I spent twelve years at companies the recruiters like to list on the front of their emails. I will not pretend the brand was meaningless. It opened doors, it got my mortgage application through the first stage of a high-street bank's underwriting in about forty minutes, and it mostly stops the aunties at weddings from asking when I plan to do something serious with my life. But the things it actually taught me about engineering are not the things you will see in the cheerful blog posts written by people still employed there.
Here are five.
1. Scope is almost always more important than technical brilliance
The strongest engineer I worked with at Google was a woman called R, who could rewrite a race-conditioned indexing pipeline in her head on a flight home from Dublin. She was also the most senior stuck-at-the-same-level engineer on the team. She stayed there for four years. The reason was scope. She worked on unglamorous plumbing with no customer story, no organisational pull and no cross-functional sponsor. She was producing work that, in a retrospective, would be described as "essential" and, in a promotion packet, described as "maintenance".
The uncomfortable truth is this: at a large company, your impact is measured by the shape of the hole you would leave behind if you disappeared. Technical brilliance is necessary. It is not sufficient. The engineers who got promoted picked problems with a hole-shaped future, then used their brilliance to fill them.
2. "Staff" is a political role wearing an engineering badge
Every Staff+ engineer I worked with spent somewhere between 30% and 60% of their working week on something that was not, strictly speaking, coding. They were aligning, convincing, translating, socialising, de-risking. They were writing documents that served as a pre-agreed version of reality so that meetings could be short.
When I was a Senior engineer I watched this from below and thought it looked uncomfortably close to middle management. Now I've done it I understand that the best Staff engineers are management's technical conscience. They are the people who say, in a room with VPs in it, "this project will fail in the following specific way". They can only say that because they have spent years accumulating the political capital to be taken seriously. If you want to be a Staff engineer, you will have to learn the political game. You do not have to like it.
3. The compensation curve is not what you think it is
The compensation arithmetic inside big tech is built around retention, not reward. Your offer letter is a function of what it would cost the company to replace you, not what your work is worth. This is why internal promotions almost always carry a smaller total-comp uplift than jumping to a competitor at the same level.
The genuine, discontinuous comp step changes happen at three moments: your initial hire, a level-up into Senior, and a level-up into Staff. Everything else is noise around the edges. If you are optimising your earnings and not your learning, you should be ruthless about these three moments and lazy about the rest.
4. The "brand" protects you for about three years after you leave
I've now been out of full-time big-tech employment for a little over a year. When I was interviewing for advisory work, the Meta line on the CV was doing a lot of heavy lifting. A year from now it will be doing less. In five years it will be a piece of decoration.
What is actually portable is the specific muscle you built: the fact that you have held a pager for a global service, the fact that you have shipped a migration that touched a thousand callers, the fact that you have sat through a live incident on Bonfire Night with a VP on the phone and still made good decisions. The brand gets you through the first interview. The muscle keeps you employed afterwards.
5. You will miss it more than you expect to
I did not expect this one. I'd spent the last two years at Amazon quietly furious about a reorganisation, a product decision and the internal tool I had to use to book travel. I left. I was relieved for approximately three weeks. Then I missed it.
Not the company. The company, as an employer, is largely indifferent to the specific you. What I missed was the resolution of the problems. Big-company engineering work is a steady stream of puzzles of a very particular shape: too much data, too many constraints, too much legacy, real consequences. Once you stop having those puzzles dropped on your desk every morning you realise how much of your identity was built around solving them.
I'm not going back. But I'm fonder of the place than I was when I was in it, which is, I gather, how a lot of things in life go.
If you're three years in and you're wondering whether the next three are worth it, the honest answer is: probably, and not for the reasons on the careers page.
— Nivaan