Most people hear “Google job” and picture a programmer.

That’s maybe 40% of the company.

Google employs roughly 180,000 people. The other 60% do work that sounds ordinary until you look closer. Salespeople who negotiate $50M cloud deals. Writers who agonize over the 4 words on a cancel button. A guy whose entire job is managing the company’s fleet of MacBooks.

I’ve spent 3 years tracking their job board and talking to people inside. Here’s what’s actually there. No recruiter babble. Just the roles and what they really look like.

Engineering and technical roles

You know about Software Engineer. That’s the default. But the floor is broad.

You have Site Reliability Engineers (SREs). They sit between dev and ops. Their job is to keep Search and Gmail from going dark at 3 AM. Pay is high, oncall stress is real.

Cloud Architects don’t write production code. They help customers unbreak their cloud setups. You need to know Kubernetes and networking cold, but you’re also on video calls with clients half the day.

And Research Scientists? They live in DeepMind. They publish papers. Most of them have PhDs and haven’t shipped a consumer product in years. That’s not the point. They make the breakthroughs the products ride on 5 years later.

Data Scientists and ML Engineers sit closer to the product. They build the models that rank ads and recommend videos. The split: Data Scientists analyze. ML Engineers productionize.

What they actually hire:

• Software Engineer (backend, frontend, mobile, generalist)
• Site Reliability Engineer (SRE)
• Cloud Architect / Customer Engineer
• Data Scientist / Machine Learning Engineer
• Hardware Engineer (Pixel, Fitbit, TPU chips)

Comp starts around $150k base and goes north of $300k at staff+ levels. But you’ll grind LeetCode for 3 months to pass the loop. 4 to 6 interviews. System design. Behavioral. No shortcuts.

Product and design

Product Managers at Google own the roadmap. They don’t tell engineers what to do. They figure out what’s worth building in the first place. It’s a high-leverage, high-visibility job. The application process is a lottery. Thousands apply. Maybe 0.2% get in.

Technical Program Managers (TPMs) are the unsung heroes. They handle the timeline, the dependencies, the “legal review is stuck again” blockers. If you’re almost pathologically organized and can read architecture docs without glazing over, TPM is the path.

UX Researchers get a surprising amount of respect. I’ve talked to one who spent 8 months traveling to rural parts of India to study how farmers use voice search on cheap Android phones. That’s the gig. Talk to real people. Find the friction. Bring it back to the designers.

UX Writers craft the microcopy. Error messages. Button labels. Onboarding flows. It’s writing with constraints. Every word earns its pixels or it gets cut.

Sales and cloud: where the money is

You see a Google ad? A human sold that ad slot.

Google’s ad sales team is enormous. Less visible than engineering but it’s the revenue engine. You start as an Account Executive or Account Strategist. Mid-market or enterprise. You manage a book of clients, hit quarterly quotas, and explain to a car dealer why his cost-per-click went up.

Cloud Sales is different. More technical. The job title is often Customer Engineer or Technical Account Manager. You need to demo BigQuery. You need to walk through a migration plan. The interview has a coding component, then a mock customer pitch. If you can’t explain IAM roles clearly to a confused CTO, you won’t make it.

Comp is heavy on commission. A mid-level Cloud rep can pull $180k to $220k total. But you feel the pressure every quarter. Miss targets twice and you’re looking.

The jobs nobody thinks about

Google hires technical writers. They document APIs, write SDK guides, maintain the developer docs you’ve skimmed at 1 AM. Dry work. Essential work. The good ones prevent hours of confusion across the entire developer ecosystem.

They hire chefs, facilities managers, real estate negotiators, and security guards. They employ a team of people whose full-time job is managing the global laptop fleet. Tens of thousands of Chromebooks and MacBooks. Refresh cycles. MDM policies. Asset tracking. One person I talked to described it as “IT logistics at civilization scale.”

Recruiting operations is another big group. Coordinators, sourcers, program managers who keep the interview machine running. Google gets about 3 million applications a year. Someone has to process all that.

And there’s always the weird edge: a colleague once interviewed for a role writing trivia questions for Google Assistant. The job was real. Paid decently. Required a weirdly deep knowledge of 90s pop culture.

How to actually get in

Get a referral. It won’t land you the job, but it gets your resume seen. Reach out on LinkedIn, do actual work the person can see, then ask. Cold “please refer me” messages go straight to trash.

Write your resume like a human. No “results-oriented professional with a proven track record.” Just actions and outcomes. “Shipped a redesign that increased checkout completion by 18%.” Specific beats vague every time.

And apply for the less famous roles. The Associate Product Manager job has 500 applicants by Tuesday. The Cloud Learning Services Program Manager role? Maybe 15. The smaller the pool, the better your odds.

One last thing. Google’s hiring process is slow. Weeks between stages. Sometimes radio silence. Don’t wait by the phone. Apply and move on with your life. If they call back, great. If not, you’ve got something else running.

Common questions

Do I need a CS degree?

Not really. It helps for SWE roles. But I know a Cloud Customer Engineer who majored in philosophy. Google cares about output, not transcripts. Show you can do the work.

What’s the hardest job to land?

Staff Engineer in a research group like DeepMind. Or a Product Manager role. The PM funnel is enormous and the interview process is notoriously ambiguous. People prep for years and still strike out.

Does Google hire non-technical people?

Yes. Aggressively. Sales, legal, HR, culinary, real estate, operations, writing. If a mid-sized city needs a function, Google employs someone to do it. Probably 20 someones.