Amazon as experiment
ben-evans.com
ben-evans.com
It's clear from those letters that Amazon is not an experiment, but rather a flawlessly executed plan about how an online retailer can exploit their inherently better capital flows to grow very quickly.
It's almost the opposite of the modern startup approach.
Not: "We need to borrow money and take investment to get big and dominate the market".
Instead: "If we drive our costs down and focus on fast fullfillment or orders/inventory turnover, we'll have all this spare money sloshing around to reinvest, and happy local customers. Therefore we can't help but get big.".
(There's a lot more to it than this.)
The first letter shows that they already understood the economics of web retail very well in 1997!
Accusing them of lacking pride is also definitely missing the mark. Amazon are optimizing customer satisfaction and rate of fullfillment. They enter all these other markets exactly because they are proud enough to believe they'll win. Amazon wouldn't enter a market aiming to be 2nd best.
I haven't gotten to the present yet, so I didn't know that Amazon had opened high-street retail shops. The earlier letters said that they wouldn't do that unless they figured out how to make it scale faster than their expenditures, like their online business.
This isn’t just a byproduct of high speed/low drag commerce, they intentionally and aggressively focus on and invest in customer satisfaction.
Anecdotally and most certainly with exception in this very audience they are easily in my top three if not GOAT for direct human to human empathetic and effective customer service. Google in comparison leaves me feeling like I’m twisting random knobs on someone else’s lab project.
https://www.allthingsdistributed.com/2006/03/naked_answers.h...
Back then when he claimed that Amazon did a very good job of being customer focused, I was thinking: maybe, but whatever, I'm not sure what's special about their customer experience compared to others. Today, I'm thinking: wow, now that's a real example of audacious vision both executed and realized.
https://www.amazon.jobs/en-gb/principles
As a customer and as a former interviewee, their motives were always very clear.
The question was simple: why do gift cards have to expire? The answer that nobody wanted to say: because we don't want to leave money on the table.
It was a frustrating exercise in futility as the customer priority took a back seat at the first sign of decreased liability/free money.
Also we would cancel orders for relatively minor pricing mistakes where I believed we should have ate those costs as a teachable moment.
On the other end though, as a customer I'm more scared of Amazon than walmart. As an Amazon.com customer, I consciously avoid returning unless I absolutely have to because I care about my account. I fear amazon will honor my ridiculous requests to a point and ban me at some point. I (unsuccessfully) tried to return an air mattress to Walmart after a week of use because it didn't hold air perfectly (they never do). I don't fear being banned by Walmart.
From my experience i've had the opposite approach. Anything that doesn't live up to my standards or simply doesn't work out i return. I return ruthlessly and i've never had any problems. However my rate of return is still less than 1 in 20 and i purchase a fair amount.
I've never heard of anyone being banned for returning items on amazon?
My understanding from working on online game time / game bux cards (a little different from gift cards but not really) is that it makes the accounting much easier if your gift cards have an expiration date. Because you can't properly recognize the revenue until the gift card is used / expired, and if there is no expiration date, then you could potentially have revenue sitting around forever that you could never recognize. Terrible from an accounting / finance perspective. So it's not really greed it just makes things much easier to deal with and makes your books more sane.
That is my understanding anyways. I could be wrong.
Thus serving the convenience of the accountants over one's patrons.
Any sensible company should account for the gift card balances as a liability on their balance sheet exactly the same way a bank does its depositors' money.
The potential problem is that predicting how much liquidity you'll need to fulfill those liabilities in a given year isn't so easy. Banks are regulated to ensure that depositors are never out of their money. Some businesses have gone under as a result of getting this wrong with gift cards.
And yet there's this problem with fraud! Counterfeit products, Kindle Unlimited scammers, etc. I'm not totally sure how to reconcile the two. Possibly it's that the customer service mentality is very strong in the group that's responsible for saying "yes, you can return that" but not as strong in the engineering/product teams in retail?
(AWS is functionally a different organization than retail, which is probably worth remembering for these discussions.)
Their web site design leaves much to be desired. And it's heavy. Before I upgraded my PC, I dreaded opening a few Amazon tabs - it would bring my PC to a crawl. No other retailer had this problem. In the old days, with a single click, I could go to the Reviews page. Now the product page has no immediate click to the Reviews - I have to scroll down and click on it. Clicking the link on the top just takes me to the shorter review at the end of the page (and I have to scroll again to get to See All Reviews).
When I write a review, I have no way (that I know of) to get notified of comments to my review. If there's an option for it, do let me know. I just checked and I can't even find a link to see my own reviews in my profile's page. Anyone know where it is?
I've also sold by FBA. Their web site for sellers looks like it was made in the 90's, and discoverability is atrocious. I have to keep trying different links to see if it will take me to the report/inventory view I'm looking for. And I've dealt with several bugs on those pages (e.g. updating a price would not work sometimes).
Then there's their policy of comingling, so that when you buy a brand new item that is "Shipped and sold by Amazon.com", there is no guarantee you're not getting the item that an FBA seller put into their warehouse. Hence, no way to guarantee you didn't get a fake product.
Then when you check out sites like fakespot.com, you'll find how many products have incredibly gamed reviews. Amazon's reviews was what set them apart, and is now horribly unreliable. And then add to that mix the fact that many reviewers are actually reviewing the fake product that was shipped to them instead of the bona fide on. How do I know if the product is bad from the reviews? I can't.
And then to top it all off, when looking at one product, it's actually a combination of several ones that you have to pick. In some cases this makes sense - say I'm buying a particular brand of slippers, and they have different colors, that's fine - as long as the "model" is the same. But when I'm looking at a whiteboard, and the other option is a set of markers, that's ridiculous. The overall rating of the whiteboard has ratings of the markers mixed in. Now if I want to know the rating of just the whiteboard, I have to go through extra steps (click on Format, etc).
> Within AWS, that same pattern has recurred many times. For example, we invented DynamoDB, a highly scalable, low latency key-value database now used by thousands of AWS customers. And on the listening- carefully-to-customers side, we heard loudly that companies felt constrained by their commercial database options and had been unhappy with their database providers for decades – these offerings are expensive, proprietary, have high-lock-in and punitive licensing terms. We spent several years building our own database engine, Amazon Aurora, a fully-managed MySQL and PostgreSQL-compatible service with the same or better durability and availability as the commercial engines, but at one-tenth of the cost. We were not surprised when this worked.
Didn't Gmail start as someone's 20% project, rather than something that was "systematically rethought from first principles"?
Gmail's creator, Paul Buchheit, said it was not a 20% project:
https://time.com/43263/gmail-10th-anniversary/
excerpt: >Gmail is often given as a shining example of the fruits of Google’s 20 percent time, its legendary policy of allowing engineers to divvy off part of their work hours for personal projects. Paul Buchheit, Gmail’s creator, disabused me of this notion. From the very beginning, “it was an official charge,” he says. “I was supposed to build an email thing.”
Even though Amazon failed with the Fire Phone, I wish they'd try again! I think Amazon is uniquely poised to do this. Have a half-smart phone that includes their back-end services (Alexa, shopping, movies, music, storage for your photos) and a selection of curated apps (Uber, Lyft, some popular games) and a decent browser, and you may have a good phone for older people, people who want a nice phone at, say, a $150 price point but don't want to deal with complexity
It might be just a bit of nostalgia, but I honestly miss my Fire Phone a lot. It did everything I needed it to, was fast, and actually fit in my pocket.
[1] Unlike generation-old iPhones that get slowed to a stop by software updates, the thing never got updated, and so it was still quite good [1.1]. Yay, I guess?
[1.1] I came from a really cheap BLU-branded phone, and so anything was better, really.
The Fire Phone arguably was Jeff's pet project. E.g. the four cameras facing forward were his idea if I remember well. Then it failed. What should have happened, based on leadership principles, would have been for Jeff to own it and admit it (vocally self critical and so on). That didn't happen, my impression was that blame was put on the lower echelons of the organization. For that was, in retrospect, the point when leadership principles started to be used as a political tool by some people. After that, at least for me, the culture changed as well. It took more than a year for that effect to trickle down, so. Amazon was still a good yet challenging place to work, but it felt much more like any other large corp out there.
Are we starting to see that effect of culture change on results ? How ?
Or did they somehow managed to prevent that ?
One side effect of all that was managers being more risk averse. Some initiatives that could have been at least tried started to be blocked by red tape until the political environment was favorable. Did I mention that Amazon already was a pretty political place?
This form factor would be revolutionary for many people, and amazon could pull it off.
Some people truly believe in the LPs. Some people truly believe they can be used to get ahead and justify whatever actions they want to do. I've never worked directly with Jeff, so I have no idea what his approach to them is.
The Fire Phone? That best analogy I've heard for it is that the Fire Phone is to Jeff as the "Homer" car is to Homer Simpson[0].
[0] https://imgur.com/a/o0Q6fw6 and https://simpsons.fandom.com/wiki/The_Homer
EDIT: It's true that some people really believe and live the LPs on a daily basis. My first biss was one of them, Amazonian from basically day one in Germany. He retired earlier this year. Since I assume he never sold one single share he received (frugality anyone?) that really means retirement now. Pitty that he left, so. After I did, but still.
They can be the same people. There's an incoherence in any ethical system based on a set of values.
The false justification isn't necessarily intentional. Motivated reasoning[1] is a natural cognitive bias, so while you're cherry-picking values that support your desired ends, it doesn't feel like you're doing it.
A better way to approach the LPs is to recognize that they're smart business practices, but that they're subordinate to professional ethics, which in turn are subordinate to a moral code.
Oh, sorry, I meant that they are different people. Though I've heard of and met one or two people who might be crazy enough to be both.
The way Amazon likes to solve problems is by creating platforms.
So they created FBA, and now others discover products for them.
They also created their affiliate program, and that's a good way(combined with other business Intel methods) to discover online buying experiences with potential, and than scale them.
And when they do see something non-standard with potential, they invest.
For example: printed t-shirts. They copied that idea.
And how do they win here?
A combination of a lot of traffic, the best logistics, and maybe cheaper manufacturing costs(due to scale), combined with Amazon's execution.
And that could be a general method for creating new experiences.
This model is very hard to beat, since usually the emerging platform operate at a loss, which is big barrier to entry.
The key to win here is not to fight the emerging platform but the supporting platform. For example, if one attack AWS, it would likley take down the eCommerce part.
People visit mall stores sporadically enough that they might forget about the bad experience they had or hope it might be different or feel they don't have a choice.
Thus mall stores can damage their brand image for a quick buck; in fact they are often brain washed by their own advertising and have no idea how the customer sees them.
Grocery, on the other hand, has a customer that does back at least once a week, if the brand image is not upheld they can get similar goods elsewhere.
That awful whole foods at union square with the cattle chutes at the checkout might look good compared to the local bodega in NYC, but does not look good compared to a good suburban supermarket (Wegmans). It is one case where urban folk could be out of touch with the broader market.
Describes every facet of a city IMO.
Everything that's boxed, bottled, or bagged, I can get from Amazon. Instead of making a list I just scan the barcode and order and throw the empty one away (or put the almst empty one in the back in the fridge)
Also worth noting that since that whole foods opened, the grocery situation throughout a lot of manhattan improved in that they opened quite a few more locations that also were rather large compared to the older grocery stores around.
This all said, I think the real market in a place like manhattan is for grocery delivery. Fresh Direct when they opened was tremendously popular right away, and given amazon's hold in the online purchasing space, within manhattan (and other cities/neighborhoods where few people drive), online grocery purchasing experience matters more for the brand than in store experience.
Where I think they do get it right in that sort of space is the Subscribe and Save programme which must cannibalise a fair amount of supermarket purchases. The "set and forget" nature of it works really well for stuff like nappies/diapers (it's a bit more painful for things like toothpaste because the quantities/frequencies aren't quite right).
Historically the quality of their fresh produce was also relatively poor (not particularly fresh and lower grade I think, although that just put them on a par with eg Asda or Morrisons) but I don't know if that's changed since the Whole Foods purchase (I'd like to think so).
https://www.amazon.co.uk/1908-Sears-Roebuck-Catalogue-Schroe...
There’s also a copy on the Internet Archive, I’ve yet to find the portable forge:
https://archive.org/details/catalogueno11200sear/page/760
(Scales is singular, isn’t it?)
Such an amazing catalog, I could spend many hours paging through it.
Of course, AWS is a different story.
What fixed costs are you referring to? It seems to me that almost their entire cost structure is variable and quite low, compared to competitors.
1. Private labels enjoy about 10% extra margin. Amazon is currently scaling private labels and exclusive products(same thing), like mad.
2. Low margin doesn't mean a business is bad, just hard. Walmart is an amazing business, with low margins.
3. Let's say amazon creates a system that can supply everything, to your home, in a few hours, at the lowest price. And customers will really like that.
How many other companies can build that ? maybe 1-2 in the US.
So now you have a duopoly.
4. Amazon controls demand. In the internet age, that's the most important thing.
Could they manufacture their own products ? be responsible to purchase inputs for their manufacturers ? Build something like AWS for manufacturing ?
Who knows.
5. Amazon is growing rapidly in B2B. Not sure, but i believe products there have much higher margins.
And any major corporation can use the same supply chain to make the same private label goods. All commodity products are a race to the bottom.
2. Low margin doesn't mean a business is bad, just hard. Walmart is an amazing business, with low margins.
Not a bad business, but low margin businesses don’t command the multiple that Amazon has.
3. Let's say amazon creates a system that can supply everything, to your home, in a few hours, at the lowest price. And customers will really like that.
Yes and then you increase the marginal cost for shipping....
How many other companies can build that ? maybe 1-2 in the US. So now you have a duopoly.
You don’t need many companies to erase margins if they are competing on price.
4. Amazon controls demand. In the internet age, that's the most important thing.
Amazon doesn’t control demand. But Amazon retail has the highest marginal costs of any of the tech companies including Apple. Netflix, Google, and Microsoft (software not hardware and cloud) have much lower marginal costs. Software has basically zero marginal costs.
Could they manufacture their own products ? be responsible to purchase inputs for their manufacturers ? Build something like AWS for manufacturing ? Who knows.
Apple is considered the best in supply change management in the world. They have huge bargaining power because they sell a lot of a few goods. Amazon would be selling a relative few of a lots of stuff.
5. Amazon is growing rapidly in B2B. Not sure, but i believe products there have much higher margins.
Businesses have an entire procurement department to make sure they are getting the lowest prices. Those margins should be even lower.
The industrial, Full Truck or Container Load B2B supply chain is different beast. Not that Amazon couldn't figure it out, especially the logistics side. They are doing anyway already for themselves. Organization wise I think it is a different story. And in that field I always have a hard time to understand all the potential profits customers leave on the table. At my current gig procurement is buying full containers from Asia that are than directly delivered to our sites. They buy from a European trader only placing orders in Asia after we did. And the Asian supplier is only starting production after the trader ordered. Crazy business, right? And there are a lot of B2B customers like out there, other people's lazyness seems to be a solid basis for a company.
// You don’t need many companies to erase margins if they are competing on price.
Can you explain ? Because my mental model of this is wireless carriers, cable TV, etc, with good margins.
// Apple is considered the best in supply change management in the world. They have huge bargaining power because they sell a lot of a few goods. Amazon would be selling a relative few of a lots of stuff.
Xiaomi does that sucsessfuly over a large range of electronic products.
Also, sure Amazon won't get to 50% margins like in software.
But could they get to 10% ? Or maybe 15% ?
And they'll surely grow, globally.
Would that be enough for the p/e ?
Cable carriers usually have a monopoly in the places they operate. We have seen carriers have to be more competitive as T-mobile got aggressive as prices did fall.
And they'll surely grow, globally.
Growing selling physical goods aggressively means setting up more warehouses, more expensive logistics, catering to different languages, etc.
In the worlds biggest markets - India and China they would have to compete with entrenched incumbents and in the case of China, government subsidized businesses.
Amazon is 25 years old.
I saw a comment on Medium the other day that summed up my feelings on Amazon:
Third, his “business” isn’t even a business. Amazon, for 99% of its history, lost more money than it earned. The reason it even still exists today is because it was propped up by truly gigantic loans in the form of stock shares and VC capital. Propping up a non-profitable company until it grows big enough to crush any and all competition is literally the Communist Chinese model of “capitalism.”
Fourth, all of Bezos’s “wealth” is built on the backs of overworked, underpaid, and horrifically mistreated employees who aren’t even allowed enough time to use the bathroom, a sizeable percentage of which are elderly folks who, in a more just society, would be enjoying their retirement.
https://medium.com/@lifeinromania/jeff-bezos-is-a-fraud-458f...
Another note from another site The Street:
Using this estimate, the overall business, without AWS and Prime, lost about $2 billion last quarter. Although advertising is included here, too, this loss is almost all from the retail side of the business.
https://www.thestreet.com/opinion/amazon-is-losing-money-fro... operations-14571703
I don't doubt that Bezos is wicked smahhht but it doesn't take a genius to figure out that if you borrow so much money that you can run a business at a loss for decades and undercut everyone that you'll be successful. Except he clearly isn't even able to do that.
And those people are being rewarded for it..so what?
> Except he clearly isn't even able to do that.
Except he does. They optimize for cash flow, not income, and they're absolutely crushing it in that department.
Let me kindly refer you to Amazon's cash flow statement:
https://www.marketwatch.com/investing/stock/amzn/financials/...
For more fun reading: https://25iq.com/2014/04/26/a-dozen-things-i-have-learned-fr...
> Fourth, all of Bezos’s “wealth” is built
Bezos wealth is built by shareholders who invested in Amazon. What you're referring to is a moral and regulatory criticism, which frankly shouldn't be a criticism of Bezos himself but rather the people who govern his company and wealth.
> So maybe that’s the real test of Amazon’s pride: can it work out how to let us shop, rather than just buy?
Why even write this article?