The Relentless Jeff Bezos
stratechery.com
stratechery.com
I like to joke with my friends that AWS is responsible for the greatest second-order GDP growth on the planet, if you add up all of the market caps of all companies that have been able to build off of AWS (or their competitors). Rivaled only by MS Excel, as far as I'm concerned.
Up until this point, I thought you were simply right (also: worked at AWS 2008-2014 and have a specific perspective on why I think you're right).
Then, when you mentioned MS Excel, I thought: brilliant. Simply brilliant. I feel I'm one of the few that sees MS Excel as the biggest software winner of the past 40 years. So when I see someone that thinks like me, I rejoice.
> worked at AWS 2008-2014 and have a specific perspective on why I think you're right
Would you be able to talk about this? I’m curious.
1. We had an in house production engineering team, manage our own BGP and PoP points.
2. As we started to scale, we required active-active replication across multiple datacenters with high SLA requirements, which spawned a year-long project with multiple engineers to build in-house.
3. As the team grew, we required some sort of pub/sub system to enable cross-service communication — which spawned a year long project with a handful of engineers to build in-house.
4. We had to deploy our own load balancers in our datacenters and ensure that they scaled as we grew.
Today, my tiny little team has access to tooling that enables pretty much everything I described above for a low monthly price. Elastic load balancer gave us (4) right out of the box, RDS (or Dynamo) gave us (2) right out of the box, SQS + SNS gives us (3) right out of the box. Within the AWS VPC in us-east-1, we hardly ever think about (1). Today, we just had a need to do OCR (optical character recognition) of receipts because it was becoming untenable to have humans read every single on of them. Once upon a time, that would have been its own big engineering workstream. Instead, we had an intern integrate with AWS Textract (https://aws.amazon.com/textract/) and called it a day.
These anecdotes just scratch the surface; if you go on https://aws.amazon.com/, every single one of those products were once months-to-years-long multi-engineer projects that had to happen in-house at every tech company as recently as 2012. The sheer breadth of what AWS allows startups with limited resources to do just can't be understated, and as a startup employee myself, I appreciate that first-hand.
I'm glad it's been useful for you though.
I can literally pay for a $5 VPS and get off the ground in mere weeks as well. The only thing I don't get is the potential performance of AWS, but you have to pay extra for that at AWS as well, so paying extra outside of AWS for it isn't outlandish.
Now, AWS has a lot of value addons that can make it worth it/useful, so perhaps there's a point there but it's not nearly as noteworthy as the other poster is making it seem.
Fast forward three days later and I get email from Amazon out of the blue saying they realized the audio of the episodes in the season were messed up. Not only did they fix the audio problems (the next time I queued up an episode, it was fine), but gave me a full refund. This was totally unsolicited and it blew my mind.
Yeah, they're backsliding. They need to get their house in order.
This isn't a naive statement - this is a reflection of other markets that have had decades longer to manifest in the various societies compared to Amazon.
Food quality is higher, waste is lower, trust in government is higher, social programs can succeed, infrastructure projects are cheaper - all of this is universally true across progressive nations from Western Europe to Canada, compared to the US.
What I don't like of Amazon now is how it is flooded with Chinese crap. It looks like Aliexpress.
If you peruse some of the sneakerhead forums, you'll see a lot of complaints about counterfeit dealers on Amazon. Nike used to partner with Amazon so you were assured your Nike's were coming directly from Nike. Nike killed the deal in 2019 and so now you have no idea where those Nike's and other sneaker companies products are coming from:
And the Wall Street Journal reported this week that Nike was “disappointed the deal with Amazon didn’t eliminate counterfeits and give the brand more control over gray-market goods,” which are legitimate items that Amazon sellers buy from distributors or retailers, rather than from Nike itself.
The problem for Nike is, removing itself from the Amazon platform won’t fix that problem. In fact, it’s likely to get worse. Amazon will now look to acquire more “gray market” Nike products to fill in the inventory gaps Nike’s departure will leave behind.
https://www.vox.com/recode/2019/11/14/20965434/nike-amazon-p...
Now Chinese Nike knockoffs are showing up under dubious brands like "fashion" or "Vitike":
https://sneakerhistory.com/2019/11/14/8-unbelievably-fake-sn...
The question wasn't whether it happens in America. The question was whether it also happens in Europe (personally I've always assumed it does but my point is your articles don't address that despite being in response to that question).
Similarly if you bought a pair and they didn't fit, your only option is to throw them out and buy a new pair. No reselling.
Gray market is very far from chinese knockoffs though.
Nike trying to cast distributors and retailers as "gray market" is a little gross.
Jessica — not her real name — has spent well over $15,000 on Amazon this year, buying everything from Halloween decorations to a queen-size inflatable mattress. She's purchased over 700 products, including three vacuum cleaners, six desk chairs, and no fewer than 26 pairs of earbuds. And even though most of the products are cheaply made, she’s given each a 5-star review. The twentysomething who lives on the East Coast isn’t a bad judge of quality — the companies that sell these products on Amazon reimburse her for the purchases.
Third-party sellers know what it takes to make it on Amazon: Get good reviews and a high search ranking. But attracting genuine customers is tough, so some sellers use a reliable cheat — bribes. Because of Amazon’s vast scale, inscrutable algorithms, and capricious enforcement of its own rules, unscrupulous sellers and paid shills largely get away with it.
https://www.buzzfeednews.com/article/nicolenguyen/her-amazon...
In fairness to the GP, it does seem less prevalent in the UK (either that or the reports about fake reviews are greatly exaggerated in general) but there's definitely been a few reviews here that have "smelt off" -- either before purchasing or after having received the product. And I definitely know these bribes do happen too as I know of a few people who have been contacted. All of which refused to take the money.... but please do take this anecdotal sampling with a pinch of salt.
Because Amazon has exactly the same problems in the UK, Germany, and Spain.
Delivery and logistics are outstanding. But the rating/review/counterfeit/duplicate issues are a complete dumpster fire, and have made me pull back from using the site.
Maybe I just don't buy brand stuff?
I've bought stuff that turned out to be shit, but I never thought there was a "real" version of it anywhere, just a shitty product.
Unfortunately, while it's relatively straightforward to e.g. detect fake reviews in a static dataset, it's a constant arms race when you're doing so in an online adversarial setup. Same goes for counterfeit 3P goods.
Disclaimer: I work for Amazon, but not anywhere near the retail side.
The same would happen if e.g. half of Amazon customers would use one of these solutions.
The difference between being online versus being in an adversarial scenario largely depends on how / to what degree you're visibly impacting the bottom line of the scammers.
If those things didn't work reliably, I don't imagine the bad sellers would keep doing it
If by that we mean "token goodwill gestures where it doesn't hurt our bottom line", then yes.
If they cared for the customers they'll have improved things that affect hundreds of thousands or millions or all of them (tons of such long-standing, including for the Kindle, which has languished as hardware in neglect-land).
Doing a fuss to make the release day the same for 12 customers (or 100) in the Isle of Man is not "caring for invididual customers". It's either a token empty feel-good gesture within the company, a whim of some superior, or a "nice story" empty marketing stunt.
Turns out I'm not the only one:
https://www.avforums.com/threads/chili-tv-audio-compression....
Everybody loves that Amazon and Prime bought and made new episodes of the Expanse, right? Well, on my stock Sony Bravia Android TV, if you are watching the Expanse with English subtitles, when the text for the subtitles is on, the screen is brighter. It's noticeable, and quite annoying, and this ONLY happens on the Expanse. Other apps and other shows on Prime work.
Having said that, one of the reasons why I'm a huge Amazon fan is because they have such great service. For example, I purchased a PS4 that died 3 days in and it was the easiest thing in the world to get it sent back and replaced.
It's one of the reasons I generally trust ordering expensive items through Amazon.
That is small time customer service. Even Costco and Nordstrom’s and Trader Joe’s do it.
But the big time stuff, like Amazon providing retail products via a trusted supply chain, are where they screw you. It’s a low margin business, and Amazon wants as little to do with it as possible. So they removed the filter for showing only items shipped and sold by Amazon.com years ago, and they commingle inventory, so you have no idea of the supply chain of your product. Could be coming from some random product by some third party seller. Who knows?
Obviously, Amazon wants that 15% top line revenue from being a platform rather than the 2% to 5% they get from sourcing and selling retail products. That doesn’t align with my interests as a consumer.
When?
Living in the city, I think it's easy to quibble and groan about their offerings or subtle changes in their service quality. But sometimes it's helpful to step back and marvel at the sheer scale of what they built.
I've attempted to buy from other places and the experience has always been so terrible that I just end up buying even more from Amazon. As an example, I bought some stuff from Lowes, they sent me the wrong items, and my only choice was to drive a few hours to return it. Unless I price my time at $0, this isn't worth it. Another incident happened with Home Depot, where their delivery person was some random dude who banged on the door loudly in the middle of the pandemic saying that they absolutely had to get my signature on a <$100 order. I begrudgingly opened the door and signed, and then decided to never buy from them again. Something like this happens every time I try to buy from anywhere that isn't Amazon. On Ebay, the last time I used it I received a very used item when it was supposed to be new. The reason Amazon is doing so well is just that everyone else is so terrible. I've had issues with them as well, but it literally takes me minutes to resolve any problems.
I think his mother was very good at encouraging his interests and nurturing that relentless spirit.
Btw, a bit of trivia, try typing relentless.com into your address bar.
I've written to cs@amazon.com about it a couple of times and not gotten any reply whatsoever.
“Import costs include the Import Duty (II), the Merchandise and Service Circulation tax (ICMS) and the Brazilian Federal Value-Added Tax on Manufactured Products (IPI). Import duty is a federal product-specific tax levied on a CIF (Cost, Insurance, and Freight) basis. Rates usually vary between 10% and 35%. The IPI is a tax levied on finished products (whether foreign or domestic), resulting from some sort of industrial process. The IPI is not considered a cost for the importer, since the value is credited back to the importer. IPI rates range between 0% and 15%. The ICMS is the VAT levied by states and applies to the legal, physical, or economic circulation of goods, whether imported or not, the services of transportation and telecommunications.”
https://santandertrade.com/en/portal/international-shipments...
source: brazilian, talked to people that know more about this than i do.
Relevant thread from earlier today that discusses, if relentless.com was considered as a name for the company before it was amazon.com - with Amazon employee #2 chiming in.
https://news.ycombinator.com/item?id=26007814
Seems like relentless.com was first registered on 24th of September 1994 (a Saturday), just 37 days before amazon.com was registered on the last day of October 1994 (a Monday).
Thanks for all that, Jeff.
Many claimed that he was possibly one of the greatest CEOs in the technology sector, sharing a podium with Steve Jobs and Bill Gates.
The new blockchain-based millionaires/billionaires will be minted starting these days.
No, not cryptocurrency. Please don't conflate the two. Cryptocurrency is a side effect of blockchain.
- supply chain transparency (with the right crypto model you can do things like that different parties have different visibility into transactions)
- food traceability
- fighting counterfeit items (on amazon!)
- a review website where reviews cannot be tampered with (bye bye google or amazon removing negative reviews)
- logging stack for IT operations that cannot be tampered with
I don't fully understand that reason. It was trivial to get one of the millions of books not in stock at a book store even before Amazon. You could just ask for it, or call them, and they would order it. You could usually pick it up the next day. After all, they were using exactly the same infrastructure to order the books as Amazon. So basically any half-decent bookstore was offering millions of books decades before Amazon did, and any half-decent bookstore would also do home delivery.
1. You had to go through their order process, talking to someone, completing the ordering process, and then coming back to pick it up. If you live 45 minutes from the book store, that gets to be a pain.
2 You have to know the book existed. I don't think I ever really ordered a book, because for the most part, I didn't know what alternatives existed to what they offered. Amazon helped fix that. Instead of finding say one book on Python, there were a dozen on Amazon's site.
Going to the store in person? Or calling them on the phone??? It's not 1874 anymore, nobody wants to do either of those things.
Signal-to-noise. I'd love having a knowledgeable consultant to help cut out the noise.
For instance, I'm kinda-sorta in the market for a new mechanical keyboard for my home office. (I got one for work years ago and might want something a little different.) The price point is enough that I want to have some confidence before an online order. The Internet has tons of data that are nowhere close to giving me the knowledge I need to make an informed decision about whether it will satisfy my desires.
Crucially, I also don't know exactly what questions to ask to winnow the universe of possibilities into something more manageable.
Being able to consult with someone in the absence of having access to a plethora of keyboards would be huge—the feedback is killer.
I suspect books are actually very similar in this regard. The world of books is huge, and having a guide to get you to what's helpful is really important. I don't have time to read everything to get to the Good Stuff.
Every friction point in the customer experience causes potential customers to lose interest and give up on their prospective purchase. So if your potential customer has to take two actions, spread across multiple days, then all else being equal you're (hopefully obviously) going to get lower conversion/purchase rates and lower customer satisfaction than if they can impulse-buy in one action.
I was repeatedly told that working on games outside office hours was not allowed, then later was permitted but only if you were willing to give Amazon all related IP. Oh, and you could not collaborate with people outside Amazon, so no game jams with friends.
To say nothing of the treatment of workers in the FCs, relatively low engineering compensation, etc.
Also, starting an online business is easier than ever due to democratized access to infrastructure that was once only available to large blue chip companies with access to 1000s of high skilled engineers, and has seeded an industry of small/medium sized tech companies that has resulted in further consumer surplus and wealth creation.
Not saying that there's an obvious "correct" point of view, but once you add a little more nuance, it actually does make everything worth it for a lot of people.
On the backs of warehouse and distribution workers who put themselves at serious risk for Covid (or just backbreaking work environments).
And while it's definitely difficult that they have to put themselves at serious risk for Covid, that's true for any essential worker too. Ultimately most people agree that this is a worthwhile tradeoff to ensure that people continue to get access to food, supplies, COVID tests (in the case of frontline nurses), etc.
I'll re-iterate, there's no obvious correct answer, and it's difficult to agree on the best way to litigate these issues. One thing I can say for sure is that blithe one-line HN comments is the worst possible way to litigate them, much less understand them.
Workers pass out in warehouses and aren't allowed toilet breaks. And many are gig-economy contract hires who have zero chance of ever seeing a 401k or any other permanent employee benefits.
Nobody is eulogizing anything. The system isn't dead, it's alive and well. What I, personally, am doing is simply appealing to nuance.
> Workers pass out in warehouses and aren't allowed toilet breaks
Yes, this is bad. Nobody is arguing that it's all sunshines and daisies for Amazon Warehouse workers. The argument is that the reality is more complicated because they are well compensated relative to their similarly-skilled counterparts elsewhere. And it's not like they're not allowed toilet breaks as a rule, the reality is that Amazon's warehouse policies are decentralized and in one documented situation, there was malpractice. This is undeniably bad and should be fixed, and one would be a fool to argue otherwise. Notwithstanding (objectively bad) isolated incidents like this, in the aggregate, it is arguably a good thing that 1.3 million unskilled people are gainfully employed with competitive benefits, and all in service of providing convenient supply chains to the world especially in times of widespread pandemic and distress.
> And many are gig-economy contract hires who have zero chance of ever seeing a 401k or any other permanent employee benefits.
At least as it relates to the warehouse workers in the United States, this is factually untrue.
The hours are predictable, and there's no oncall rotation so you can reasonably plan your life around the difficult working hours, and you never have to worry about not affording health care because Amazon health insurance is quite generous.
That a warehouse worker might benefit less from 401K matching doesn't refute the central argument: Amazon provides high pay and generous benefits atypical of entry level unskilled work that doesn't require a college degree. Not only that, they offer this to 1.3 million people. As a fun exercise, I recommend just trying to do the back-of-the-napkim math and try to work out the total cost to providing $15/hr (at a minimum), generous group health insurance (usually working out to about $1000/month in premiums), 401(k) matching, paid parental leave, and then multiplying that by 1 million (my rounding error there is more people than employed at most large tech companies).
* This is specifically parental leave, not just general purpose leave. And it's only 6 weeks for adoptive parents.
* There are reports of people losing healthcare coverage while on leave - right after having a baby? Really?
* Reports of people being put on PIP after their leave was over, which as we all know is a threat that you might lose your job
* Reports that people are not pushed to use their leave and there is sometimes pressure to not use it (risk of PIP, “Boy, will you be sorry if you use it.”)
* Parental leave doesn't extend to hundreds of thousands of temporary, contract, and part time workers
https://www.fastcompany.com/3053093/whats-missing-from-amazo...
I would also say it's not a great argument to say how much Amazon spends in aggregate going "above and beyond" on its employees. I doubt it really affects their bottom line in any material way.
$15/hour is also looking to soon become the minimum wage. I guess a pat on the back they did it before they were legally forced to. $32k a year is not what I'd call "high paying" though, even for unskilled entry level pay. Many waiters and bar tenders in big cities get paid a lot more than that.
> * This is specifically parental leave, not just general purpose leave. And it's only 6 weeks for adoptive parents.
I'm not sure that I insinuated otherwise.
> * There are reports of people losing healthcare coverage while on leave - right after having a baby? Really?
> * Reports of people being put on PIP after their leave was over, which as we all know is a threat that you might lose your job
> * Reports that people are not pushed to use their leave and there is sometimes pressure to not use it (risk of PIP, “Boy, will you be sorry if you use it.”)
Do you have an actual link with context? Otherwise this is just unfalsifiable conjecture. The links in your article are 410'ing. Keep in mind that you can't draw conclusions from one off incidents. Even non-Amazon workers have work horror stories — they happen. Whether this is emblematic of the experience of all 1 million-odd workers is another question entirely.
> * Parental leave doesn't extend to hundreds of thousands of temporary, contract, and part time workers, contrary to you saying it covers 1.3 million employees.
Sure, but that's true anywhere. What we do know is that Amazon does not rely predominately on temporary contract workers. It also heavily narrows down the scope of your argument to "Amazon probably doesn't treat its contract workers well", and I might agree with that. I also, personally, think that Amazon is still a net-good to society in spite of that because I'm not an absolutist.
> $15/hour is also looking to soon become the minimum wage. I guess a pat on the back they did it before they were legally forced to. $32k a year is not what I'd call "high paying" though.
First of all, it's not even obvious that $15/hr is soon to become the minimum wage, Joe Manchin has signaled opposition to it (https://twitter.com/bpolitics/status/1356676773971456003). It'll most likely be left out.
Secondly, $32k a year isn't the total compensation, because you have to take into account the monetary value of Amazon's group health insurance policy (it's around $1000/month).
Thirdly, $32k (+ ~$12k) is around the median wage in the US, and that happens to be in the top 5 in the developed world (https://en.wikipedia.org/wiki/Median_income#Median_equivalen...).
And fourthly, this is for an entry-level job that doesn't require a college degree or any sort of training, it's obviously not going to be 6 figures, nor is it going to be on par with (say) teacher's salaries or police officers.
Back in the old days, THE CLOUD came from IBM or CompuServe. Now it comes from Amazon and Microsoft. Big deal. It's still way more expensive than buying hardware, and now every programmer also has to work as an incompetent sysadmin and DBA.
I guess that if you see a spike in traffic, AWS will let you bankrupt yourself from the bills rather than take your site offline for an hour.
I'm willing to concede that Amazon may well have driven overall consumption up. Also, anything being shipped by air is causing way more emissions by weight than any amount of people driving to the stars. But I still wonder how this balances out against the efficiency of a delivery van; I don't think the answer is that obvious.
I think you might be misreading my comment to suggest that Amazon is unambiguously good. Read it again, I'm trying to argue that it's not unambiguously bad, and that we need more nuance when discussing these issues. I'm not sure if it's possible for an argument to be any more nuanced than that.
Regarding the climate issue, we agree! It's a big problem. But again, we need to put our nuance hats on:
1. While Amazon is a substantial contributor to CO2 emissions, it's nothing compared to that of the coal industry, airlines, and everyday drivers of automobiles. That being said, it's worth noting that US renewable energy consumption surpassed that of coal for the first time last year -> https://www.eia.gov/todayinenergy/detail.php?id=43895
2. Amazon is simply meeting the needs of its consumers. Ultimately there's a cost to convenience (there's no denying that), and it comes at the cost of the climate. We agree that we'd be better off if externalities were priced in. The fact that externalities are not priced in isn't really Amazon's fault, because they're in no position to enact such policies as a whole. They only have control over their own policies, which brings me to...
3. Amazon is on track to power their operations with 100% renewable energy by 2025 (https://www.aboutamazon.com/planet/climate-pledge), and invested $2 billion (with a B), in green energy companies (https://www.aboutamazon.com/news/sustainability/amazon-annou...). So it's not like they're doing absolutely nothing in their power to at least check their own emissions.
Now, all this said, is it still not worth it? Maybe, I don't doubt that there exists some people that believe that the warts are big enough that none of this is worth it. But it isn't obvious or straightforward, and there are just as many people (if not more) that do think this is all worth it, warts and all.
How's that for nuance?
What I'm saying is that I find it difficult to praise a company for its success when that success is at least in part predicated upon costs that they do not bear. In other words "profits for them, costs for us all". All that could be solved, or at least improved, by adequately pricing in those externalities, or even restricting them all together (e.g. for emissions: carbon tax + hard cap).
And I don't buy the "they're not in control of whether or not externalities are priced in", they have immense lobbying power to de facto subvert democracy and overpower the will of the people. They can kill or stall legislation that harms their bottom line, and of course they will! It makes them more money so why ever wouldn't they?
What evidence do you have that Amazon, specifically, has done this? And why would Amazon work towards becoming 100% renewable by 2025 (surely at a high cost to them) or investing $2B in climate action (that's roughly how much they'd have to pay in carbon pricing anyway), if they actually wish to lobby against carbon pricing legislation? That is internally inconsistent.
I have exactly _once_ received an item that was obviously fake.
Every time I purchase elsewhere I am shocked to be charged shipping, wait days for the item to be shipped, and go through hell if I need to return something. Amazon wins because it’s a better product in every way.
i think you can have a better than 50/50 chance of getting a decent item from amazon. you just have to be picky about what you buy from them. for example, i would never buy an apple charging cable from amazon. no matter what the listing says, it is probably going to be a cheap chinese knockoff that might set my house on fire. for that particular case, i'd buy it from apple instead.
about ten years ago, i paid US$1500 for a brand-new sony teevee, from amazon. it has always worked fine, and i am still using it today. (i'm glad i bought it before "smart" teevees that phone home to the mothership became the norm.) my trust in amazon was much higher in those days. really doubt i would ever do that again. for any moderately expensive piece of high tech gear, i'd try to buy it directly from the manufacturer, or from a reseller that specializes in that particular type of thing.
if you limit your amazon purchases to medium-expensive items that are not easily cloned in china, you'll probably be fine. it is kind of a shame that you have to be so cautious, though.
okay, actually, i lied: i wouldn't buy anything at all from amazon these days. they are looking far too much like a monopoly for my tastes, and i do not approve of the way they treat their minimum-wage workers. it has been a struggle, but i recently abandoned my amazon account, created in 1998, and have moved 100 percent of my substantial amount of e-commerce to other sites: chewy.com, target.com, crutchfield.com, b&h photo and video, newegg, ebay, and so on. almost all of them have a substantially worse experience than amazon did, but i am willing to suffer a bit of inconvenience for this particular bit of principle.
Amazon doesn't have minimum-wage workers, at least in the United States [1]. The starting wage at Amazon, $15 USD per hour, is more than twice the U.S. minimum wage.
[1] https://www.nytimes.com/2018/10/02/business/amazon-minimum-w...
UK minimum wage is £8.20. Amazon warehouse basic day rate is £9.70, less deductions.
> That's not the starting wage, that's the "average" wage as reported on GlassDoor.
No: see the link in my previous post. In the United States, the lowest starting wage for Amazon workers (including warehouse workers) is $15 USD per hour.
Bonus points for getting the manufacturers themselves to link to this company directly from their own websites. Obviously, they'd have to cut a sweet deal for providing the fulfillment services, but for many items, I'd rather pay a slight premium to ensure I'm getting the real thing.
In the past 20 years I've probably ordered about a thousand things from there and have never received a counterfeit item.
I'm not saying it is good for technology, or innovation. But one cannot deny the advantage it provides to key Amazon stakeholders.