How WayBlazer's Dynamic Personalization Drives Results for Travel Companies

Scott Goldberg
VP of Business Development & Marketing
Chatbot on a mobile device

Finding one or two core objectives to pursue is good practice for any company. The best core objectives have a long horizon — ie, things that will be true for that company well into the future. At WayBlazer, one of our core objectives is to create the most personalized travel recommendations possible. We believe travelers will always value great, relevant recommendations.

Of course, we’re not alone. Personalizing your product or service is important to every company in every industry, and most companies are aware of the tools available to help accomplish that.

We’ve seen many different flavors of personalization in the travel industry. Some companies use past booking behavior to predict the right travel products to show a potential customer. Companies like Adara and Sojern follow a traveler throughout their booking process, tracing their interactions across travel sites to better understand the holistic picture of their trip. Airlines like Delta personalize their in-flight service for loyalty customers by referring to them by name and understanding their past preferences.

There’s no single right answer. Any effort to personalize your products and services should, in theory, help achieve some of your objectives.

Our approach at WayBlazer is pretty simple: we try to understand the intent of your next trip, then make personalized recommendations accordingly. And we do this wherever you might be, whether on desktop or mobile, through display ads or messaging platforms. Here’s how we do it…

Personalization starts with intent

There are differing opinions on the data required to personalize a recommendation. Our experience is that trip intent is the best place to start. And we believe that all trips, regardless of whether they’re business or leisure, share 5 basic attributes:

  1. Where are you going?
  2. Who are you going with?
  3. Why are you going?
  4. When are you going?
  5. What are you planning to do there?

The only thing that varies is the specificity of each attribute. A bucket list trip to celebrate a big milestone 2 years from now typically has low specificity (“I want to go somewhere tropical for my 10th anniversary”) whereas a trip to an industry trade show 2 weeks from now has high specificity (“I need a hotel with a gym in Chicago from June 16–18 near The Loop”).

Either way, getting answers to those 5 attributes, regardless of the level of specificity, creates a phenomenal starting point for making personalized recommendations. How to do that, however, isn’t quite as straightforward.

Personalization doesn’t fit in a box

Many people remember The Net, the 1995 thriller about a computer programmer who uncovers a conspiracy. Watching Sandra Bullock’s character book a flight and pick her seat online was mind blowing at the time.

Our co-founder and chairman, Terry Jones, founded Travelocity in 1996, just after The Net came out. It was a major shift in travel planning and booking capabilities.

Travelocity’s home page in 1996

So major, in fact, the basic functionality hasn’t changed in 22 years.

Travelocity’s home page in 2018

Destination, Dates, and Number of Travelers. That was the Trip Intent machine established in 1996 and it continues to be the model for nearly every travel company today.

What if you, as a traveler, don’t have perfect information on those three attributes? Good luck.

Trip Intent doesn’t fit in a box.

  1. Where are you going? Maybe to San Francisco. Maybe to San Jose. I’m not sure which is better for my meeting in Palo Alto. Or maybe I don’t know at all. Maybe I’m trying to decide. Maybe I’m thinking about Europe. Or South America. Maybe I’m thinking about Florida. What’s the weather like in Florida in October?
  2. Who are you going with? I haven’t decided yet. My wife is coming, I know that, but we’re not sure about the kids. Our in-laws might come too. Or they might babysit the kids. Or maybe I’m traveling solo for work and a colleague may or may not be joining. It depends who attends the meeting on the other side.
  3. Why are you going? To get away! To experience life. To read, relax, and try new foods. To go to the beach, play with my kids, get a massage. To go sailing or scuba diving. To go wine tasting. To see a concert, play golf with my friends, meet new people. Or to see clients, to entertain them, take them to dinner, catch up with them face to face…
  4. When are you going? I’m not sure exactly. I know I want to go in August. Or maybe August 5–13. Or maybe in the early fall. I have 2 precious weeks of vacation this year and I want to use it right. Or maybe we haven’t set a date for our meeting in Palo Alto yet. We’re waiting on the CFO to reply. We may not know until the day before.
  5. What are you planning to do there? Read a lot and sit by the pool. Or shop for cool things to take back home. To learn how to kite surf. To take a cooking class. To attend a conference and network.

With so many possibilities, essentially the only time the typical travel booking site works for many of us is when we have perfect information about the 3 attributes available: Destination, Dates, and Number of Travelers. And that can be difficult to nail down.

Which is why, at WayBlazer, we spend a lot of time obsessing over natural language processing and chatbots.

Trip Intent means different things to different people

Travel is a very expensive, precious and, yes, personal thing. You could ask 1,000 people to tell you about an approaching trip and you’ll get 1,000 completely different descriptions. The information provided against the 5 attributes of a trip will differ from one person to the next.

It is therefore foundational to our approach at WayBlazer that we provide travelers a way to describe their trip without restricting their inputs to boxes or dropdowns or bubbles. Natural language, for us, means just that: travelers need to have the room to describe, in their own words, what it is they’re looking for. And we, in turn, need to understand those descriptions, regardless of the words chosen or the order in which the information is provided. It’s a challenging endeavor, but it is essential to understanding trip intent.

We also need to be infinitely flexible and intelligent in handling each interaction. 71% of sessions with one of our hotel brand customers begins with the traveler describing where they want to go. But the specificity varies greatly. Some travelers know the exact city they’re going to. Others search by region or state. Some search by continent.

We built a chatbot to understand more about a traveler’s trip intent. When we got started with Leading Hotels of the World, one of our first customers, we had a single line of natural language input. Most sessions began by addressing only one attribute, often times a destination-based search (ie, “I’m interested in Mexico”). As you’ll see in the next section, our technology thrives when we have more travel intent, and with Leading Hotels we had no mechanism to derive that.

So we designed a chatbot to pull that out. Now when a traveler says “I’m interested in Mexico,” we’re able to ask the right follow-up question to pull out more intent. One of the follow up questions to a destination-based search might be “Who are you going with?” When someone replies “With my wife and 2 kids” or “Solo” we have a lot more trip intent with which to personalize the recommendation.

Show me you understood my Trip Intent

Understanding intent is only half of the equation. We take the information gathered from the traveler and show them why a set of recommendations is right for their trip. This is key to reducing purchasing friction.

WayBlazer’s personalization at work with Emirates Vacations

WayBlazer’s personalization at work with Emirates Vacations
Travelers need the following evidence, at a minimum, to know they were understood:

  • Visual proof: I asked for a hotel with a pool or a rooftop bar or a romantic dinner on the beach, or perhaps a place to host a meeting or print copies of a presentation. So don’t show me the image of the lobby or parking lot. Show me the images that reflect my trip intent first. Even better, understand what it is I’m looking for and show me other things I might care about as well. People that are interested in the spa are likely interested in the gym, for instance. So after showing me what the spa looks like, show me the gym even though I didn’t ask for it
  • Written proof: I’m taking my family to San Diego. Explain to me, by prioritizing the descriptions or reviews, why this hotel or vacation package or restaurant is optimal for me. Explaining or better yet showing its proximity to things families like to do in San Diego, such as visiting the zoo or Legoland
  • The sort order: A list of recommendations is a ranked list. Tell me why one recommendation came before another. What makes the first recommendation better than the second? Why is it better for my trip?

Be where your customers are

Our approach to personalization at WayBlazer doesn’t end with understanding the traveler’s intent and showing them relevant recommendations. We believe you need to be wherever the traveler is, whether on your website, mobile app, or even in display ads or 3rd party platforms like Facebook Messenger.

We’ve brought our chatbot and personalized recommendation engine to any platform that will support it…through the direct channel desktop site with customers like cheapoair.com, through voice search on IHG’s mobile app, through chatbot ads with customers like Emirates Vacations, and this month through the Facebook Messenger page of 2 of the largest hotel brands in the world. Consumers today demand convenience, and part of that means reaching their brands of choice wherever and whenever they want.

It’s another form of personalization we will always aim for at WayBlazer.

Conclusion

Our approach to personalization at WayBlazer is grounded in the belief that we can reduce travel booking friction by understanding the traveler’s intent and reflecting that intent in the recommendations. And we’ve seen amazing results. For one of our hotel brand customers we’ve increased conversion by over 17% and our chatbot ads are increasing engagement by 87% over the standard approach.

Those outcomes goes way beyond incremental revenue for a travel company. It reveals a new model for travel planning and booking, one that addresses a clear pain point for travelers.

The ability to sit down at your computer and book a trip on your own time and without assistance was a major shift for the travel industry in 1996. Today, 22 years later, the ability to describe your trip in your own words and make booking decisions much faster is, in our view, the next shift.

Related Posts

  Photo of a business man inside a hotel room

AI Strategy for Hotels goes Beyond Alexa

Proposals for new technologies are hitting hoteliers’ desks at breakneck speed: mobile kiosks, mobile keys and concierges, artificial intelligence (AI), virtual reality, even robotic staff. It’s a lot to digest when you’re tasked with figuring out wh…

WayBlazer
The Intersection of AI and Travel
  Photo of snorklers

What Does Personalization Even Mean Anymore?

In the early 1900s, psychologists coined the term semantic satiation, the phenomenon when a word or phrase is repeated so much that it temporarily loses meaning. The listener perceives it as repeated meaningless sound.

WayBlazer
The Intersection of AI and Travel