Conversation

Interesting as a strong leadership belief in a good leader. He’s right it’s a sign of dysfunction in a good, centralized hierarchical org. For decentralized networks, basically every result/algorithm I’ve read about tldrs to “communicate more, between more pairs of points”
Quote Tweet
Jeff Bezos’s counterintuitive idea: “Communication is a sign of dysfunction.”
Show this thread
Image
6
134
I’ve worked with good/bleeding edge centralized hierarchical orgs and the entire philosophy of good management can be boiled down to “do what Amazon does.” It’s amazing how far ahead they are of state of the art MBA theories of good management, at least at that scale.
1
54
I’ve only recently started working with good *decentralized* orgs (first paid gig was this year) and my prelimary conclusion: flip almost everything you think is correct about centralized hierarchy management by default. And best if you flip the *best* playbook: Amazon’s.
3
51
A good decentralized org would basically be an anti-Amazon in a good way. Flipping *bad* centralized hierarchy models is tempting but dumb. Tempting because most people turn to decentralized network structures out of frustration with bad centralized hierarchies.
2
33
For eg: Good centralized orgs do centralized bureaucracy *well* and if you flip that you’ll think creatively about good decentralized bureaucracy. But if you knee jerk react to bad centralized bureaucracy, you’ll just create a big of anarchist un-process that doesn’t work
2
31
Heard good things about the working backwards book, but haven’t yet read. And I guess now that it’s over, I can share that I did a 5-year gig with Amazon that just wrapped. One of the best gigs of my free-agent career. I’ve liberally applied managemebt ideas from there elsewhere.
2
23
I’d say about 1/3 of everything I’ve learned about good management at bleeding edge scale in the last decade comes from Amazon. Another third comes from working with Jim Keller over the same period across 4 companies. Last third is everything else.
1
19
Replying to
Will wrap and publish my art of gig book for consultants in the next week or two, but I feel I’m not yet ready to write the management book I eventually want to. Ambitions kinda along the lines of The Essential Drucker. But need to truly grok the decentralized side to same level
1
18
2x2: small vs big, and centralized hierarchy vs decentralized network. I’m mostly on the “big” side. I’ve logged hours at the small scale but tbf there’s not a whole lot to say on small-scale *management* per se. It’s 90% about product, funding, and chemistry of 3-4 key people.
3
14
Minimum scale at which my kind of management theorizing seems to be meaningful and useful is like 25 people, which is also my gating criterion. I almost never work with anything smaller unless there’s a product angle I understand particularly better than others (which is rare)
2
7
I’m dangerously close to being a SPOF at YC, as are 3-4 others, but we are getting better steadily at being robust to “critical node dropout” every year
3
10