Hay Increases New User Conversion Rates by 300% with Amazon EMR

As big banks grapple with how to securely digitize their operations, FinTechs are increasingly filling the gap for consumers seeking easy access to mobile spending accounts. One such FinTech is Hay, who launched to the Australian public in February 2020 as a mobile-first payments and money management solution.

Hay is a mobile app offering a Visa card and digital transaction account that users can top up via direct transfers from their bank accounts. The FinTech has also started building the B2B aspect of its business, Hay-as-a-Service (HaaS), selling its platform to financial institutions seeking to modernize legacy payment systems. Born on the Amazon Web Services (AWS) Cloud, Hay has the agility to quickly adapt its products to the needs of diverse customer segments.

In a FinTech world, companies that operate exclusively online are known as neobanks. Often described as the next wave of banking, neobanks are FinTech disruptors with a fully digital front end and backend. Hay is in the process of obtaining a Restricted Authorised Deposit-Taking Institution (RADI) license, which is the precursor to becoming a bank recognized by the Australian Prudential Regulation Authority (APRA).

While APRA encourages a cloud-agnostic approach for security in banking services, Hay’s founders worked closely with regulators to validate that a solution built entirely on AWS was appropriate for Hay. This solution would not only offer the same—if not higher—security, but also halve Hay’s time-to-market and reduce risk. “AWS is one of the world’s largest technology companies, and we believe the concentration risk of working this closely with such a company is far less than the risk of our engineers making mistakes trying to keep our banking platform aligned on different technologies,” says Jason Latham, chief information officer at Hay.

Hay built its infrastructure across multiple AWS Regions and Availability Zones to establish strong encryption and resilience. Its app adheres to the Payment Card Industry Data Security Standard (PCI DSS) and uses the Mambu cloud-native core banking platform. As an AWS Partner Network (APN) Select Technology Partner, Mambu subscribes to the shared responsibility model for clear designation of security and compliance roles.

The Hay app launched at the onset of the COVID-19 crisis when the lockdown in Australia was in effect. Despite a general decrease in consumer spending during the lockdown, the app attracted more than 9,500 users in its first six months of operations, largely from word of mouth. The company is growing 25 percent month-on-month, and management anticipates that figure to rise to 75 percent given the overwhelming response to its B2B offering.

By using a microservices architecture, Hay is now able to customize its offering according to each bank’s requirements. “We provide a universal adapter for financial institutions to connect to modern technology, even with legacy systems. Our modular approach allows financial institutions to upgrade their existing technology in small chunks, without a significant investment,” explains Andrew Laycock, chief executive officer at Hay.

To offer a positive customer retail experience, credit card companies such as Visa and Mastercard stipulate payment authorizations be completed within 200 milliseconds. However, banks typically don’t have the technology to stop a fraudulent transaction right away, and will only freeze a card when suspicious behavior is flagged post-purchase. In contrast, the Hay platform possesses a bespoke backend authorisation platform. Supported by Featurespace—a machine learning–based Risk Hub—fraudulent transactions are prevented from completing and such transactions are flagged for investigations that are enabled by Amazon Athena within Hay’s “single pane of glass” operational platform.

B2B customers also benefit from lower fraud rates on Hay transactions, which are made possible by the FinTech’s data aggregation expertise. Hay engineers have built a data lake using Amazon Simple Storage Service (Amazon S3) that pools information from the company’s users. “Our fraud risk engines have a large, highly diverse dataset to learn from. This allows us to leverage behavioral machine learning analytics to protect users across our platform from fraud attempts,” Laycock says.

The ability to process high volumes of data at a near real-time pace with Amazon EMR likewise enables rapid releases, and frequent iteration has become an inherent part of Hay’s DevSecOps practice. In its first six months of operation, Hay released 21 feature updates of its retail app and has since averaged a new release every two weeks. The company is constantly evolving to accommodate its users’ ever-changing lifestyles with its steady stream of new features.

Expanding business abroad is also part of the company’s plan, starting with Southeast Asia. “AWS offers a high level of flexibility and scalability, so we can handle infinite volumes of data as our business expands,” Laycock says.

Needless to say, data is at the center of Hay’s marketing strategy, with teams closely monitoring consumer behavior to improve adoption and retention rates. Shortly after launching, Hay observed a high rate of user drop-offs within a few days of retail users downloading the app. By using Amazon EMR in conjunction with third-party tools, the company was able to process and visualize target customer data to address issues and adjust the user experience accordingly. Within a week and a half of testing and making changes, Hay had increased conversion rates for new users by 300 percent.

Recently, Hay released its Tiny Big Rewards referral program to retail customers. The inspiration for this program, which provides referral bonuses and a 1 percent cashback on transactions below US$7.3 (AU$10), came from insights extracted from spending data. The company’s analysis showed that 40 percent of all transactions on the Hay app were below this dollar value.

Since implementing the program, the number of such transactions has increased by 20 percent. “We used data to better understand our customers’ spending habits and came up with a reward program that actually changed how people interacted with our product,” Laycock says.

Overall, Hay is confident in its AWS technology stack and is pleased with the service relationship. “We feel really supported by AWS. The after-sales support has been great, and I like the AWS egalitarian approach to clients. It doesn’t matter whether we’re a big bank or a tiny agency, we get the same high-quality service,” Latham concludes.