The issue is this view doesn’t scale beyond a small organization. There are a variety of reasons many of which I already pointed out but including you can’t hire uniformly P99 leaders.
The thing Bezos did with Amazon was create a scale free organizational culture which is resilient and highly adaptable. You can’t do that by adopting processes and organizing artifacts that depend on perfect execution by everyone everywhere all the time against some ideal. You have to build processes that are resilient to inadequacy and even incompetence yet still be successful at all levels reliably. When you’re managing an organization of over a million employees with a pretty flat org structure this becomes even more important. Saying “delegate effectively” is not a resilient thing - setting up a structure that ensures delegation happens but executive leadership is aware of and involved in enterprise critical decisions is hard to do.
One way Andy Jassy does this is he requires the documents to be read in his meetings to always use a specific style including Oxford commas. If he reads the document and there aren’t Oxford commas he ends the meeting and you have to reschedule - which can take months. So, you really are certain you have made the most succinct document according to a protocol that’s very low cognitive load for him. He delegates most decisions to his team and they to theirs but at the scale of aws or Amazon, there are some decisions he is a part of. And that number is a lot because Amazon is enormous, not because he doesn’t trust his team or delegates.
But some things he doesn’t. At aws he never delegated pricing decisions. He scrutinized any pricing change in detail. If you did a good job and everyone on his team already was bought in he invariably had something incredibly insightful no one else thought of. He would send it down and his subordinates would often be empowered to approve it. But he always reviewed pricing at least once. This was less about micromanagement and more about choosing to apply his time against what he felt his org should really care about. Margin, cost, scale, and customer experience of these things.
No offense, but nature of these replies sounds more like someone parroting tech speak than someone applying principles that have a context beyond just tech or AWS.
>scale free organizational culture
The fact that the meetings are bottlenecked by decision makers ability to synthesize information imply it is not, in fact, “scale free”. Scale free would imply there are not such bottlenecks.
I agree it’s about developing resilient processes. What you allude to is not that, because it implies single points of failure within decision-making, bottlenecks etc. It doesn’t come across as a clear understanding of true process-oriented culture.
>he reads the document and there aren’t Oxford commas he ends the meeting and you have to reschedule
I can’t know, of course, but I suspect this has more to do with ensuring due-diligence than document formatting. It’s the same thing Van Halen did in the 1980s by requiring brown M&Ms removed from the bowl in their dressing room. It was a quick heuristic to ensure the venue read their rider/contract completely and adhered to it because a lack of due diligence in set design, pyrotechnics etc. would have been a major safety issue. I’m willing to bet checking for a lack of Oxford commas is shorthand for “what other details did they miss?”
>And that number is a lot because Amazon is enormous
Again, this implies the opposite of “scale free” culture. A true scale free org wouldn’t have any nodes with a large number of connections. A large number of decisions does not mean any single individual has to involved in those decisions directly. Why is he not willing/able to select someone capable of making those decisions? (Honest question to understand the dynamic).
The thing Bezos did with Amazon was create a scale free organizational culture which is resilient and highly adaptable. You can’t do that by adopting processes and organizing artifacts that depend on perfect execution by everyone everywhere all the time against some ideal. You have to build processes that are resilient to inadequacy and even incompetence yet still be successful at all levels reliably. When you’re managing an organization of over a million employees with a pretty flat org structure this becomes even more important. Saying “delegate effectively” is not a resilient thing - setting up a structure that ensures delegation happens but executive leadership is aware of and involved in enterprise critical decisions is hard to do.
One way Andy Jassy does this is he requires the documents to be read in his meetings to always use a specific style including Oxford commas. If he reads the document and there aren’t Oxford commas he ends the meeting and you have to reschedule - which can take months. So, you really are certain you have made the most succinct document according to a protocol that’s very low cognitive load for him. He delegates most decisions to his team and they to theirs but at the scale of aws or Amazon, there are some decisions he is a part of. And that number is a lot because Amazon is enormous, not because he doesn’t trust his team or delegates.
But some things he doesn’t. At aws he never delegated pricing decisions. He scrutinized any pricing change in detail. If you did a good job and everyone on his team already was bought in he invariably had something incredibly insightful no one else thought of. He would send it down and his subordinates would often be empowered to approve it. But he always reviewed pricing at least once. This was less about micromanagement and more about choosing to apply his time against what he felt his org should really care about. Margin, cost, scale, and customer experience of these things.