Happy Go Makes Members Happier with Improved Loyalty Platform on AWS

By migrating to AWS, Happy Go has improved the rating of its mobile app from 1 to 4.3 out of 5 stars and delivers in-app response times of 500 milliseconds. Happy Go is Taiwan’s leading bonus point member platform, offering customers rewards for purchases with partner merchants. The company uses Amazon ElastiCache to maintain low latency, Amazon Kinesis Data Firehose to transform data for streaming, and AWS Lambda for serverless provisioning. 

In Taiwan, shopping brings happiness—at least for the 10 million cardholders who are members of the Happy Go bonus point platform. Established in 2005, Happy Go partners with department stores and major retailers, such as Sogo and FamilyMart, to reward their customers with points that can be redeemed for purchases. Happy Go has the highest market penetration of point cards in Taiwan, with one in two people registered as members of its platform. Its bonus point platform is also available in China and Japan.

Before 2019, Happy Go was only available as a physical card, with servers running in an on-premises data center that supported all merchant transactions. That year, the company embarked on a digital transformation and began developing a mobile app that members can use to redeem and receive points at various merchants’ point of sale. The business built its app on premises but decided to migrate most of its workloads to the cloud shortly after—a decision prompted by the need to support its overwhelming volume of transactions. Happy Go chose to migrate to the Amazon Web Services (AWS) Cloud because one of its founders had experience with AWS and valued the platform’s quality of customer service. “We believed AWS was the best choice to proactively support us for a swift cloud migration,” says Rodem Chung, IT director at Happy Go.

One of the challenges Happy Go faced with digitalization was ensuring dynamic response times for its app. Handling application traffic spikes during promotional events and prime shopping seasons, as well as matching its members’ points with merchant services, must be instantaneous. “Matching, certifying, and authorization have to be fast, which imposed a huge burden on our initial on-premises infrastructure,” explains Dragon Chen, IT senior manager at Happy Go.

The company implemented Amazon MQ managed message queueing service for Apache ActiveMQ to “park some events for processing” and reduce the stress on its on-premises servers. It also relies on Amazon ElastiCache to power real-time app responses. Using these services has enabled the company to maintain latencies of 500 milliseconds, which is below the internal service level agreement of 1 second.

Developers have also recently begun taking advantage of AWS Lambda to run and scale code automatically in a serverless manner, which saves time by eliminating the need to provision servers. AWS Lambda is currently used during merchant promotions when vouchers for cash or popular merchandise are offered to cardholders. During such periods, redemption happens simultaneously in merchant branches across the country, and the Happy Go app must provide instant updates on the status of vouchers or products left in stock. “Using AWS Lambda for voucher promotions has reduced the time spent on coding and related development procedures by 50–70 percent,” Chen explains. In addition, Chen also hopes to extend the use of AWS Lambda to all Happy Go app services and is currently running feasibility tests.

Happy Go claims to have the largest transaction database in Taiwan, so it can easily track customer spending and consumption patterns to provide deep consumer insights. The company offers complimentary data and analytics services—known as Go Survey Market Research Consulting—to large partner merchants, as well as to firms outside the Happy Go network for a fee.

To extract behavioral data from the Happy Go app, developers rely on Amazon Kinesis Data Firehose, which they use to centralize data on the AWS Cloud and transform it into a structured format for streaming. Most analytics functions at Happy Go are still performed in-house with an on-premises data warehouse, and the company uses AWS Direct Connect to securely transfer streaming data from the AWS Cloud to Happy Go’s servers. The company plans to build AI-based analytics on the AWS Cloud to deliver personalized customer recommendations in the app.

Throughout migration, AWS representatives regularly visited Happy Go and made suggestions on infrastructure improvements. “Any time we faced a challenge, AWS always provided a professional and timely response and solution,” Chung says.

Happy Go employs just two engineers to manage its infrastructure and has been able to maintain this staffing level throughout its digital transformation. “If we had tried to develop and run everything in-house, we would have needed at least four to five people to support the current implementation,” Chen says. “AWS provides a lot of services to support digitalization and improve our app—services that would have been difficult for us to create internally. With AWS, we can invest more of our resources into improving the customer and merchant experience, which is the primary purpose of Happy Go.”

Customer satisfaction is on the rise, as is evident from Happy Go substantially boosting its app’s online rating in one year since migrating to AWS. The Happy Go app formerly received just 1 out of 5 possible stars on the Android app store, but the rating has since increased to 4.3 stars. According to Chung, this is the highest ranking in Taiwan for any loyalty card or point system app. He attributes the improvement to the ease of using the app for point redemption and to quick response times for logging in and out. Since migration, the app has had 99.999 percent uptime, which has contributed to its success.

Customers have also benefited from more regular updates to in-app features. “By using AWS, we are able to develop and introduce a new app feature every month, compared to at least three to five months in the past. We’ve had a great experience working with AWS and are highly satisfied with the level of service provided,” Chung concludes.

To learn more, visit aws.amazon.com/serverless.