Localization at Scale

In 2021, Spotify announced support for 36 new languages at its Stream On event, bringing the total number of supported languages on its platform to 62. The effort, called Scaling Translations, was enabled by a combination of dedicated engineering work and a revamped localization process. The release was delivered in six months, on time and within budget, with the launch beginning the day after the public announcement.

Why expand language support

Spotify was available in 26 languages when the project began. The rationale for the expansion was reach: supporting more languages was seen as a way to engage users whose native languages were not yet covered and to grow the total addressable market. The bet was that localization would be among the strongest drivers of growth, in line with what industry peers have reported.

The historical approach to adding a language had been tied to launching service in a new country or region. That meant the underlying architecture was never adapted to make new language launches easier. As the tech stack grew, adding even a single language became complicated — a manifestation of Conway's law in which the organization's structure is reflected in its software.

Preparing the ground technically

Before funding and staffing were secured, the team spent two months assessing the technical landscape. They interviewed software engineers, product managers, engineering managers, and business-side colleagues across the company. They also examined source code repositories, system diagrams, and production metrics and telemetry. The findings were compiled into a 25-page technical scoping document that covered existing systems, their interactions, use cases, business metrics, required capabilities, and the expected effort.

After the direction was approved, a small dedicated engineering team was hired six months before the scheduled release date. This marked the first time dedicated engineers worked alongside the localization team on a language-launch project. Their work included defragmenting the tech stack, automating linguistic quality-assurance tasks for mobile apps, and using automation to update 30+ backend services in a coordinated fashion to support the new languages.

The localization approach

Historically, launching in a new market with a new language took significantly more time. To pull off 36 languages in six months while maintaining quality, the localization team focused on six areas:

  • Minimum viable product: The product was stripped down to what matters most to users in local markets, with a philosophy of delivering quickly, learning, iterating, and expanding where needed.
  • Nimbleness: Hard decisions were made early to keep progress moving with minimal resources and ramp-up time, avoiding analysis paralysis.
  • Vendor partnerships: Vendors were included from the start. They contributed market and language research, helped build the business case, participated in strategic decisions, and built out linguistic teams and quality processes. They were given context and visibility into the project's goals and approach.
  • Cross-functional support: Dozens of internal teams had to be identified, aligned, and bought in. Their roadmaps and priorities had to be reconciled with the project timeline.
  • Scope flexibility: As awareness grew, stakeholders began requesting their areas of the product be included. Scope was managed tightly from a prioritization standpoint, but changes were still part of the process throughout.
  • Communication: Dedicated Slack channels and Jira dashboards for bug reporting from localization testing were set up. A regular meeting cadence for critical workstreams and daily stand-ups helped maintain visibility and collective decision-making.

Outcomes

The project had a positive business impact on the user base and drastically reduced the percentage of users not served in their preferred language. It also made the product more diverse and inclusive from a language perspective.

Scaling Translations demonstrated that challenging the standard approach to international expansion and product launches can help a business move faster. The team believes it proved that launching at speed does not have to come at the expense of quality or cultural relevance in the user experience. The project also reinforced the value of remaining nimble, focusing on the most important problems first, iterating with fast feedback loops, and managing scope and stakeholders tightly.

The initiative required a change in mindset and ways of working. But with a solid business case, organizational alignment, and the right use of technology, it succeeded — and has opened up new globalization opportunities that hadn't previously been considered feasible.