Based in Amsterdam, Travelbird was founded in 2010 with the mission of taking the stress out of planning vacations online. It found that around half of vacation buyers did not know where they wanted to go, on average spending 55 minutes browsing across 17 websites before booking. Travelbird wanted to simplify the experience by providing well-packaged, great-value deals designed to inspire travelers with a “booking to back-home” experience comprising travel, accommodation, and extras like dining, entertainment, museum tickets, or spa packages. It has grown rapidly from a 50-person operation in 2010 to 700 people today, and operates in 17 countries.
As a web-based company Travelbird has always relied on cutting-edge technology to host its website and platforms, and deliver deals to customers. At the start of 2014, it had planned to engage in television advertising, and wanted to ensure that it had a scalable infrastructure to support the expected influx of new customers. It looked to Amazon Web Services (AWS) as a possible path into the cloud.
In addition to building high scalability into its infrastructure, Travelbird also wanted to analyze and derive business intelligence (BI) from the data in its own systems running on AWS. It hired Rob Winters to head up this operation, who leads a team to support A/B testing on all products, and provide centralized data so the company can explore it in a meaningful way to gather insights.
Travelbird CTO Philipp Wassibauer recalls, “We looked at other cloud providers, but chose AWS since it is the most mature cloud platform, and supported the technology and components we needed.” The travel firm had been using Amazon Simple Storage Service (Amazon S3) for several years so it was already familiar with AWS and its technologies. After spending one month preparing its infrastructure for the move, Travelbird migrated its systems to AWS in just one night.
Today, it runs its entire infrastructure on AWS, everything from web and database servers to BI applications. Along with Amazon S3, it uses Amazon Elastic Compute Cloud (Amazon EC2) instances, runs its databases with Amazon Relational Database Service (Amazon RDS), and uses Amazon Redshift for data warehousing.
When Rob Winters joined the company, he and his team were able to quickly set up BI processes using existing services like Amazon EC2 and Amazon Redshift. They also added services like Amazon Kinesis for real-time data processing over a large, distributed data stream; and the Hadoop-based Amazon Elastic MapReduce (Amazon EMR) to distribute data and processing across a resizable cluster of Amazon EC2 instances.
Winters says, “We found it fairly straightforward to get the environment up and running. My team and I were able to get it up and running without any support from AWS at all.”
By setting up a series of BI processes on the AWS infrastructure, Winters and his team have been able to gain new insights into the company’s data, and drive business results through better-informed decision-making. After just four months, one-fifth of the company’s employees are already using this data every day.
Winters says, “For the first time we can see all products online and their inventory levels. A few weeks ago that was impossible. We identified that about 10 percent of our products were not actually sellable. By cleaning those up we can avoid directing people to products that aren’t available.”
By building data pipelines and reports in AWS, Travelbird has reduced the workload for many of its managers. This has freed up time to work on more valuable projects. According to Winters, “Before, managers had to make manual adjustments to reports to determine who got credit for a sale, for example. Now, we can do it all automatically. For some managers that means a 25-30 percent reduction in their workload.”
It’s also provided insight into customer preferences—specifically what people look for and when. This means Travelbird account managers can tailor vacation packages more accurately for certain times of year, right down to the number of nights each package lasts. “Running our BI applications on AWS has changed our operational model,” Winters says. “It’s vastly improved our ability to find the best customized deals for our customers.”
Winters also lauds the speed at which his team can deploy, experiment with, and scale new features. He gives one example: “We recently needed to upgrade a reporting server. In a company with a fixed infrastructure you’d need to go through a long acquisition and implementation process. With AWS, we were able to upgrade in three minutes to match increased demand on the server. We’re definitely able to get our BI features out faster as a result.”
To learn more about how AWS can help you manage Big Data, visit our Big Data details page.