Conversation

1) Zig Zag and hiring: why FTX is going to keep growing as others cut jobs

200

650

3,079

2) A few months ago, when BTC was $44k, I wrote a document which lead with: "Today, February 8th, 2022, I’m not announcing a hiring freeze. Which, in many ways, makes it like every other day. But I think it’s time for us to substantially slow down hiring for a bit."

7

15

319

3) But before we get there, let's start in 2018. One of the first lessons I learned as a CEO: Make sure your team can mentor employees as fast as you hire them. Otherwise, as I learned the hard way, everything gets worse.

11

74

595

4) I've looked into hypergrowth companies. Time and time again, growing from 200 employees to 2,000 doesn’t seem to 10x your productivity as a company. In fact, sometimes it doesn’t even 1x your productivity. Sometimes, the more you hire, the less you get done.

10

105

722

5) Why exactly is it that hiring doesn’t scale the way it should? a) Coordination becomes really hard b) Diffusion of responsibility: if 5 people could theoretically do something, maybe no one will feel like they actually have to do it, and so it won’t get done

8

29

549

6) c) Companies lower the bar as they hire more, and over time the average coworker people interact with becomes a worse experience d) Incentives become harder to align as people have less and less of a sense of what each other are doing

7

18

388

7) This is, as far as I can tell, the most common reason that successful companies decay. It’s a race–between how quickly new employees understand the company's culture, and how quickly people are hired. The faster you hire, the harder it is to keep everyone on the same page.

5

15

374

8) So maybe you can grow by 50% each year and still mentor each new recruit. But if you grow by 300%, each new employee only has 4 months to learn before they have to teach. And meanwhile we have to make sure teams are coordinating and compensation is fair.

7

16

295

9) Otherwise--if you're not careful--your company slowly decays into Moloch before your own eyes. (slatestarcodex.com/2014/07/30/med) ---- They broke their backs lifting Moloch to Heaven! Pavements, trees, radios, tons! lifting the city to Heaven which exists and is everywhere about us!

slatestarcodex.com

Meditations On Moloch

I. Allen Ginsberg’s famous poem on Moloch:What sphinx of cement and aluminum bashed open their skulls and ate up their brains and imagination? Moloch! Solitude! Filth! Ugliness! Ashcans and u…

7

16

239

10) And so, when we hit around 250 people, we slowed down hiring. Not because of money--we were, and still are, strongly profitable--but because we wanted to make sure we could actually integrate everyone we hired into FTX's working culture.

10:59 PM · Jun 6, 2022Twitter Web App

Replying to

11) We have 10% as many employees as others, but when it comes to output, I would take on any of them. We built FTX, originally, with two developers. Having 25 today allows us to scale. Never underestimate what each developer can get done in the right environment.

8

40

459

12) And because we hired carefully, we can keep growing regardless of market conditions. Because we exponentially scaled our revenue and productivity, not our expenses. But more importantly, because each person we add takes on a huge opportunity, and a huge responsibility.

13

15

342

13) A lot of reporters have reached out asking if we're going to be scaling back. And our answer is the same as always: We're going to keep pushing forward.

22

14

263

14) More will be joining than leaving, though--as always--we'll make sure to add people at a sustainable rate. And, as always, we'll make sure to uphold a high standard.

22

7

227

15) The amount each teammate can do is enormous. And while we expect a lot of each other, it's incumbent on us to create an environment where people can thrive.

27

9

237

16) The first truth of the 10x engineer is that they're just a smokescreen for their 100x counterparts. The second truth is that they aren't usually born that way. Some of our devs who were rejected by another employer now produce more than that company's whole team.

37

25

458

17) Which is a lesson that I learned the hard way, when I grew a company from 5 people to 10 to 25 to 10, all in the matter of a few months.

28

7

218

18) Anyway, on February 8th, I didn't announce a hiring freeze. And today I'm not lifting a hiring freeze--because there never was one. And because the truth is that, for FTX, things are the same as they always are. We should hire great people, and be great to them.

35

25

437

19) Sometimes, when others Zig, you Zag.

98

63

647