My colleague Eric Tyree shed some light on predictive analytics, so there's no need for me to dive into definitions and theories. Let me take the pragmatic approach instead and share with you some of the coolest developments we’re working on at the moment. We may not be Gandalf, the wizard from Lord of the Rings, but data enables us to create our own magic.
“A wizard is never late, nor is he early. He arrives precisely when he means to,” Gandalf famously said. A pilot we’re running with machine learning company Flyr enables us to do just that. We buy before a ticket goes up in price or wait because we know it’s going to go down. The proof of concept shows us this can make around two percent of travel savings appear. Magic!
On the hotel front, we're starting to work with a different startup on a similar project but looking at hotel prices. As with Flyr, timing is everything, and making the reservation at the right time can make four percent of hotel costs disappear.
Both partnerships, along with CWT Price Tracking, will enable us to implement the most efficient and sophisticated air booking and hotel booking processes in the world. The great thing about this is – and here comes the magic – the analysis is running in the background, completely invisible to the traveler, who does not need to worry about a thing.
Less choice, not more
There's also a lot of data-enabled magic around the personalization of our customer interface, myCWT. Throwing up 5,329 hotels and 1,289 flights for a trip from Barcelona to Paris when we know our client is going to be there at 9:00 am, stay two days at the hotel next door to his office, and be back at 7:00 pm makes no sense. Refining the results each traveler will be able to see based on stored knowledge is the next-level personalization our predictive analytics offers to clients.
What to do with the time
All these advances will become even more significant as people increasingly manage everything through their mobile phones, with just the sound of their voice. That’s the promise of our app, CWT To Go. Tell the app where you want to go and when, and CWT technology will work its magic to:
- Understand what the traveler is saying.
- Feed the information to a system that knows the traveler's preferences, and makes reservations accordingly.
- Recommend two to three bundled options for the traveler to choose from, eliminating the hassle of booking flights and hotels separately.
- Issue tickets using our predictive models to optimize costs.
- Leave the traveler happy at having completed the reservation in 10 seconds and getting 30 minutes of their life back.
Now, as Gandalf wisely said, “all we have to decide is what to do with the time that is given us.”
Blog author: Carlos Sánchez, Senior Manager Big Data Analytics and Product Innovation, Carlson Wagonlit Travel