Hacker News Reader

Comment by greendave | Hacker News Reader

> Before the layoffs, Cruise had 1,800 full-time employees.

Their hiring rate has been remarkable. Less than 3 years ago (6/2017) they had 200 employees. Just over a year ago (3/2019), they had ~1000.

View on HN

Wow that is ridiculous. There is absolutely no way to have those numbers without a major compromise in the quality of hires.

There are 7,500,000,000 people in the world. A lot more than 800 of them are good at their jobs.

800 hires a year means ~15 people starting every week. Every one of those 800 positions likely received hundreds of resumes, which needed to be sifted through. Then you need to have calls with each candidate, narrow down the pool further. Then all the onsite rounds. For an initial team of 200 that is an impossible task.

The "initial" 200 people don't hire the next 800, they hire the next 1.

Then the 201 people hire the next 1, and the 202 hire the next one, and so forth.

I don't think many people on this board comprehend that there exist professionals at most companies whose sole responsibility is to identify, recruit, and onboard new employees.

Or that most companies are willing to hire people who aren't perfect candidates because a fair amount of internal training is expected. Hiring only unicorn candidates who already have the perfect qualifications and don't need any training/onboarding is a peculiarly Silicon Valley phenomenon.

To hire so many you have to hire a lot in parallel. Also, even once hired, probably the engineer shouldn't have "interview more engineers" as his first priority.

Note that the average or median “quality” of employees isn’t really that relevant if you just need more employees to do more total work.

Eh, mythical man-month and all that. As always, the answer is somewhere in the middle, but I'd be surprised if most companies didn't see effects from median employee quality dropping.

That myth concerns adding more engineers to a project that is already started. Of course that doesn’t mean that 100 engineers can’t do more projects than 10 engineers.

Of course, but AV parallelization is limited by the fact that you're all working on a single, tightly-interconnected system. The reality is somewhere in between "do more (fungible, separable) total work" and "dump engineers into a single project". Onboarding costs are substantial for a fast-growing AV company.

Do people have experience of that ever being a success? Because I would ball park it and say my team can still basically be productive at 33% growth per year - for every 1 new employee you need about 3 established employees. I've got to imagine that the vast majority of those new employees don't know what they're' meant to be doing, and let's face it, they've probably got 16 different people doing the exact same infrastructure job that 16 different teams thought was essential, hired for, and didn't trust anyone else to deliver.

Reply on news.ycombinator.com