Case Study: Etsy's 'Seamless' Payments Platform Migration
Etsy's engineering team has detailed its migration of its core payments platform using the database scaling system Vitess. The case study highlights the importance of deliberate orchestration, phased rollouts, and robust communication across engineering, product, and business teams to achieve a 'seamless' transition at scale.
The migration of Etsy's payments platform was a direct response to the company's explosive growth, which saw its gross merchandise sales skyrocket by 106.84% in 2020 to $10.28 billion. This surge in transaction volume pushed two of the company's core payments databases to their vertical scaling limits on the Google Cloud Platform, creating a high-risk situation that could have led to performance degradation or even lost transactions for sellers. The 18-month project, running from December 2020 to May 2022, was a collaborative effort between three key teams: the Payments Platform, Database Reliability Engineering, and the Data Access Platform. This cross-functional "squad" model is integral to Etsy's engineering culture, which emphasizes autonomy and shared ownership. These small, cross-functional teams are responsible for maintaining their own codebase and addressing technical debt as part of their regular workflow. To ensure a smooth transition, the engineering teams created a full staging environment that mirrored the production setup, allowing them to repeatedly test the migration process with a clone of production data. This meticulous preparation was crucial for building confidence across the teams and with leadership, demonstrating a commitment to minimizing risk for a system that is critical to seller livelihoods. The success of this approach meant the migration was completed with no downtime or impact on normal operations. The project involved moving 23 tables with over 40 billion rows from four unsharded databases into a single, horizontally scalable environment managed by Vitess. By migrating only the highest-volume tables, the team was able to reduce the load on the primary payments database by 60%. This not only provided the necessary headroom for future growth but also exemplified Etsy's engineering principle of targeted, high-impact initiatives. This infrastructure overhaul was part of a broader technology strategy at Etsy to support its rapidly expanding marketplace. During the same period, the company's active buyers grew to 95.5 million by 2024, with annual revenue reaching $2.808 billion. The ability to scale its core payments platform was essential to reliably process the growing volume of transactions from its global user base. The leadership approach within Etsy's engineering organization empowers managers to provide technical direction and strategic focus while fostering a culture of experimentation and data-driven learning. Engineering managers are expected to partner closely with product and analytics teams to ensure that technical projects directly contribute to business growth. This was evident in the Vitess migration, where the technical necessity of scaling the database was directly tied to the business's need to support its expanding marketplace. Communication and influence are key responsibilities for Etsy's engineering leaders. They are expected to "pitch the work their team is doing, both up and down the business, by convincing them of your vision." For a project of this scale, this would have involved regular updates to executive leadership on progress, risks, and the ultimate business value of ensuring a stable and scalable payments platform. The successful migration of the payments platform is a testament to Etsy's engineering culture, which combines a high degree of team autonomy with a focus on blameless post-mortems and continuous learning. This allows teams to tackle complex, high-stakes projects like the Vitess migration while minimizing disruption to the millions of buyers and sellers who rely on the platform.