interview

Travel group Expedia aims to predict your holiday plans

Bill Goodwin

Travel group Expedia is developing technology that will predict the needs of its customers before they phone to book a holiday or query existing hotel bookings or flight plans.

The $4.7bn travel company aims to forewarn its call handlers on the likely reason for a call from a customer before they pick up the phone, by piecing together data about their journey, travel plans and travel history.

463735099 (2).jpg

Expedia, which owns brands including Hotels.com, Hotwire and Trivago, also plans to alert travellers to flight delays, changes in their itinerary and other problems by text, and offer them alterative travel options.

The move is part of a drive by the travel company to focus on customer service, at a time of increasingly intense price competition among travel providers, says Mikko Ollila, senior manager for product management at Expedia, in an interview with Computer Weekly.

“You can actually service the customers better in a way that makes the traveller feel safe and confident. I think that is one of the areas where we can truly fight for customer loyalty,” he says.

Intelligent travel data

Expedia, which is investing several million dollars a year in the project, plans to use tools from Pegasystems to provide customers with tailored advice and services.

The quicker the agent can get through the entire transaction, the happier the customer usually is

Mikko Ollila, Expedia

The company is using a combination of its own Java-based in-house booking software and Customer Process Manager (CMP) case management software from Pegasystems to manage customer calls and bookings.

It plans to build on the data analytics capabilities of CMP to predict what deals they may be looking for in the future by analysing their travel and booking histories.

Expedia’s ultimate aim is to eliminate the need for customers to call the company by anticipating their needs in advance.

If a flight is cancelled or delayed, or a customer needs to change a hotel reservation, the system will contact the traveller and offer them two or three alternative choices.

And if customers do call in, rather than requiring them to go through their entire travel history the call centre agent will already know why they are calling and have options lined up to help them, says Ollila.

“The representative should be able to ask, 'Hey, is this so and so? Are you calling about your flight that seems to be changing on Wednesday?',” he says.

The software will help call centre staff suggest intelligent options for people who call Expedia with vague requests like, “I want to go to Las Vegas, what have you got for me?”.

And it will be able to predict what customers are likely to book in future and offer them appropriate deals. For example, if a family with two young children register on Expedia, it's not a big leap to assume they might be interested in a trip to Disneyland or SeaWorld, says Ollila.

Expedia's core software

Expedia uses two core software applications – Voyager and Pegasystems' Customer Process Manager (CMP).

Voyager is a java-based platform, developed in-house, which allows its 9,000 call centre staff to book flights, hotels and other services for customers.

Pegasystems' CMP is case handling software which Expedia uses to automatically record and document its interaction with each customer.

Expedia has been using both pieces of software since 2009.

Faster transactions mean happier customers

The company first rolled out Pega’s CPM software, in 2009, to manage and document incoming calls and emails to call centre staff.

The software is able to record details of the call automatically, eliminating the need for call centre staff to take copious notes.

This means staff can process calls much more quickly, and that means happier customers, says Ollila.

“The quicker the agent can get through the entire transaction, the happier the customer usually is. They don’t want to hang out on the phone and talk to us,” he says.

It is surprising how many call centres still rely on customer service reps taking extensive notes during phone conversations, he adds.

Keeping the software simple

Ollila says he has learned a number of lessons from implementing Pega’s software.

One of those is the importance of recruiting software engineers who have a real passion for the technology. 

“We did not go out and say, 'We are going to have this system that combines Java and Pega. Now let's go out and recruit the optimal, most passionate engineering workforce for this.' We did not do that, but we should have done that,” he says.

And although Pega’s software is designed to be easily customisable, Ollila learned that it's better not to customise too much.

The problem is that each time you customise Pega, you have to repeat the work whenever you upgrade to a new version of the software.

“We probably over-customised our system along the way a little bit, then had to scale back in favour of more longevity and scalability,” he says.

Balancing the short-term demands of the business with the longer-term need to get the technology right has also been a challenge, says Ollila: “You have to fight for the long term.”

Experimenting with data analytics 

Ollila plans to begin work on expanding the data analysis capabilities of Expedia, using Pega’s tools, from next year.

He is keen to try out projects on a small scale, let them fail if necessary, and experiment with different approaches to improving customer services. This is one of the most challenging parts of the work, he says.

“The challenge is to build it in a way that gives you a longer-term, relentless capability to experiment and know exactly what the right message is for the right customer,” says Ollila.


Email Alerts

Register now to receive ComputerWeekly.com IT-related news, guides and more, delivered to your inbox.
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
 

COMMENTS powered by Disqus  //  Commenting policy