like when people started to use mobile devices it took them quite a long time, they get faster because they figure out how to make it work. Interestingly, we see a higher conversion rate on bot than on mobile devices, the majority of our queries come from mobile now. With this business being global we have a big presence in Asia, which we see as a lead indicator for the West about what’s happening, and that is that people use voice to do a booking, which is actually a lot faster than typing, and if they’re being able to book travel using voice in large volumes it’s just a matter of time until that spreads West – as we saw with the shift to mobile, which we saw first in Asia.

I think that those who have been in travel a long time know that things take a long time to change, and I believe that AI will still take a long time to change.

A major shift regarding the costumer satisfaction level is the recent introduction of Partner Quality Score, the star rating system against partners. This is a first move away from price comparison to product comparison and that’s affected the booking experience comparison and also have customer service that feeds into that star rating. There’s been a massive increase in how people now select in terms of choice. It used to be that they went to the cheapest regardless. Now people are moving away from the cheapest and actually going towards a balance between price and booking experience. These allowes companies to actually potentially charge a premium for that service and customers will be happy to pay it because they’ll know it’s a better experience.

With this there’s been a few challenges, one of them is regarding the UX, how do people think about how it? How do we make it clear? It’s about booking experience rather than the flight experience for instance. We had a lot of work to do around about that, testing, making sure it was gonna work and this was a multi-month project, working with UX group support, working with teams, testing it, etc. Then you got the partner relationships, we’ve got to talk to them a lot because partners are hugely important to explain why we’re doing it and explain how we’re approaching it and actually give them time to change their operating models to ensure that they are given suitable customer service and other things that they know they can be scored against, so they’re not so suddenly downgraded. You have to be really sensitive about how you roll these major changes out.

What I think is the fundamental issue, and the requirement therefore for data science and the AI that sits on top of it, the range of queries and choice in travel is astronomical, I mean even from a flights point of view we're taking three billion queries a day and looking them up in kind of a fraction of a second for our users.

Questions & Answers

What is the stage of AI in the travel technology industry?

A: You know, there is a spectrum of tech and travel and some are very much still in that kind of brand end of stuff which to some degree I think is somewhat head-in-the-sand and others are right at the forefront of tech. There is no doubt that, I think those who have been in travel a long time know that things take a long time to change and that also might have been part of it. So, I believe AI will still take a long time to change. What I think is is the fundamental issue, and the requirement therefore for AI, is, well first, for Data Science and then the AI that sits on top of that, is the, uh, the range of queries and choice in travel is astronomical.

 

What is a major challenge in the travel technology industry?

A: Massive queries, huge data. Every day we’re catching and looking this up. Because if you think about it, you got 10,000 airports so that gets you to 100 million different combinations of flights. Thousands of itineraries, hundreds of different partners flying those, different times, different schedules, different prices. And then you have different providers selling that same itinerary. So, the choice is absolutely vast. The challenge is to help the consumer actually articulate and narrow down their searches to get the right solution for them.

 

What makes Skyscanner unique?

A: Our ability to add thousands of partners in a direct connect way. A lot of travel companies rely heavily on GDS’s to get the live pricing, whereas we have 1,200 direct connects with partners. When you move into fragmented markets where there’s a lot of low-cost carriers who do not have pricing available through a GDS, that allows us to have much better coverage of flights and prices than others may have who are who are solely kind of operating with GDS reliance. It’s a huge advantage for us – it’s also a massive challenge.

 

What technologies is Skyscanner using today and planning to use in the future?

A: We’re starting to experiment with bots and voice. So, we now power Alexa so you can run Skyscanner queries through Alexa but using voice and bots. Right now we only have bots in English language, but it’s interesting even to see the engagement increase from when we first went out earlier in the year. Month-to-month we’re seeing 200% growth rates or sit quarter on quarter as more and more people are starting to engage this way, and what’s interesting is the queries are getting shorter as people start to figure out how to use it. Kind of initially when people started to use mobile devices it took them quite a long time, they get faster because they figure out how to make it work. So, we’re seeing the queries and the time gets shorter, but interestingly, we see a higher conversion rate on bot than we do on mobile devices, which is incredible – how people are starting to use this to query. So, we’re pretty excited by it.

 

How does this back end process affect the costumers’ satisfaction level?

A: A major shift for us recently is that we introduced Partner Quality Score, which is the star ratings that you’ll now see against partners. This is a first move away from price comparison to product comparison and that’s affected the booking experience comparison and in that we also have customer service that feeds into that star rating. We’ve seen a massive increase in how people now select in terms of choice. It used to just be they’re going the cheapest regardless. Now we’re seeing people move away from the cheapest and actually move towards a balance between price but also booking experience. And that’s great because what you can then see is guys who have great customer service and great booking experience can actually potentially charge a premium for that service and customers will be happy to pay it because they’ll know it’s a better experience.

X

For a written transcript copy of this interview please fill in the information below:

Thanks! we will be in touch soon

Close
Focus Mode

Contact Request

Close

We will call you right away. All information is kept private