Left Google 11 months ago after 4 years as an L5 SWE, mostly in the Ads org. People still ask me if I miss it. Complicated answer.
The WLB depends almost entirely on your team and your manager, which is the least satisfying thing to hear but also genuinely true. My first two years I worked maybe 45 hours a week max. My last two I was consistently pushing 55-60 during product cycles. Same company, different team. The culture that's advertised in recruiting material reflects some teams and is completely unrecognizable in others.
What's real: 20% time exists on paper. Whether you can actually use it is manager-dependent. I never used it. OKRs create a grind in Q3 and Q4 that most people don't mention upfront. Everyone is sprinting toward impact season. The internal tooling is genuinely excellent. That part of the lore is true. The peer review cycle (Perf) shapes behavior in ways that aren't always healthy. You learn to document your impact obsessively, which is a useful skill but exhausting as a constant background task. Geographic variation matters. NYC office culture felt noticeably different from MTV. Smaller offices tend to have more autonomy, less visibility.
What's overrated: The free food is great but it's also a mechanism to keep you at your desk. The campus vibe is real but it can become a bubble that's hard to escape. I knew people who hadn't cooked a meal in two years and thought that was fine.
What's underrated: The opportunity to work with genuinely exceptional engineers is real. The internal tech talks, the breadth of systems you can learn from, the ability to switch teams without resigning. These are legitimately differentiating.
My overall read: Google is a great place to be L4 or L5 and building skills, harder to sustain long-term if you care about scope of impact vs. internal politics. The culture has shifted noticeably since the 2023 layoffs. The psychological safety that used to feel baked in is more fragile now.
If you're optimizing for total comp at 2026 market rates, it's still very competitive. If you're optimizing for autonomy or mission clarity, you'll likely spend a lot of time frustrated.