Hacker News Reader

Comment by anarbadalov | Hacker News Reader

Jesus, i would not want this guy's job, nor would i want to work for him.

"His colleagues describe a demanding manager who is impossible to bluff. During the holidays, Clark’s team meets daily to review metrics about Amazon facilities around the world. Clark can zero in on any signs of trouble since he’s been scanning the same figures for 20 years. A key metric is the estimated delivery time shoppers see when they view a product on Amazon, which is broken down by geography. If the number is moving in the wrong direction, Clark demands an explanation and a solution. The culture is far from pleasant, but current and former insiders marvel at its effectiveness."

View on HN

Having been at a very low level of this I can confirm the daily reviews. And yes it is extremely effective, if you have the right people.

The delivery time thing is, IMHO, ober simplified. Not sure if Clark is reviewing that life, but it is reviewed for the previous day. At least on the warehousing side of things, lead times are closely watched as these directly impact promised delivery dates for everything stored there. Same goes for transportation lead times. The worst thong to happen is the "red button", if that is pushed a complete warehouse looses the Prime status, meaning only standard delivery is enabled. Happens quite regularly so, and I even remember a Big Red Button when the whole network was affected. It just shouldn't be to long.

The most important metric at my level was missed shipments. More explicitly shipments missed due to the warehouses fault (ready to ship on time but not picked up by the carrier). Even worse, shipments that missed the promised delivery date. It speaks volumes about Amazon's logistics systems that all this could be figured out before 9 o'clock in the morning for the previous day. And it is one reason why you learn so much there.

Regarding people, Clarks ability to manage the network not only depends on the low level guys (like me back the day) but also managers in between. If these managers got their priorities right the system is less harsh then one might think. If managers play politics, not so much.

What’s better? Waiting for a yearly review to talk about why something 6 months ago was low? It is better to adjust these problems as they came up.

That way everyone’s on the same page. There’s no doubt that X is an important metric and that your manager is informed of it everyday so that it isn’t a surprise when it comes to a review later on.

Why is unpleasantlness necessary? What's wrong with having high standards while treating people with respect?

What in the quoted paragraph seems like lack of respect to you?

I think it depends entirely on how it's done, really. Presumably the people who he reports to, have the same figures, so they will go into the meeting knowing that they should have thought a bit about what the problem is and should be able to suggest some solutions. It can be firm, but still friendly and collegiate. Or it could be monstrous blame-game.

Can be both and I saw both. Generally, Amazon's KPIs are pretty well aligned at least at the basic level. Also, at this level things are handled friendly. Well, as friendly as logistics during peak can be. Higher up things, at least from my experience, can be different. Together with Amazon's lean organisation, I was in regular contact with my director's director, this politics do spill down to the purely operational level. Not that great for operations. But unless there is a serious fuck up that causes a major disruption and potentially stock prices, this won't change.

That being said, the system is still so much more effective and efficient than any competition.

It depends on how much you value money, I guess. Everyone that reports directly to Dave is probably going to be a VP making 800k/year-several million.

What are the mechanisms in corporations that determine these figures? Always been super duper curious how the salary of a VP is "set"...

Not at the VP but at the C level you can get a feel by reading SEC proxies. Amounts, composition and sometimes methodology/justification are included.

VP level at Amazon has a massive scope and Amazon pays very well at high levels + their stock has performed very well recently

There are consultancies that specialize in compensation benchmarking. One is Radford Consulting. They survey companies for pay info. Imagine they also get intel from clients.

As with every role, you pay the minimum you must pay in order to attract and retain the people with the experience you need.

It is usually some combination of consultants using bogus logic, and other executives trying to bump their own pay by hiring very expensive colleagues.

I have never been convinced it is worth paying these kind of numbers. I have looked at all kinds of businesses, I have looked at businesses where productivity is power-law distributed (i.e. one good employee can be worth 10,000+ bad ones) and it never adds up...again, even if you look at situations where productivity is very unequal (which it isn't in almost every business, and very much isn't at Amazon)...it doesn't work.

One particularly toxic part of Amazon is that they pride themselves on only hiring the "elite". Well, if your market cap is hundreds of billions and you hire tens of thousands people a year...you aren't hiring the "elite" (genuinely, there are only three or four CEOs over the past five decades who are truly proven execs). So you have this awful culture where everyone views themselves as this unbelievably valuable "alpha dog". Inevitably, this results in them lighting huge stacks of other people's cash on fire (the real danger isn't hiring the 80 IQ guy, it is the guy with a 110 IQ who thinks he is a 160 IQ guy).

The majority of AMZN's tech employees (ie. average SDE) know they are not elite and often talk about how they consider themselves failures for not being getting into better paying companies.

I didn't say either that employees are elite or that they think themselves to be elite. I said that is the hiring culture, which is something quite different.

You did say

> So you have this awful culture where everyone views themselves as this unbelievably valuable "alpha dog"

In my experience this may be true of upper management, but it simply isn't true of most employees.

Depending on the RSU package, more in the million range. Already for directors that wasn't so uncommon between, say, 2013 and 2018.

Most of his direct reports at L10 are going to be $M’s and for those with larger scopes, $5M+ is very easily possible. For someone like Dave on the S-team he’s going to be $10M+ and I couldn’t even begin to guess at how much over that.

It isn’t incredibly uncommon for L8 offers to be made around $1.5M USD today in AWS - the market is very competitive for excellent leaders in Seattle. Long-time employees can potentially make more: appreciation of their equity; you have specialized or niche knowledge which commands a premium; retained because AMZN can’t easily replace you.

In summary, senior leadership positions at Amazon are extremely well compensated. It’s very rare indeed to lose someone good to another company over money.

Throwaway because its taboo for an L8 or L10 to be openly discussing $M+ compensation packages.

L8 in AWS is =/= to L8 in Fulfillment.

You were in the Andy Jassy world- Pay is higher. The Dave Clark world pays less- despite what you hear elsewhere.

Every Large FC has an L8 director... they don't all make 1.5Mil. Smaller FCs are normally lead by lower levels but again... that pay would never scale.

Reply on news.ycombinator.com