SaaStr AI规则：国际收入达5%即应投入，含落地清单
You’ll Probably End Up With 40% Of Your Revenue from “International.” Invest Earlier There.
给出了明确的触发阈值（5%）和极低成本的执行清单，创业者可直接对照自家数据检查并行动。
53% of Figma’s revenue is international
Figma 53% 的收入来自国际业务
49% of Cloudflare’s revenue is international
Cloudflare 49% 的收入来自国际业务
49% of HubSpot’s revenue is international
HubSpot 49% 的收入来自国际业务
46% of MongoDB’s revenue is international
MongoDB 46% 的收入来自国际业务
36% of Twilio’s revenue is international
Twilio 36% 的收入来自国际业务
26% of Snowflake’s revenue is international
Snowflake 26% 的收入来自国际业务
20% of Okta’s revenue is…
Okta 20% 的收入来自……
— Jason ✨👾SaaStr.Ai✨ Lemkin (@jasonlk) September 5, 2026
— Jason ✨👾SaaStr.Ai✨ Lemkin (@jasonlk) 2026年9月5日
Amjad Masad announced Replit’s London office the other day. It’s the company’s first international office.
Amjad Masad 前几天宣布了 Replit 的伦敦办公室。这是该公司首个国际办公室。
Replit was founded in 2016. It has more than 50 million users and is on track for $1 billion in run-rate revenue by the end of this year. The $400M round at $9B was earmarked in part for expansion in Europe, Asia, and the Middle East.
Replit 成立于 2016 年。它拥有超过 5000 万用户，预计今年年底年化收入将达到 10 亿美元。估值 90 亿美元的 4 亿美元融资轮次中，部分资金用于在欧洲、亚洲和中东地区的扩张。
So: ten years, 50 million users, a billion in run-rate, and the first office outside the US opens now.
所以：十年时间，5000 万用户，十亿级别的年化收入，而第一个美国以外的办公室现在才开设。
It’s the normal pattern, of maybe waiting … just a bit too long ;). And we pulled the numbers on a dozen public B2B companies:
这是正常的模式，也许只是等待的时间……稍微长了一点 ;）。我们统计了十几家公开交易的 B2B 公司的数据：
Eight of twelve are at 40% or higher. But many of them got to 20%+ international before they put a single person on the ground, because the product was available everywhere and someone in Berlin or São Paulo found it.
十二家公司中有八家达到或超过 40%。但其中许多公司在派驻任何人员之前，其国际收入占比就已超过 20%，因为产品随处可用，柏林或圣保罗的用户发现了它。
But if you put a great team there, you grow even faster.
但如果在那里组建一支优秀的团队，增长速度会更快。
85% of Figma’s users are outside the US. 53% of the revenue is.
Figma 85% 的用户在美国以外。53% 的收入也来自美国以外。
Figma has now opened a Bengaluru office, added local data hosting and governance for India, and localized for Brazil. India is its second-largest market by monthly active users. All of that is recent. But the gap it’s addressing was visible in the data for years before anyone acted on it.
Figma 现已开设班加罗尔办公室，为印度提供本地数据托管和治理，并针对巴西进行了本地化。印度是其月度活跃用户第二大的市场。所有这些举措都是近期的。但它所解决的问题在几年前就已在数据中显现，只是当时无人采取行动。
MongoDB 45%-46% international for 3+ years, Datadog 28%-30%
MongoDB 国际收入占比连续三年保持在 45%-46%，Datadog 为 28%-30%
- Cloudflare, sitting near the top of this table at 49%, grew US revenue 41% last quarter against EMEA at 30% and APAC at 32%.
- Okta’s US revenue grew 12.6% while non-US grew 9.1%.
- Klaviyo: international up 33% versus 22% in the US
- HubSpot: international up 23% versus 17% domestic
- Snowflake: non-US up 39% versus 34% US
- Figma: international up 48% versus 46% total
- Cloudflare 以 49% 的比例位居本表前列，上季度其美国收入增长 41%，EMEA（欧洲、中东和非洲）地区增长 30%，APAC（亚太）地区增长 32%。
- Okta 的美国收入增长 12.6%，而非美国地区增长 9.1%。
- Klaviyo：国际业务增长 33%，而美国业务增长 22%
- HubSpot：国际业务增长 23%，而国内业务增长 17%
- Snowflake：非美国地区增长 39%，而美国地区增长 34%
- Figma：国际业务增长 48%，而总业务增长 46%
Klaviyo added regional hubs in Dublin and Singapore, and international went from 39.5% to 42% in a year
Klaviyo 在都柏林和新加坡增设了区域中心，一年内国际收入占比从 39.5% 上升至 42%
Klaviyo’s US revenue was $215.8M of $370.6M last quarter. A year earlier the international share was 39.5%. International revenue grew 33% against 22% in the US, and the mix moved almost three points in four quarters.
Klaviyo 上季度的总收入为 3.706 亿美元，其中美国收入为 2.158 亿美元。一年前，其国际收入占比为 39.5%。国际收入增长了 33%，而美国收入增长了 22%，在四个季度内，收入结构比例移动了近三个百分点。
Two hubs. Not a global rollout. And it moved a number that had been flat for most of the companies on this list.
两个枢纽。并非全球推广。而且它推动了一个在该列表中大多数公司都长期持平的指标发生了变化。
Snowflake at 26% and Okta at 20% are the direct version of the same problem
Snowflake 的 26% 和 Okta 的 20% 是同一问题的直接体现。
For self-serve products, underinvestment means leaving free demand unconverted. For enterprise sales-led products, it’s more literal: revenue only appears in a country after you’ve staffed it, so the international share is a direct readout of headcount decisions made two years ago.
对于自助服务产品，投资不足意味着未能转化免费需求；对于以企业销售为主导的产品，情况更为直观：只有在当地配备人员后，收入才会出现，因此国际收入占比直接反映了两年前做出的人员配置决策。
Snowflake and Okta aren’t at 26% and 20% because international demand is weak. They’re there because the deal requires people on the ground and the people aren’t there yet. Snowflake’s non-US revenue is now growing faster than its US revenue, which is what happens once you start hiring.
Snowflake 和 Okta 的国际收入占比分别为 26% 和 20%，并非因为国际需求疲软。而是因为交易需要当地人员支持，而目前人员尚未到位。Snowflake 的非美国收入现在增长速度快于其美国收入，这正是开始招聘后会出现的情况。
We should have gone much harder much earlier. One of the many mistakes I’ve made. Like many things it’s just focus and priority! https://t.co/DDFghbNBGb
我们应该更早、更大力度地推进。这是我犯下的众多错误之一。与许多事情一样，关键在于专注和优先级！https://t.co/DDFghbNBGb
— Todd McKinnon (@toddmckinnon) September 6, 2026
—— Todd McKinnon (@toddmckinnon) 2026年9月6日
The SaaStr AI rule: once one country is 5% of revenue, lean in
SaaStr AI 法则：一旦某个国家的收入占比达到 5%，就应加大投入
Here’s the threshold I’d use. Not 20%. Not “when it’s material enough to justify the cost.” 5%.
以下是我会使用的阈值。不是 20%。也不是“当它大到足以证明成本合理时”。而是 5%。
At 5% with zero local investment, that country has already proven the three hard things: people there can find you, the product works for them, and they’ll pay through a buying process you never designed for them. You cleared the hard part by accident. What’s left is the cheap part.
在零本地投资的情况下，达到 5% 意味着该国已经证明了三个难点：当地人能找到你，产品对他们适用，并且他们会通过一个你从未为他们设计的购买流程进行付费。你意外地攻克了最难的部分。剩下的只是容易的部分。
Lean in at 5% does not mean a lease and twenty headcount. It means:
在 5% 时加大投入并不意味着租下办公室并增加二十个编制。它意味着：
- Local currency and local payment methods, not just a translated pricing page
- Support hours that cover their business day
- Pricing tiers set for that market
- One person on the ground in that timezone, whatever their title is
- Data residency if you sell to anyone regulated
- 提供本地货币和本地支付方式，而不仅仅是翻译后的定价页面
- 覆盖他们工作时间的支持时段
- 为该市场设定的价格层级
- 在该时区有一名实地人员，无论其头衔是什么
- 如果你向任何受监管的客户销售，则需满足数据驻留要求
That list costs a fraction of what founders assume, and it’s the difference between Klaviyo moving three points in a year and MongoDB printing the same number three years running.
这份清单的成本远低于创始人的预期，而这正是 Klaviyo 一年内移动三个百分点与 MongoDB 连续三年实现相同数字之间的差异所在。
The mistake is treating 5% as too small to warrant investment. 5% with no investment is the strongest signal you will ever get about a market. It’s demand that survived your neglect.
错误在于认为 5% 太小而不值得投资。在没有投资的情况下达到 5%，是你所能获得的关于某个市场的最强烈信号。这是在你忽视下依然存活的需求。
When To Open Your First Int’l Office? Maybe When You Have Even Just $1.5m in ARR There (Updated)
何时开设第一个海外办事处？也许当你在那里仅有 150 万美元 ARR 时（更新版）
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力