Make searching simpler

In any given location there can be hundreds of hotel rooms available. Finding the one that ticks all the boxes and comes in on budget is no mean feat.

To make it easier for our customers to find the perfect hotel, we tailor our search results based on insights gained by analysing booking data and customer ratings for each property, making the most popular hotels easier to find.

We also use analytics data to help us understand what our customers are searching for when they enter our website. For instance, if someone has just been looking at ratings for hotels in Paris on Trip Advisor, it’s highly likely that they’ve visited us to book a hotel there, so it makes sense to present them with a tailored page of hotels in that location.

Equally, someone who is searching for budget hotels is unlikely to find a page of five-star accommodation particularly helpful. 

Avoid interruptions en route

It’s a fact of life for retailers that some people will always change their mind and go elsewhere at the very moment when you are expecting to make a sale.

By analysing exit rates across our site, we are able to identify where things are going wrong and make changes to improve the customer journey.

For instance, you could find that a bug in the system causes problems each time a page is loaded or that something else is giving people cold feet. For an online retailer, this could be a limited selection of payment methods or a lack of transparency over terms and conditions or delivery rates. 

Using big data to analyse the points at which people exit your site can also unearth some interesting differences between your customers. For example, we found that drop-out levels were higher in some countries than others.

After a little investigation we found that drop-out levels were highest in the countries where online shopping is least established and customers need a bit more reassurance that it is safe to enter their payment details.

With this insight in hand we were able to review the customer journey in these countries to assure our customers that our payment system is secure.

Get to know your customers 

Since launching our mobile site in 2010 and our first apps a year later, we’ve found that there are big differences in how people book hotels on mobile and tablets, compared to the web.

Analysis of our booking data has shown that mobile users typically make same-day booking for short stays at nearby hotels, whereas tablet users typically act more like our website users and spend a longer period planning longer trips further afield. 

This insight has helped us to optimise the design of our apps and streamline the booking experience for mobile users. Now, when you open our app, you are presented with a selection of hotels in your location that have rooms available that night. 

Streamlining the customer journey in this way, doesn’t just make finding a hotel easier, it gives users a great first experience, which we have found is especially important on mobile.

Data from our apps shows that many of our customers will download our app and keep it on their phone until they need to book a hotel room, which can sometimes be several months later.  

Room to learn

As with all companies that have embraced big data, we are still learning about how to generate the most value from it.

Although big data is very useful for identifying macro trends, we have found that some of the most interesting insights come from findings that affect much smaller groups of people. To identify these smaller insights, we look for anomalies when analysing data and dig down to try to work out what the cause is. 

However to do this, you must have a full and accurate data set. This can sometimes be easier said than done. One of the challenges that we’re facing is how to join up the dots between connections, when customers use multiple devices to plan and book trips away.

We know from speaking to our customers that many of them will research hotels through one channel before switching to another to make a booking, but this journey is very hard for us to track and analyse if a customer hasn’t been logged in throughout the entire process. 

Of course, that’s not to say that it’s impossible. We’re looking at how we can better understand the multi-device customer journey and how we can change the design of our website and apps to make it easier to tie data from the two together. 

At times big data can be a big challenge, but the insights are well worth fighting for.