Hiring software engineers and engineering leaders from Big Tech (Part 1)
從大型科技公司招聘軟體工程師和工程領導者（第一部分）
A dive into why hiring Big Tech talent can be a bad idea for startups, a look at cases when it works, what makes it hard to recruit from Big Tech, and how to do it
深入探討為什麼對初創公司來說，招聘大型科技公司的人才可能是一個糟糕的主意，分析何時這樣做是有效的，招聘大型科技公司人才的困難之處，以及如何實現這一目標
Before we start: the Korean translation of The Software Engineer’s Guidebook is out! If you are based in Korea, you can get it from Hanbit Media (the publisher), from Kyobo, from YES24 and from Aladin. The book is a two-volume, 568-page edition. The book is a lot more than "just" the original book: it includes an appendix with insights from five prominent Korean developers, including engineers from large Korean companies and startup CTOs. This addition provides a local perspective and practical insights on software engineering. This is the first translation of the book – other languages like German, Japanese and Chinese will follow in the coming months!
在我們開始之前：《軟體工程師指南的韓文翻譯已經出版！如果您在韓國，可以從 Hanbit Media（出版社）、Kyobo、YES24 和 Aladin 獲得這本書。這本書是兩卷本，共 568 頁。這本書的內容遠不止於“僅僅”是原書：它還包括附錄，裡面有五位知名韓國開發者的見解，包括來自大型韓國公司的工程師和初創公司的 CTO。這一補充提供了本地視角和實用的軟體工程見解。這是該書的首次翻譯，德文、日文和中文等其他語言的翻譯將在接下來的幾個月內推出！
《軟體工程師指南》的韓文翻譯。點此獲取。.
There are many standout software engineers and engineering leaders in Big Tech, and it’s easy to assume that hiring them is a sure win for any startup and scaleup. But counterintuitively, recruiting techies from Big Tech is often very difficult for startups. Sometimes, it’s simply very hard to get tech professionals interested in a smaller company, even when they’re a good fit.
在大型科技公司中，有許多出色的軟體工程師和工程領導者，並且很容易假設聘用他們對任何初創公司和成長型公司來說都是一個必勝的選擇。然而，反直覺的是，對於初創公司來說，從大型科技公司招募技術人才往往非常困難。有時，即使他們非常合適，讓技術專業人士對小公司產生興趣也非常困難。
A few weeks ago, we dug into reasons why software engineers quit Big Tech. In this article, we look into ways to attract folks to startups.
幾週前，我們探討了 為什麼軟體工程師會離開大型科技公司。在這篇文章中，我們將研究吸引人們加入新創公司的方法。
For this piece, I talked with techies in senior roles at startups. Keeping identities anonymous, this deep dive covers:
在這篇文章中，我與初創公司的高層技術人員進行了交談。為了保護身份，這次深入探討涵蓋了：
Why Big Tech hires are often poor fits for startups
為什麼大型科技公司的招聘人員常常不適合初創公司When hiring from large companies doesn’t make sense
當從大型公司招聘不再合理時When it does make sense
當它有意義時Why is it hard to hire from Big Tech?
為什麼從大型科技公司招聘這麼困難？How to “poach” from Big Tech
如何從大型科技公司“挖角”
Part two of this mini-series will cover how to pitch opportunities to Big Tech folks, with advice from hiring managers at startups about their successful approaches.
本迷你系列的第二部分將探討如何向大型科技公司的人士推銷機會，並提供來自初創公司招聘經理的成功方法建議。
The Pragmatic Engineer deepdives related to this topic:
實用工程師深入探討與此主題相關的內容：
1. Why Big Tech hires are often poor fits for startups
1. 為什麼大型科技公司的招聘人員常常不適合初創公司
Let’s start with the elephant in the room; it’s a terrible idea to hire someone from a big company into a small, scrappy startup. Here’s the founder of a data startup on their personal experience:
讓我們先談談明擺著的問題；從大公司聘請人員到小型、靈活的創業公司是一個糟糕的主意。這是某個數據創業公司的創始人分享他們的個人經驗：
“Some of our hires from Google wanted to replicate all Google’s processes/culture, and completely failed. One staff engineer was the worst hire I can remember; they were so certain of their excellence and Google's superiority, that they ignored what made our company outstanding.”
“我們從 Google 招募的一些員工想要複製 Google 的所有流程和文化，但完全失敗了。有一位員工工程師是我記憶中最糟糕的聘用；他們對自己的卓越和 Google 的優越性充滿信心，以至於忽視了我們公司出色之處。”
An ex-Big Tech cofounder of an AI startup offers their experience:
一位前大型科技公司的共同創辦人分享他們在人工智慧創業公司的經驗：
“We've had Big Tech folks consistently fail our interviews on some rather fundamental engineering best-practice questions. We don't ask Leetcode questions and never will, but we found that BigTech candidates (Meta, Google, Stripe) had a particularly hard time with basic system design and coding questions.”
「我們發現大型科技公司的候選人在一些相當基本的工程最佳實踐問題上，面試表現持續不佳。我們不會問 Leetcode 題目，也永遠不會，但我們發現大型科技公司的候選人（如 Meta、Google、Stripe）在基本的系統設計和編碼問題上特別困難。」
There are other reasons, including:
還有其他原因，包括：
“Entitlement.” One thing mentioned by a few folks at startups is that some recruits from Big Tech are noticeably pleased about that fact, with a “I worked in Big Tech, therefore I’m awesome” mentality. Of course, it’s understandable to feel pride at having got into Big Tech and gained valuable experiences, as a career achievement. But when joining a startup from Big Tech, it seems sensible to be driven more by curiosity and humility, than judging a new workplace by the old one.
「權利感。」一些創業公司的人提到，來自大型科技公司的某些新員工對此感到相當滿意，抱持著一種 「我曾在大型科技公司工作，因此我很厲害」的心態。當然，能進入大型科技公司並獲得寶貴的經驗，作為職業成就，感到自豪是可以理解的。但從大型科技公司轉到創業公司時，似乎更應該以好奇心和謙遜為驅動，而不是用舊有的標準來評價新工作環境。
Startups do operate very differently from large companies, and the best way to make a difference and not alienate colleagues is to soak up a new environment, first!
新創公司確實與大型企業的運作方式截然不同，最好的方法是先融入新環境，這樣才能有所改變而不會讓同事感到疏離！
Success in Big Tech is often about managing optics, sometimes without real stakes. A founding engineer shares that there are plenty of seemingly successful engineering leaders in Big Tech who operate well, run great meetings, have excellent project management skills… and still ship lackluster products.
在大型科技公司中，成功往往與管理形象有關，有時甚至沒有真正的利害關係。一位創始工程師分享道，在大型科技公司中，有許多看似成功的工程領導者運作良好，會主持出色的會議，具備優秀的專案管理技能……但仍然推出平庸的產品。
Some characteristics can appear as ownership and agency, when they’re not. So, it’s easy to hire folks who are good at following processes, but not at being owners. Former Stripe product manager Shreyas Doshi describes this in the thread, “Operators optimizing for optics.”
某些特徵可能看起來像是擁有權和主導權，但實際上並非如此。因此，雇用那些擅長遵循流程但不擅長擔任負責人的人是很容易的。前 Stripe 產品經理 Shreyas Doshi 在這個討論串中描述了這一點，“為了外觀而優化的操作員。”
Lack broad experience with tools. A founding engineer at a fintech startup shares:
缺乏對工具的廣泛經驗。一位金融科技初創公司的創始工程師分享道：
“I came across folks with FAANG experience who did not even know JOINs on SQL! This was because they've only queried their internal non-relational datastore.
我遇到了一些有 FAANG 經驗的人，他們甚至不知道 SQL 中的 JOIN！這是因為他們只查詢過內部的非關聯數據存儲。
I had a friend who bragged about 10x-ing the QPS on a service at Google, but when I asked how they'd approach a Flask app running Postgres, they were completely clueless as to where to even start.
我有一個朋友自誇在 Google 的一個服務上將 QPS 提升了 10 倍，但當我問他們如果要處理一個運行 Postgres 的 Flask 應用時，他們完全不知道該從哪裡開始。
There's real skill in navigating FAANG stacks, but it's frequently using internal tools that someone else wrote for a promo packet, with little bearing on the "stitching together open source tools" of startup-land.
在駕馭 FAANG 技術棧方面確實需要真本事，但這通常是使用其他人為推廣包撰寫的內部工具，與創業界的「將開源工具拼湊在一起」關聯不大。
Many ex-FAANG people are unprepared for the upfront cost of learning the ecosystem outside of their silo. Non-technical startup founders or executives don't predict this; they just see the elite job background, and assume all candidates from that background will be strong in a role.
許多前 FAANG 的人對於學習自己所處範疇之外的生態系統的前期成本感到不知所措。非技術性創業公司的創始人或高管並未預見到這一點；他們只看到精英的工作背景，並假設來自該背景的所有候選人在角色上都會表現出色。
Focus on things startups don’t care about. An ex-Google engineer working at a startup says:
專注於初創公司不在乎的事情。一位在初創公司工作的前谷歌工程師說：
“Most FAANG engineers I've met do years of work without ever talking to a customer. In the ZIRP 2010s especially, they never had to worry about a cost, ever.
“我遇到的大多數 FAANG 工程師在多年工作中從未與客戶交談過。特別是在 2010 年代的零利率環境中，他們根本不需要擔心成本。”In a FAANG environment, there's a lot of focus on things that your early startup shouldn't care about – but which FAANG engineers do!
在 FAANG 環境中，有很多事情是早期創業公司不應該關心的，但 FAANG 的工程師卻非常重視！
These include: 這些包括：
A deep engineering ladder and promotion process
深厚的工程職階與晉升流程Expectations of consistent and/or relaxed working hours
對於一致或放鬆的工作時間的期望Make most decisions in meetings
在會議中做出大多數決策Architecture reviews 架構評審
Restarting work because someone found a technical snag that prevents a hypothetical scaling event
因為有人發現了一個技術問題，阻礙了假設的擴展事件，因此重新開始工作Technical things: 技術相關事項：
Ceremonies for "clean code" (whatever that means)
「乾淨程式碼」的儀式（無論那意味著什麼）Building for future scalability
為未來的可擴展性而建設Copying the tech stack of their previous Big Tech workplace.”
複製他們之前在大型科技公司工作的技術堆疊。
2. When hiring from large companies doesn’t make sense
2. 何時從大型公司招聘不合適
Big Tech talent can have a magnetic pull, but the quotes above indicate there’s plenty of ways that it can not work out in small workplaces. Circumstances when it doesn’t make business sense for a startup to hire for a Big Tech profile, include:
大型科技公司的人才具有吸引力，但上述引用顯示在小型工作場所中，這種情況可能會有很多不如意的地方。當初創公司聘請大型科技公司背景的人才不具商業意義的情況包括：
No real need for Big Tech expertise
不需要真正的科技巨頭專業知識
Many startups don’t actually need Big Tech expertise, especially not in leadership. An engineering manager at a startup in San Francisco explains:
許多初創公司其實並不需要大型科技公司的專業知識，尤其是在領導層方面。一位在舊金山的初創公司擔任工程經理的人解釋道：
“Leadership that has only operated at Big Tech often doesn’t know the realities of operating at a smaller scale. For example, planning years in advance at a startup is usually a waste of time because things change so quickly. But such planning is required in Big Tech!”
“在大型科技公司中運作的領導者，往往不瞭解在較小規模運作的現實。例如，在初創公司提前幾年進行規劃通常是浪費時間，因為變化來得太快。但在大型科技公司中，這種規劃卻是必需的！”
Unfamiliar with “startup infra” and pace. A downside of hiring from larger companies is that Big Tech engineers and managers are often used to shipping faster. In some Big Tech companies, they might have mostly been building on top of sophisticated, Big Tech-specific infrastructure, and be unfamiliar with common cloud infrastructures which many startups use, like AWS, GCP, GitHub Actions or similar tools. Outside of Amazon, Big Tech companies almost always use their own infrastructure, not public cloud providers. Google doesn’t use GCP.
對「創業基礎設施」和節奏不熟悉。從大型公司招聘的缺點之一是，大型科技公司的工程師和管理者通常習慣於更快的交付。在一些大型科技公司中，他們可能主要是在複雜的、大型科技特定的基礎設施上進行開發，對許多創業公司使用的常見雲基礎設施（如 AWS、GCP、GitHub Actions 或類似工具）不太熟悉。除了亞馬遜，大型科技公司幾乎總是使用自己的基礎設施，而不是公共雲服務提供商。谷歌不使用 GCP。
A startup founder in Boston says:
一位位於波士頓的創業創始人說：
“Some Big Tech companies are particularly bad at honing skills that translate to startups. For example, Google engineers usually focus on very small product surface areas, and all the work is on very specific Google infra stack.”
一些大型科技公司在培養可轉化為創業公司的技能方面特別不佳。例如，谷歌的工程師通常專注於非常小的產品表面範圍，所有的工作都集中在非常特定的谷歌基礎設施堆疊上。
No business model for paying Big Tech comp
沒有支付大型科技公司補償的商業模式
Big Tech companies typically generate around $400,000 to $1,000,000 in revenue per employee, while being extremely profitable. It is thanks to this kind of revenue generation that they can justify paying senior-and-above hires $500,000 a year or more in total compensation (in the US: adjusted to regions, but still paying top tier compensation.)
大型科技公司通常每位員工產生約 $400,000 至 $1,000,000 的收入，同時獲利極高。正是因為這種收入產生方式，他們才能夠合理地支付高級及以上職位的員工每年 $500,000 或更多的總薪酬（在美國：根據地區調整，但仍然支付頂尖薪酬。）
If a startup has a business model to eventually generate this kind of revenue, it means the business fundamentals exist to compete with Big Tech on comp. But if the business isn’t forecast to earn so much revenue, then paying the same kind of compensation as Big Tech isn’t sensible, nor practical.
如果一家初創公司有一個商業模式來最終產生這種收入，這意味著該企業具備與大型科技公司在薪酬上競爭的基本條件。但如果預測該企業的收入不會如此之高，那麼支付與大型科技公司相同的薪酬就不合理，也不實際。
Pure software startups often have a theoretical business model to get to Big Tech revenues. This is why it makes sense for such startups and scaleups raising venture funding to offer similar base salary and equity. These businesses then need to execute: grow their market and revenue.
純軟體創業公司通常擁有一個理論上的商業模式，以達到大型科技公司的收入。因此，這些創業公司和成長型公司在籌集風險資金時提供類似的基本薪資和股權是有道理的。這些企業接下來需要執行：擴大市場和收入。
Focus on speed, efficiency, and doing more with less
專注於速度、效率，以及以更少的資源達成更多
Most of Big Tech is used to doing more with lots of resources. For example, it’s impressive that Meta built the social media site Threads in 6 months, got 100 million users in the first week, all with a starting team of 12 people, but this was done by building on top of Instagram’s infrastructure. Things like the storage and compute layer did not need to be built from scratch.
大部分大型科技公司習慣於利用大量資源來做更多事情。例如，Meta 在 6 個月內建立了社交媒體網站 Threads，並在第一週內獲得了 1 億用戶，這一切都是由 12 人的初始團隊完成的，但這是基於 Instagram 的基礎設施進行的。存儲和計算層等部分並不需要從零開始構建。
Compare this with the Bluesky team building its social network from scratch: it took much longer, done with very little Big Tech experience. And it’s not a given that all Big Tech engineers can “do more with less” well, which is essential at early-stage startups. But sometimes it does make sense to hire from big places; Bluesky hired Dan Abramov from Meta. We cover more about How Meta built Threads, and How Bluesky was built in deep dives.
將這與 Bluesky 團隊從零開始建立社交網絡進行比較：這花了更長的時間，且幾乎沒有大型科技公司的經驗。而且並不是所有大型科技公司的工程師都能在早期創業公司中“以更少的資源做更多的事情”，這是至關重要的。但有時從大型公司招聘是有意義的；Bluesky 從 Meta 聘用了 Dan Abramov。我們在深入探討中涵蓋了更多有關 Meta 如何建立 Threads，以及 Bluesky 是如何建立的。
Among the Big Tech companies, Amazon is typically the closest to operating infrastructure like a startup, by running on AWS services. We cover more about why Amazon is a common source of startup hires, later.
在大型科技公司中，亞馬遜通常是最接近像初創公司一樣運營基礎設施的，因為它依賴於 AWS 服務。稍後我們將深入探討為什麼亞馬遜是初創公司招聘的常見來源。
To solve novel problems in scrappy ways
以靈活的方式解決新問題
If the goal is to get from zero to one in a difficult problem space by using as few resources as possible, Big Tech probably isn’t the place to do it. The biggest companies are good at solving novel problems with lots of resources, but are getting better at solving common, well-understood problems with fewer resources (headcount). Generally, Big Tech isn’t where a scrappy mentality for building novel solutions on a budget thrives.
如果目標是在困難的問題空間中以最少的資源從零到一，那麼大型科技公司可能不是最佳選擇。這些最大的公司擅長用大量資源解決新穎的問題，但在用較少的資源（人力）解決常見、已知的問題方面變得越來越擅長。一般來說，大型科技公司並不是在有限預算下建立新穎解決方案的拼搏心態所能蓬勃發展的地方。
A good example is AI companies. Google has an applied AI team that is easily 10x the size of OpenAI. And yet, OpenAI out-executes Google in novel product releases. Google, to its credit, is pretty good at catching up in problem areas that are well understood, such as shipping enterprise-ready APIs, or enabling its AI solution (Gemini) for enterprise Google Workspaces. We cover more on how OpenAI ships so fast in a deep dive.
一個好的例子是人工智慧公司。Google 擁有一支應用人工智慧團隊，其規模輕鬆是 OpenAI 的十倍。然而，OpenAI 在新產品發布方面的執行力超過了 Google。值得一提的是，Google 在解決一些已經充分理解的問題領域方面相當出色，例如推出 企業級 API，或 為其人工智慧解決方案（Gemini）啟用企業 Google 工作區。我們在深入探討中更詳細地介紹了 OpenAI 如何如此快速地推出產品。
An engineer at an early-stage startup puts it like this:
一位早期創業公司的工程師這樣說：
“In the absence of real stakes, many ex-FAANGers I've met view the focus on code and architecture quality as "doing the job of software engineering" and providing value.
在缺乏真正利益的情況下，我遇到的許多前 FAANG 員工將對代碼和架構質量的關注視為「在履行軟體工程的工作」並提供價值。In early-stage startups, the goal is to hit product-market-fit as fast as possible, it’s not to get high-quality code out the door. This difference means the day-to-day work is also different. Software engineers at startups should focus on what customers care about, and much less on what other software engineers care about.”
在早期階段的創業公司中，目標是儘快達成產品市場契合，而不是推出高品質的代碼。這一差異意味著日常工作也有所不同。創業公司的軟體工程師應該專注於 客戶所關心的，而不是其他 軟體工程師所關心的。
Related to this last point, here’s a deep dive on how to thrive as a founding engineer in a startup.
關於最後一點，這裡有一個深入探討 如何在初創公司中作為創始工程師蓬勃發展。
To operate completely differently from Big Tech
以完全不同於大型科技公司的方式運營
When building a company in a very different style from Big Tech, hiring from those places makes less sense. For example, when building a full-remote workplace, hiring from companies which mandate being in the office for most of the week, isn’t optimal. Of course, there are plenty of people in Big Tech who are tired of how things work there, and would like to try new ways of working. These people can bring valuable experience, without being tied to Big Tech processes.
在建立一個與大型科技公司風格截然不同的公司時，從這些地方招聘的意義就不大了。例如，在建立一個完全遠程的工作場所時，從那些要求大部分時間必須在辦公室工作的公司招聘並不是最佳選擇。當然，許多大型科技公司的人對那裡的運作方式感到厭倦，並希望嘗試新的工作方式。這些人可以帶來寶貴的經驗，而不受大型科技公司流程的束縛。
Why hire from Big Tech, anyway?
為什麼要從大型科技公司招聘呢？
If there’s no strong reason for hiring from Big Tech, why do so? Startups need a very good story to tell Big Tech folks in order to close them, even with compensation packages that match Big Tech. If that compelling story has yet to be written at a fledgling startup, then why bother paying the top of the market?
如果沒有強烈的理由從大型科技公司招聘，那麼為什麼要這樣做呢？初創公司需要一個非常好的故事來吸引大型科技公司的員工，以便能夠成功招募，即使薪酬方案與大型科技公司相當。如果在一家新創公司尚未寫出這個引人入勝的故事，那麼為什麼還要支付市場最高的薪資呢？
3. When it makes sense to hire from Big Tech
3. 何時從大型科技公司招聘是明智的
Despite the downsides mentioned above, there are naturally plenty of reasons to hire from large, high-performing companies! These include:
儘管上述提到了一些缺點，但從大型高效能公司招聘自然有很多理由！這些理由包括：