100 Years of Advancing Destinations

Itineraries at the Crossroads of Machine Learning and Travel Psychology

Author: Joy Lin
Posted: March 16, 2015
Blog Topics Covered:
1 of 1

Today's marketing environment is approaching a one-to-one level dialogue, and travel planning is no different. With machine learning driving predictive analytics and powering programmatic placement of our marketing messages, destination marketers have many opportunities to work collaboratively with such technology to place the most creative and inspiring messages in front of the right customer at the right time.

Gilad Berenstein, CEO and Founder of Utrip, a personalized day-by-day trip planner powered by artificial intelligence, joins us in this interview on his background with the power of data and how Utrip is using it to travelers' (and destinations') greatest advantages.

Describe one of your earliest memories of seeing data at work.

Data brings power to the people. I learned that when I was very young. My father is an engineer and business owner. He is also an early adopter in using artificial intelligence on large datasets to identify patterns and create efficiencies. I grew up watching my father harness the power of data to solve problems and produce better outcomes for customers.

How had your initial experience or experiences with data and intelligence directed your career path since

As an entrepreneur, I have paid close attention to how different industries leverage data and intelligence. I have learned what to do and more importantly, what not to do. For example, I watched as innovative companies like Farecast, one of the first businesses to use machine learning, became obsolete. Farecast used algorithms to sort through data and predict price fluctuations. It could literally tell you when to purchase an airline ticket. It was a game-changer. But those predications relied on a constant flow of fresh data. When Farecast was sold, the acquiring company did not purchase new pricing data. Without fresh data, the accuracy of the predictions declined. So a company that was delivering real value to customers became obsolete. Another lesson was the absolute value of human expertise. Judgment is a big part of the equation, and takes intuition and experience. I learned you had to have both: human expertise coupled with the power of data to deliver the best results.

How does the intelligence that Utrip leverages help travelers? How does it help vendors?

Rather than making a traveler sort through a long list of activities and reviews, we use algorithms to do it. We ask for a traveler’s interests and budget and then use predictive technology to create a personalized itinerary—in minutes. For vendors, the biggest question we help them answer is why travelers behave as they do. Through bookings it is clear that a flight or hotel was purchased, but businesses don’t know why that action took place. And when you understand the why, you understand how to influence additional people to take the same action.

On a scale of one to human, how far along the path would you place machine learning? What are the current limitations and do you see them going away in the next five years?

There are things that machines do better, and things that humans do better. Humans bring judgment and creativity into the process, something we can’t expect from machines. If you want a list of the 5,600 different restaurants categorized by cuisine and sorted by neighborhood, that’s a job best done by an algorithm. But if you want a list of the best restaurants in Seattle, asking renowned chef Tom Douglas is the way to go. To get superior results, you need a combination of data, machine learning and local expertise. And I don’t see that changing in the next 5 years.

What’s on your information wish list? If you were Big Brother, what is one thing that you would love to know or track about travelers?

I would love to peek in on a traveler during their trip and see exactly how long they spend on each activity. A person who visits the Louvre for one hour is very different from the person who visits for seven hours. We are starting to gather some of this in-location data with our mobile app, but we aren’t quite there yet. And I want to be, because funneling that data into future recommendations would be the ultimate in trip personalization.

Gilad Berenstein joins DMAI in person at the Marketing Innovation Summit April 16, 2015 to share more about personalization and psychology in travel decisions.