Some of the things the author observes, like having 56 different frameworks that roughly do the same thing, or so many business processes running in Excel, are really not about dysfunction, and are actually "by design", but in a nuanced way.

Amazon has a strong thesis about how organizations should function and how they should execute work / produce useful outputs, and part of that thesis is, to be incredibly and nigh-unbelievably averse to coordination costs. You can also think of this as, a consequence of the frugality mindset leading to an aversion against premature optimization. Yes, it is wasteful to some degree to have 56 different frameworks, but OTOH, the ROI of spending the effort to consolidate them, simply isn't there to justify fixing it... if it takes a team of 2-3 engineers 6 months to consolidate it, well guess what, in the same time, that same team of engineers could launch a $10M-$100M business opportunity (seriously). And trying to put a single central team in charge of maintaining a universal framework that can support all current and future needs forever, simply never works out (eliminating gatekeepers is a common refrain, in service of moving faster). Compared against the ROI of actually delivering value to customers, who cares about some extra frameworks?

Or put it another way: Amazon does not optimize for waste elimination, or "cleanliness" of organization, or unity of purpose; Amazon optimizes for deploying capital (in the form of launching software/services/products) as quickly as possible across as many opportunities as possible, and then rapidly evaluating and doubling down on winners. Is there waste? Absolutely (I mean, what VC has a 100% hit rate?)! Does it seem dysfunctional? Probably. Is this way of working effective (meaning, does it actually function at achieving the goals of the organization)? I think the growth of Amazon in the past 20+ years speaks for itself.

Does this way of working put people/employees first, or optimize for their happiness/well-being? No, not particularly, but neither do many industries (investment banks, public school teachers, or as you point out, the restaurant industry.). Should that be fixed or at least improved? Well, I'd like to think that's why I'm here, and also why post-retirement Bezos has pivoted to working on making Amazon "Earth's best employer" [1] (which is a significant shift from the original focus of being, "Earth's most customer centric company"). I realize those sound like platitudes, but it actually is a meaningful statement, showing a reallocation of energy/focus between customers vs employees.

[1] https://www.cbsnews.com/news/jeff-bezos-amazon-employee-care...

---

EDIT: One further thing I wanted to clarify: if I am arguing in favor of anything, it is that Amazon as a whole organism, is far from dysfunctional at the macro level. But if you zoom in to individual pockets, individual orgs or teams, then yes absolutely there are many, many bad leaders, bad managers, poorly planned roadmaps, over-burdened/over-allocated engineers, terrible on-call cycles, etc., like what the author of this post observed. You are always taking a risk when joining Amazon, that you might end up in one of these bad situations. How high is that risk? Without trying to be overly precise, I would venture that it's closer to 80% good and 20% bad, then vice versa (but you don't have to believe me).

Another curious thing I've learned to appreciate in my team here, is that there is practically no other company in the world, where you would get the opportunity to learn how to operate a culture like this, at this scale, with this level of effectiveness (imperfect though it may be). How valuable or translatable is that experience though, I'm not sure... it's not like I can feasibly take my knowledge, and start up an Amazon-killer, or really get the opportunity to exercise my knowledge of how to design a company culture like-this-but-better, at this type of scale. But it sure is fascinating to watch it tick, from the inside.