[crdb]

I wish it was the case (wouldn't that be convenient for my reputation) but it was a more standard culture fit problem, they didn't think we'd work well together. Specifically, the job involved a lot of cross-office coordination and I'm more of a "get the job done", technically focused kind of manager than the more political nature of the managerial role in question (political in the sense that you need to help multiple teams work together harmoniously and efficiently, not House of Cards).

They were quite open with feedback which is another reason I have good feelings towards the company. The whole process was very quick - responses in days. Compared to what my friends tell me about Google's sometimes multi-year process, it's a dream.

I don't know how much I can say about their internal structure but what I found very cool (especially as a former Rocket Internet employee) was their focus on and methods for detecting and spreading knowledge across the entire organisation, which is sort of decentralised even if many important decisions are made in California.

This is something we never got right at Rocket, where companies and regions are much more independent and have wildly differing methods and standards. It was very rare for things to cross over beyond the "startup code" (the PHP e-commerce codebase that is sent over to be deployed when a venture is opened) and some marketing methodology. The other large companies I've been familiar with are better at sharing information, but tend to have a global hub where most of the thinking is done anyway (i.e. knowledge tends to radiate outwards).

[hissworks]

Thanks for the insights. Sounds like a fair procedure.

Curious to hear more about what you found cool about their communication methods. I've been thinking about this a lot as I'm surprised how inefficient a lot of Big Corp communication methods can be, especially those with a field/regional organization. Inclined to agree that knowledge radiates outwards, but tinkering with ideas about how to best receive feedback inwards and what are the most effective methods of radiating that knowledge. Easier said than done.

[crdb]

I can't speak specifically about Uber (not sure what is fair to disclose or not) and definitely not about Rocket. However, I can speculate on what I learnt from both. It's not (or not "just") about "communication methods" but an organisational problem.

In the hypothetical case where I'd be running a company like that, I'd want a small team of people who are both technically and operationally extremely able, and personally likeable and politically capable; preferably people who have proven themselves and have credit internally (or such achievements as to reach the same reputation even as an external hire).

I'd task them to overview everything that is happening globally, coordinate trials and experiments so that several hundred could run at the same time globally, and share back information and results across the world. They'd be in effect the repository of knowledge of the company, the institutional memory (so much wheel gets reinvented!) and they most crucially wouldn't be based in HQ in the Bay Area, but a distributed global team constantly roaming and meeting all the local teams. They wouldn't have or need authority to "get things done" by steamrolling over the local office (as the IBM Black Team did [1]), instead convincing them that it's a good idea to try.

They would include specialists in every discipline that is important to the company from finance to statistics, to make sure that there is always at least one team member that can be on the same level as a local specialist (both to avoid taking stupid decisions, and to be credible in conversations).

[1] http://www.t3.org/tangledwebs/07/tw0706.html