How to move from big-tech to a startup (and vice versa)

A staff engineer I know left Amazon for a 30-person Series A in Berlin. Six weeks in, he Slacked me: "I don't know how to build anything without seven internal services I now don't have." He figured it out. But the first two months were rough in a way the signing bonus hadn't warned him about.

The big-tech to startup transition (and the reverse) is one of the most common moves in a tech career. It's also one of the most consistently underestimated. People think they're changing employers. They're actually changing jobs.

What big-tech actually gives you

Before you decide to leave, be honest about what you have.

At a big tech company, you get:

• Infrastructure that just works. Deploy systems, observability, security review, on-call rotations — someone else built all of it.
• Peers who are consistently senior. Bad engineers exist, but the floor is high.
• Process that slows you down but also catches your mistakes.
• A brand on your resume that opens doors for the next decade.
• Compensation that, stacked with RSUs, is hard to beat outside FAANG.

None of that is a trap. It's actual value. The question is whether you need it.

What startups actually give you

At a 20-person startup:

• You own whole systems, often before you're ready to.
• Your decisions ship the same week, not the same quarter.
• You learn by doing things badly for the first time, in production.
• Equity that's probably worth nothing but might not be.
• A team where you know every person's name, strengths, and dog.

The speed is real. The dysfunction is also real. Small companies have fewer problems than big ones, but the problems they do have hit you directly because there's no one between you and them.

Going from big tech to startup

The failure mode: you bring big-tech expectations into a place that can't meet them.

You expect:

• A staging environment. There isn't one.
• A design system. There isn't one.
• A PM who writes specs. There isn't one — or the PM writes one paragraph in Notion and expects you to figure it out.
• Quarterly planning. The plan changed last Tuesday.

What works: treat the first 60 days as a new job, not a promotion. Even if your title is "senior" or "staff," you're a new hire at a company whose constraints you don't understand yet. Don't show up on day 3 proposing to rewrite the deploy pipeline. Ship something small. Learn why things are the way they are. Propose changes after you've earned credibility by delivering.

The staff engineer I mentioned — he got through it by deleting his own assumptions. He stopped asking "where's the X service?" and started asking "what's the 20% of that functionality we actually need?"

Going from startup to big tech

The reverse move is harder than people expect because everyone assumes startup people are scrappy and just need to adjust.

The failure mode: you get crushed by process and think the company is broken.

You'll hit:

• Launch reviews. Privacy reviews. Security reviews. Legal reviews.
• Three levels of approval for changes you'd have pushed to prod in an hour.
• A promo system that rewards documented impact, not just results.
• Meetings about meetings.

What works: accept that the process exists for reasons you don't fully see yet. A privacy review at a 50,000-person company is not bureaucracy for its own sake. It's because the blast radius of a mistake is billions of users.

The skill you need to develop: writing down your impact in a way that the promo system recognizes. At a startup, your impact is visible because everyone sees you. At big tech, if you don't document it, it didn't happen.

Which direction is right

Some rough heuristics:

Go to a startup if: you're mid-career and bored, you want to be forced to grow in ways a big company won't push you, you can absorb 12-18 months of stress, and you don't need the compensation.

Stay in big tech if: you have real financial obligations, you're optimizing for calm, you actually like the scale problems, or you're within 2 years of a level-up that will change your earnings for a decade.

Go to big tech if: you've been at startups for 5+ years and never seen what "good engineering process" looks like, you want the resume credential, or you're tired of things being broken all the time.

A note on timing

The worst time to make this move is when you're running from something. If you hate your current job, fix that first or at least understand it. You'll carry the underlying dissatisfaction across, and two months in you'll wonder why the new place also feels wrong.

The best time is when you're doing fine where you are and still feel pulled somewhere else. That pull is useful information. Running-away energy is not.

The takeaway

Big tech and startups aren't two flavors of the same job. They're different jobs that share a skillset. Moving between them is a real transition, and the people who do it well treat it as one — they stay curious, delete their assumptions, and give themselves six months before deciding if it was the right call.

If you're about to jump either direction, don't pick based on the job description. Pick based on what you actually want to spend your next two years doing.