作为数据科学家重新定义影响力

In domains where accuracy and visibility are critical, data science looks different. The work often involves modeling event lifecycles, reconciling data across systems, and building simple tools to make things observable. It’s less about running A/B tests and more about understanding how complex systems behave.

在准确性和可见性至关重要的领域，数据科学看起来不同。该工作通常涉及建模事件生命周期、跨系统协调数据，并构建简单的工具来使事物可观测。它更少涉及运行 A/B 测试，而更多涉及理解复杂系统的行为方式。

I encountered this while working on Figma’s Billing infrastructure, but the lessons apply to many high-stakes, complex domains. Here are five takeaways that can help data scientists have more impact, regardless of the problem they’re working on.

我在从事 Figma 的 Billing 基础设施工作时遇到了这个问题，但这些经验教训适用于许多高风险、复杂领域。这里有五个经验教训，可以帮助数据科学家无论从事什么问题都能产生更大影响。

Think of data science as a full-stack discipline

将数据科学视为全栈学科

The ways DS (data science) can add value varies dramatically from team to team. Some teams need rigorous experimentation frameworks, some need deep product analyses, and others rely on strong data modeling or instrumentation work. Figma hires full-stack data scientists because a single role often needs to span most of these modes—sometimes all within the same project.

DS（数据科学）为团队增加价值的方式在不同团队之间差异巨大。有些团队需要严谨的实验框架，有些需要深入的产品分析，其他团队则依赖强大的数据建模或仪表化工作。Figma 招聘全栈数据科学家，因为单一角色通常需要涵盖这些模式中的大多数——有时在同一个项目中全部涵盖。

Unlike more traditional product surfaces, Billing sits somewhere between a user-facing product and a backend system. Reliable, accurate billing directly affects the customer experience and their trust in the platform. Over time, we’ve learned that supporting Billing with data science means building domain expertise, partnering closely across functions, and developing tools that can explain, verify, or clarify system behavior. Tasks like experimentation and opportunity analysis are important, but they represent a smaller slice of the work.

与更传统的用户面向产品表面不同，Billing 位于用户面向产品和后端系统之间。可靠、准确的 billing 直接影响客户体验及其对平台的信任。随着时间推移，我们了解到用数据科学支持 Billing 意味着构建领域专业知识、跨职能密切合作，并开发可以解释、验证或澄清系统行为的工具。诸如实验和机会分析等任务很重要，但它们仅代表工作的一小部分。

I captured this sh...