I’ve always been fascinated with data science. Reading books like Freakonomics and Predictable Irrationality deepened my curiosity for understanding how people think and behave given specific circumstances. At ID90 Travel, we leverage data analytics to paint a picture of our customers’ experience across the entire application. We focus on what customers do, why they do what they do, where customers come from and where customers spend the most time.

When we talk about analytics, it’s not just about measurement – it’s about measurement with purpose. We collect and use data to optimize products, identify areas for improvement and inform upcoming business decisions. Most importantly, we use data to tell “user stories” – insights into how our customers use our product, and which of their needs aren’t yet being met – to help us make the most impactful improvements for every user’s experience.

Making Flights More Usable

Our non-revenue Flights product is our bread and butter. With nearly a million flight searches each month, we’re committed to getting you from point A-to-B as easily as possible. When we started monitoring user behavior and interactions on our site as part of our business intelligence strategy, we uncovered fascinating patterns which had previously gone unnoticed.

Below are just a few of the recent improvements we’ve released to make our Flights product more enjoyable.

Making it Even Easier with Flight Filters

It was a cold and windy Wednesday morning, the first cup of coffee not yet brewed, when the warning alarms suddenly cried out throughout the office. Our user behavior analysis had noticed a troubling trend: users searching for flights were returning back to the original search form again. And again. And again.

Not a great experience – something wasn’t working the way they needed it to.

After some digging (and a few cups of coffee,) we found the sticking point. We didn’t provide auto-updating filters available on the search results page, and because of this users were forced to input all of their criteria into the original search form. If nothing matched that very specific criteria, the user would have to go all the way back to the beginning and perform a new search.

The fix was easy and, as we soon found out, greatly appreciated.

Improved filters on search results

Now, customers can execute a very generic search and then filter their results to find the flight that most closely matches their needs – without having to start all over every time.

What Was That Airport Name Again?

Another improvement we rolled out recently was the ability to use “natural language” on flight search. Before this implementation, customers could only search for their destination with the specific IATA code of the destination airport.

Now, users can search for airports using terms like “New York”, “Dallas,” or “Seoul,” without having to know the code for a specific airport.

We believe in simplicity and familiarity. This experience reduces friction for our users, especially since they’ve become accustomed to this behavior with other commercial solutions.

improved flight search with regular expression


Hotels: Have it Your Way

We’re constantly learning from our customers. Understanding what’s most important to you is what’s most important to us. That’s why we’ve spent a lot of time gathering little bits of data to discover what makes you tick. We’ve found some really interesting things along the way too.

Our Hotels product is another area where we make constant improvements based on – you guessed it – tons and tons of data. We’re constantly learning from our users, and what they’ve taught us (and continue to teach us) about our Hotels product has resulted in some truly fantastic changes.

Location, Location, Location

improved flight search with regular expression
That old adage still rings true: our customers care more about where they stay in a given city than what the hotel comes with, or even how highly it’s rated. Not only do they care where they stay, but 10% of all neighborhoods selected are those near the airport! Our top 5 neighborhoods by search include:

  1. The Strip – Las Vegas
  2. Manhattan – New York
  3. Broadway – Times Square
  4. South Beach – Miami
  5. Los Angeles International Airport – Los Angeles

Okay, so our users want hotels near the airport – that’s blatantly obvious. But we didn’t stop there, because location’s just one factor. What kind of hotels near the airport are most important to our users?

Amenities to the Rescue!

When customers filter hotels by amenities, there is a clear winner: hotels with free airport shuttle. This makes intuitive sense. Most of our customers are airline employees who need a place to crash before their next flight, and calling an Uber to get from the airport and back is just one more unnecessary expense.

Our analytics shows customers really care about having free airport shuttle.

This, combined with what we learned about the location of preferred hotels, allowed us to make a monumental improvement in our hotel ranking algorithm. We now prioritize hotels closer to airports and those more likely to offer free airport shuttle.

What We Search For & What We Actually Book

Customers have a tendency to aim high when searching for a hotel. They want the absolute best hotel, with the absolute best price.

Don’t we all. But, as with most everything in life, often our eyes are bigger than our stomachs – or in this case, how much we’re willing to spend for one extra “star.”

We found most customers search for 4-star hotels, but most customers end up booking 3-star hotels. In addition, we found a correlation between hotel rating, average daily rate, and length of stay. Stands to reason, as the more luxurious your hotel is, the longer you’d like to stay there!

We also discovered a lack of interest in the hotel grid view (now removed). Lastly, we realized 75% of customers choose the first room, which is usually the absolute best price.

We took all of this information – every one of the million tiny data points – and used it to produce our next-generation hotel ranking algorithm. Today, when a user sees our “recommended” hotels for a specific destination, they see the properties that are most likely to meet all of their needs – the right price, the right amenities, the right location, and more.

Conclusion: Analytics Will Be Our Future

In this day and age, lack of data isn’t a problem anymore – if anything, it’s quite the opposite. The challenge now is taking every one of those million tiny data points, and using them to make a cohesive story. One that drives innovation and, at the end of the day, creates an experience that delights our users.

We take pride in leveraging data analytics to make smarter decisions that further enrich the experience for our users. As analytics evolves alongside our platform, we will continue to lead the pack in serving our customers and giving the non-revenue market exactly what it’s asked for all these years.

Author: Roger Ximenez

Roger Ximenez is the Director of Product at ID90 Travel. He has over 5+ years experience working alongside engineering teams to build the next generation of features. He is also the data and automation evangelist, promoting the use of data to streamline automation strategies.