Hacker News new | past | comments | ask | show | jobs | submit login
Amazon’s New Customer (stratechery.com)
293 points by darwhy on June 19, 2017 | hide | past | favorite | 112 comments

I had to exit at this quote:

Apple’s goal was not to build a phone but to build an even more personal computer; their strategy was not to add on functionality to a phone but to reduce the phone to an app; and their tactics were not to duplicate the carriers but to leverage their connection with customers to gain concessions from them.

I disagree here, this is hindsight and speculative, and the provided link is from 2013. The initial iPhone did not come with a market (the 'iTunes App Store') and AFAIK we don't know for sure if it had been planned from the beginning to pan out the way it did.

I am not trying to play down the success of the iPhone and how it has changed the world. Yes, it was incredibly successful, yes it all panned out nicely. I am just questioning if it was all planned to come out exactly this way from the beginning, or maybe there were some decisions made after initial success and someone came up with even better ideas.

Ben Thompson(the author of stratechery) is a smart guy, but his free articles, even though that they seem deep, do lack domain/technical knowledge. I've noticed that on the subjects of self-driving-cars, car-sharing and now on logistics/delivery/e-commerce - notice how abstract and detail free this article is.

But actually, that could be a great content authoring strategy: create a really good, smart name for yourself. write articles that the smart people of the internet like to read. give them the wrong answer, they will be forced to correct you. collect replies(including those from you your closed forum) and grab the right answer. publish and charge.

Don't know if that's the strategy. But it could work.

> give them the wrong answer, they will be forced to correct you.

Before Stack Overflow, the savvy knew that this was a great way to get help with anything Unix-related. If you just went to Usenet or mailing lists and asked "how do I X?", you'd be pummeled by uppity greybeards going "RTFM, n00b". Ah, but if you created a sockpuppet and posted the wrong answer in reply to your question, then you'd be guaranteed to get useful advice.

Yep, that's definitely Cunningham's Law [0] in action.

[0] https://en.wikipedia.org/wiki/Ward_Cunningham#Cunningham.27s...

Ha! I never knew it had a name. I thought I independently discovered this phenomenon, but I guess it's so obvious others realized it too.

Wow, this is genius. Makes so much sense.

" Ah, but if you created a sockpuppet and posted the wrong answer in reply to your question, then you'd be guaranteed to get useful advice."

I've gotten some great information on forums like this doing exactly that. I usually pose it as a challenging assertion or question rather than a statement of a fact I'm sure about. Gotta balance using that for it's benefit vs effects on one's own reliability.

Variation on that was to say something like, "Linux sucks! It won't do x." If the answers weren't sufficiently straight-forward, you could follow up with, "It's so much easier on MS!" Required a thick skin though.

It's really cherry picking and reverse engineering good outcomes at it's best. Sure Palm said this and that. What about all the thousands of others companies that paid no heed to some potential business threat and made out fine? The thing that appeared to maybe be a big threat that didn't ever happen? You can't protect and spend time on all possibilities. It's like trying to block all acts of terrorism. [1]

[1] Which is what happens after an event it then seems as if someone should have known in advance it could have happened the way that it did and taken steps to prevent it from happening the way that it did.

I distinctly remember the way Steve Jobs presented it: "It's a phone, it's an iPod, and it's an Internet communications device."

That just comes off to me as though even he didn't realize how ubiquitous phones would become, and how integral to our lives they would become, even replacing computers for so many tasks. It's my phone, my iPod, my work outside of office hours, schedules and times my workouts, my alarm clock, my camera, my GPS and mapbook, my library, my notebook, my flashlight, my scanner, my TV, the way I look at cat pictures and the way I look up where to find an AED in an emergency.

Read back the quote again to yourself.

Phone + iPod + Internet communications device...you are pretty much describing a general purpose computer, no?

And if you go back to the keynote, that was kind of the joke. He started out describing these as if they were separate products being introduced. The punch line was it was just one device. In other words, subtly mocking the idea you needed a separate device for each kind of functionality you wanted.

Lastly, Apple was a computer company from top to bottom. How else would they envision a solution to the problem of crappy phones, but to turn them into computers with really good software? That's just who Apple was.

So the idea of the Apple strategy being replacing the mobile phone with a mobile computer is only obvious in hind sight just isn't true, it was obvious as soon as the iPhone was introduced.

A general purpose computer, or a very locked down thin client. Like the world of AOL users.

I would have suggested the latter, otherwise the system would have been a lot more open.

General purpose computers give purchasers administrator/root access by default. If you don't have that, you have a thin client run by someone else.

"That just comes off to me as though even he didn't realize how ubiquitous phones would become, and how integral to our lives they would become, even replacing computers for so many tasks."

No one did, though. Go back and watch that keynote — there's thunderous applause for the iPod and phone portions of that sentence, and (relative) awkward silence for the "internet communicator" bit.

Reading through the Jobs biography, there's a great segment on the development of the iPod wherein Tony Fadell (I believe) details how he was coached to present the best mockups to Steve last.

Given the ordering of the three statements in that sentence you quoted, I have to imagine that someone at Apple — maybe not Steve — understood the eventual ubiquity of the smartphone.

Wasn't there a really big article about this?


The gist of it is the entire company wanted to do a phone because of how cool it would be. The work of three secret teams unbeknownst to Steve Jobs came together in just the right ways to make the iPhone.

"There will be web dialtone everywhere." – Steve Jobs, 1996


I found the school voucher idea to be the best part of that article. I didn't know the cost was so close for some of those public and private schools:

"I have a 17-year-old daughter who went to a private school for a few years before high school. This private school is the best school I've seen in my life. It was judged one of the 100 best schools in America. It was phenomenal. The tuition was $5,500 a year, which is a lot of money for most parents. But the teachers were paid less than public school teachers—so it's not about money at the teacher level. I asked the state treasurer that year what California pays on average to send kids to school, and I believe it was $4,400. While there are not many parents who could come up with $5,500 a year, there are many who could come up with $1,000 a year.

If we gave vouchers to parents for $4,400 a year, schools would be starting right and left. People would get out of college and say, "Let's start a school." You could have a track at Stanford within the MBA program on how to be the businessperson of a school. And that MBA would get together with somebody else, and they'd start schools. And you'd have these young, idealistic people starting schools, working for pennies."

> I asked the state treasurer that year what California pays on average to send kids to school, and I believe it was $4,400. While there are not many parents who could come up with $5,500 a year, there are many who could come up with $1,000 a year.

> If we gave vouchers to parents for $4,400 a year, schools would be starting right and left.

But if the public spends $4,400, that's not what goes to the school, since the total includes state and district oversight that is outside of the school.

This money would not be available to alternative schools unless you are suggesting that they should receive public funds without equal accountability.

Im just quoting it in case knowledgeable people show up with links to good info, proposals, or precedents. I haven't done research on it myself. I'm not suggesting anything since Im not informed enough on the topic. :)

Fine, assuming that the accountability overhead is 10% over the cost of actually doing the job and maintaining the facilities, so get a $4000 voucher.

The Mac is Steve's baby.

The i in iMac means internet.

Back in early 2000s, the prevailing silicon valley view is that the internet is and will forever be the greatest thing ever.

The iPod was already a success. Adding a phone to it is merely a formality, and an annoying one at that considering how much negotiations and concessions and leverage had to be applied to Verizon and Cingular before they settled.

But when writing trios, the best is always last.

An internet communications device, an iDevice, is really the evolution of the iMac.

The state of mind at Apple back then is that everyone absolutely knew the iPhone will be ubiquitous.

Personally I think that Jobs/Apple were expecting the Apple TV to be the next big thing. In the world of 2007, TV was still almighty and winning the living room should have been a revolution. It should have been the entry into the post-PC world, with iPad-like and maybe real TV devices following a few years later.

The iPhone was really mostly an iPod phone with internet thrown in like on the kindle.

This is obviously based 100% on gut feeling :-)

That list is pretty overwhelming when viewed all at once. None of it could have happened with the sub-3G speeds we used to have, though.

> the way I look up where to find an AED in an emergency.

Is that a specific thing, or would you just plan to map it?

It's a specific thing: PulsePoint has a couple of related apps. One will notify you (if you tell them you're CPR / AED certified) when there's a medical emergency you can help with in a public place near you (assuming the local emergency services work with them), and it'll tell you if there's an AED nearby. The other app lets you update their database of AEDs with GPS coordinates, photos, landmarks, etc. The latter is how I pass a lot of time in airports :)

Steve Jobs was absolutely dead set against 3rd party apps in the app store, which was one of the reasons the iPhone became such a huge phenomenon.



“partly because he felt his team did not have the bandwidth to figure out all the complexities that would be involved in policing third-party app developers.”

So the App Store model answered his objections. It's clear from the quotes in that article, he did want apps, but thought web development technologies would be sufficient. When that turned out not to be the case, looks like they turned to the App Store model so they could lock down the experience and keep the same degree of control as for web apps.

This was also the driving force behind the early jailbreaking efforts, because there wasn't a way to run ANY native code, be it Apple approved or otherwise.

Was in an audience when Mark Benioff discussed Steve Jobs coming to him to acquire AppStore.com as Mark owned it at the time.

Apparently Mark and Steve discussed this concept very early on.

Just did a whois history search on the domain and it changed hands from SFDC -> Apple on 03/19/2009

That's over 2 years after the iPhone announcement in jan 2007 at Macworld.. about 1 year before the AppStore launched

The phone was always one app among many, even before the platform was opened to third parties.

I agree. At the time of its initial release the iPhone "simply" was a combination of 3 main ingredients - as repeatedly stated during the presentation: Phone, iPod, Browser. And that was ambitious enough if you consider the flawless execution: While these features have been done before they always felt subpar: Music players as an afterthought and Opera Mini as a state of the art mobile browser.

The idea of apps emerged as an obvious next step and always felt like a demand by developers and users rather than something planned long ahead in Cupertino.

I am struggling to find that article in Vogue or some magazine that gave some inside details of how the iPhone came out. By my recollection, their strategy was definitely not to reduce the phone to an app. The strategy from what I remember was to add a desktop class browser to the phone and remove the limitations of a physical keyboard. Of course these had to be built on a new OS and all that. Oh there was one more - visual voicemail.

I think the point is fundamentally they were making something that could do way more than make phone calls, where so many other devices on the market were just a phone (or pager) with some extra functionality strapped on.

They started from a clean slate.

Funny that you didn't even mention RIM's Blackberry which was the dominant device for email on-the-go among busy execs, including, if I'm not mistaken Obama at that time.

Jobs was only looking for 10% of the global mobile phone market in the 2007 keynote announcing the iPhone. The iPhone was so well received it demolished Nokia and RIM's dominance in the space of a few years.

Actually BB is why I added the "(or pager)" to my comment.

BB was a two way pager (and later email) that they literally crammed audio in to (the first phone version REQUIRED a headset as it didn't have a speaker/mic).

They made the mistake everyone did. They made an X and so they tried to cram new things Y and Z into it because they were an X company. And it limited their vision of what they could do.

I'm still waiting on Visual Voicemail to come to my Carrier; i had it on a previous carrier and it was superbly useful.

From what I've heard/read it was basically a play against cannibalization of the mp3 player market by proto smartphones. There were two camps, Steve's which wanted a sleek new form factor without physical keyboard (touch input) rather than the alternate team's approach which was to shoehorn a phone onto an ipod.

I don't understand the distinction you are drawing. The original iPhone, with the addition of a usable web browser, was a general purpose computer in my mind.

Would you not call a ChromeBook a general purpose computer?

I realize there is a contingent on HN who doesn't consider a browser to be a valid virtual machine, and doesn't consider anything a real computer unless they have access to a command prompt. But that's a religious conviction, I don't see any principled difference.

Let me put it this way: the original iPhone, with only mobile Safari, was a more powerful machine in every way than every computer I owned up to probably 1999 or so, with the exception of screen size.

What am I missing?

The hardware is clearly that of a "general purpose computer". The OS and software was/is limited in ways that make me think more of an appliance, though. A complicated, customizable, but purposely limited appliance.

I can write a Javascript program, stick it on a web server, and load it in the phone's browser. Why can't I edit it on the phone and just run the program from disk? If I wanted to do so, how would I write a program to let me?

I think the core limitations for me are that I don't have direct access to the filesystem, I don't have access to the APIs required to program the device for things that I consider basic capabilities, and the vendor actively fights me being able to do those things. Any holes that they don't hit with OS and API limitations get caught during the app review when you're going to publish your code.

Compare to a Chromebook: There's a Google-provided dev mode that lets me install an OS of my choice, and a chroot environment developed by a Google engineer that gives me access to an unrestricted OS running simultaneously on the machine.

The command prompt's moot; one could theoretically write a GUI for anything and everything that it provides. The browser's "a valid virtual machine", but there are limitations of what I can do with JS, where I can host it to run the programs, and so on.

My 1st gen iPod Touch was cool after I jailbroke it (kind of a wifi-enabled single-board computer). I used to run sshd and rtorrent, and had Perl and a C compiler on it. I do similar things on my Android device running its stock OS.

I clearly presented my argument badly. I was not aiming for the general purpose computer part, but the suggested notion that the now clearly visible impact of the iPhone was intended just that way, as its main purpose, its this-is-all-or-nothing goal.

Steve Jobs and Apple in the 200x years clearly were amazing and did wonderful things, but they did not read the future, they went step by step and built on top of what they had. Some things fell into place, some things went different than expected. I can't prove it, but the shift in how we use our phones nowadays was as visible to Steve Jobs as it was to almost anyone in that field. There were predecessors. Multiple. The goal of the iPhone was to sell things, millions if possible. Everything else was in the range of: "Let's see".

It was definitely planned as a more personal computer. Bear in mind that the iPad was the original product, and they switched to a phone because the iPad would be too expensive to do at the time. So "phone" is the afterthought.

Do you have a source for that claim?

> I am just questioning if it was all planned to come out exactly this way from the beginning...

The article makes no such claim about the iPhone. The only evidence you cite is the app store. That does not refute that Apple intended on building, from the start, a "more personal computer" that "reduces the phone to an app" and "gains concessions from [carriers]". That's exactly what the original iPhone was and still is. It even launched with support for 3rd party (web) apps.

I agree, one would have to argue that concepts may have been there for some of the developers and engineers, but it's be recorded that Steve wasn't too fond of the idea. As with most products, it's been iterative improvements, that have made it the behemoth it is today.

If you exited over that you missed a great bit of analysis.

I tried to stick with it, but to me it was too obvious that he is trying to make an argument with mixing hindsight and intention.

It might make for a good analysis true, but trying to mask it as "Let me tell you about everyone's intentions, and on the next slide I show you the future" just puts me off.

A computer with apps from only one vendor is still a computer (unless you prescribe to the ideological notion that a "computer" must be "open" somehow).

To say nothing of the fact that from day one the iPhone had a best-in-class web browser. It's funny how the iPhone has been continually derided for not being open enough while having full access to the open web, the platform that is routinely held up as the future of computing by those who think local binaries are a relic of the past who also deride the iPhone for not being open enough.

Partially because it's lagging behind in "open-web technologies", although it's fair to question if all of those are really as necessary.

(Which wouldn't be an issue if you could install an alternative browser (edit: rendering engine) on it, but it's not open enough for that)

You're right. It's too bad we don't have all those great web browsers that existed on phones before. I remember how great PocketIE was. The Opera thing that sent down screenshots? The joys of WAP?

People complain about Safari but it was a REVOLUTION compared to what was available before.

I was strictly talking about the criticism now, not back then (I mean, Apple briefly tried to champion the web-only approach some of those critics now advocate and see Safari as in the way of. We'll see if it has more success this time)

Look man, you can have any browser you want, as long as its Safari.

This deal is Venmo for groceries. A tech company is filling an obvious hole in an established industry - where in many other countries, the incumbents have been flexible enough to do it themselves.

Amazon has reached the conclusion British supermarkets reached over a decade ago: The sensible way to run a perishable goods supply chain is through an existing network of grocery stores.

While WebVan was crashing and burning, Tesco (biggest supermarket) was building out a home delivery service using their shops as "free" fulfilment centres. Tesco Direct launched so early, their original client wasn't even a website - it was a desktop Windows application!

This is another odd market where somehow the US incumbents seem unable to implement an obvious feature that is common in other countries. Eventually a middleman (Instacart for food, Venmo for money transfer) do it for them - and take a cut on the way. I wonder why this happens?

> This is another odd market where somehow the US incumbents seem unable to implement an obvious feature that is common in other countries.

What other examples are there?

We have had online grocery delivery in Canada called Grocery Gateway [1] for a decade or longer. It's run by our biggest grocery company as well. The business seems to be doing fine but it's still very much niche and most people don't know a lot about it.

I assumed the US had a few local companies doing the same there. The difference is that these are not anywhere near the scale of what Amazon could do with groceries at a national country wide scale.

Overall I don't think Tesco vs Amazon is a really accurate comparison. It sounds more like all of those people who said "Nokia had smartphones way before Apple! This isn't new." Or comparing a taxi company having a mobile app to book a ride vs Uber. Or Couchsurfing vs Airbnb. Etc, etc. There are tons of examples.

There is still a huge amount of potential to replace grocery stores with ecommerce and so far only minor attempts have been made.

[1] https://www.grocerygateway.com/store/

IMHO the three biggest challenges have been cost, accuracy, and quality.

We have a lot of options for this in NYC: PeaPod by Stop N' Shop, Instacart, Fresh Direct, and Whole Foods has its own delivery service.

I've tried them all. I REALLY want to stop going to the grocery store. But with all of them, I've found:

* The price is too high to justify it. Walking down the corner to the grocery store is not that big a deal.

* You don't always get everything you ask for. They run out of stock of things, or forget things. So you often end up missing that one crucial ingredient.

* Sometimes they don't give you the quality you're looking for. Maybe the bananas or avocados are either unripe or too ripe. A couple services ask you the preferred ripeness, but then they don't end up having that so they give you a different ripeness.

* Back on the pricing: I personally tend to buy only 2-3 days of food at a time, because I want it to be fresh. If I want baked cod on Friday night, I generally want to buy it Thursday or Friday. But since I'm paying a premium for my delivery, it makes more sense to batch it into bigger orders, which doesn't work for me.

You live in NYC, and you still want delivery?

I mean, when I lived in Manhattan, the nearest grocery store wasn't much further than my refrigerator. Walking into the store and grabbing the thing you want would have been a lot faster than ordering online and waiting for someone to deliver.

I lived close to a grocery store downtown, walking distance, but I still used an online grocery service for a significant portion of my groceries. I'd get a monthly delivery of non perishable or large/heavy items such as toilet paper, paper towels, canned food, milk bags (this is Canada), frozen food, juice, pasta, etc, etc.

Having to carry large/heavy items without a car was annoying so the service was a great help.

Basically I'd walk to the store for fresh meat, deli meat, cheese and fruit/vegetables. Everything else I'd just order online. The great part of the online store is I could reorder the same list without having to pick each one out again, with some small modifications usually, so it saved a bunch of time.

It was also cheaper on a frequent basis when using the online service. The grocery stores downtown usually have a premium price on everything.

Depends how much you value your time. If you want to shop at whole foods or trader joes then you're going to stand in line. Standing in line is one of my least favorite things to do. I would rather pay a delivery fee and have someone else do my shopping and bring it to me while I spend the timing doing something else that I enjoy.

Yeah, I was thinking more the little super markets and bodegas on every corner, than whole foods or trader joe's.

Come on that's nowhere near grocery-store quality if they even have fresh foods of any kind.

> What other examples are there?

Off the top of my head, so all my counterexamples will be British, but here are a few US tech companies who built missing features for major players - whereas in other countries the major incumbents just built it themselves:

* Venmo - real-time money transfer (the UK equivalent is the bank-run Faster Payments Scheme)

* Intuit - Self-service tax filing (most UK tax is done for you. If you do have to file, you fill out a straightforward online form ["I earned £___ this year"] and enter your card details. Job done.)

* Instacart - Online grocery shopping (a standard service from incumbent supermarkets, eg Tesco/Sainsbury/Waitrose/...)

* Apple Pay - Electronic payments more secure than magnetic stripe (most of the world had chip cards ~10yrs ago; the French had them in the 1980s.)

I'd say that's a pattern. It's not that the US economy isn't dynamic, it's just...weird.

UK is the exception. Looking globally, one gets the impression that grocery delivery is a niche, with pickup being the more popular option.

In The Netherlands, the majority of the premium grocery stores have started a home-delivery option, probably in no small way helped by the density over here.

>This is another odd market where somehow the US incumbents seem unable to implement an obvious feature that is common in other countries.

This is not quite correct. Grocery stores have had delivery in large cities for a while. There's even a pretty famous scene in the movie Magnolia (1999) that revolves around it. The US is big, and has a lot of empty space, and has many different markets that are tough to coordinate. It's no surprise then than it has taken a big player like Amazon to offer wall-to-wall coverage of grocery delivery. It still hasn't been proven to be something a sufficient number of Americans even want.

>I wonder why this happens?

The UK is a much smaller market, so it could be easier to roll out new products?

I'm don't think "smaller" matters much in this case. Delivering food from nearby shops is inherently local and can be rolled out slowly.

"Denser", however, is often a very good answer. Imagine shoving every person from California into New York state, alongside the existing residents - and then shrinking the state a little. That's the size and density of England.

tl;dr: Amazon will "transform the Whole Foods supply chain into a service architecture based on primitives," using the stores as customers that guarantee a necessary level of scale.

I think this is basically true. Amazon will refactor Whole Foods' supply chain into something that serves both Amazon Fresh and the retail stores. This will allow Fresh to scale massively and dominate the category.

I would just emphasize that Amazon probably has more interest in Whole Foods then merely a guaranteed "customer". Whole Foods is bringing a lot of local supply chain relationships to the table.

But I agree 100% it's about the supply chain and not transforming retail stores into delivery hubs or anything silly like that. It won't hurt that the stores will freely promote Amazon Fresh and serve as great venues for sampling new products. But Amazon's goal here is to take over the category by massively scaling Fresh.

Geeze, I hope it compiles okay.

This whole article seems predicated on a couple of misconceptions. To whit: 1) getting bought out is losing 2) buying is winning.

The fact that Amazon has more money than Whole Foods is nothing to do with its performance in the groceries market. The fact that Amazon has bought Whole Foods suggests, in fact, that they think there's value to be had there that isn't reflected in their current market capitalisation.

And none of this actually means that the Whole Foods acquisition will work. It sounds like a good idea, and I wouldn't risk money betting against Jeff Bezos but Amazon aren't a dominant player in the groceries market yet. They've just bought one of the few successful insurgents.

In short, this is just a move in the game. Calling the match is pretty premature.

I don't see what in the article gave you the impression that Whole Foods lost. Or that Amazon won.

He drew an analogy between the Whole Foods CEO about Amazon and the groceries market and the Palm CEO about Apple and the phone market.

> The fact that Amazon has bought Whole Foods suggests, in fact, that they think there's value to be had there that isn't reflected in their current market capitalisation.

Not necessarily. They might be thinking that WF's value can be increased by integrating it with Amazon, making its market cap smaller than its actual value conditional on the acquisition.

True, but that's typically a big bet. Very few acquisitions go smoothly. Put it this way, imagine the same article being written about the Tumblr acquisition. It's unclear why he believes Whole Foods constituted an obstacle to Amazon, either.

As I say, I'm not betting against Bezos. But not every move he makes is genius, and it's currently too early to call. And groceries are _hard_: history is littered with successful firms and their disastrous expansions.

Heck, I remember reading articles about why Apple's new "physical stores" strategy was doomed to failure.

I think you missed the point of the article, which was that buying Whole Foods had nothing to do with winning or losing, but with providing a first, best customer for their grocery sales efforts.

I generally like Stratechery articles - but I found this one stretching to make everything fit into a nice, neat box - particularly the comparison of AWS to the Grocery business. I found the graphics representing rather incomprehensible (perhaps I'm simply not bright enough to get them) and it feels like someone looking for a grand unifying theory of Amazon, as opposed to something simple: the grocery business is huge ($900b/year), Amazon wants to be a part of that channel.

It's basically doing the opposite of Walmart. If there is a grand unifying theory of Amazon, I think it is also pretty simple: start with a specific niche in a category, dominate that and then grow. But that is hardly a unique approach in business. Amazon does, however, execute it extremely well.

I found this one stretching to make everything fit into a nice, neat box.

Yes. It's entirely possible that Amazon found Whole Foods for sale at such a low price that the stores don't matter much. It's the house brands which Amazon can distribute through their existing supply chain. Most of Whole Foods' house brands have reasonable shelf lives, so they can easily be slotted into the Amazon system.

Using Amazon's package distribution system to fill the stores would be inefficient. Stores get product by the pallet, not the box. Supermarket distribution is semitrailers of pallets going direct from big refrigerated warehouses to big stores. Amazon isn't set up for that.

It's not clear what Amazon will do with the stores. Probably put in Amazon lockers and cut the prices on the house brands at first.

Totally agree that all WF's white label products greatly increases the products available via Amazon Pantry.

I would also say that Amazon is well positioned to send mixed-product pallets. Stores today get entire pallets of the same product. This is incredibly inefficient. Amazon is well positioned to use fulfillment centers and robots to create entire mixed product pallets of only the products sold on any given day at a store.

This can greatly reduce the amount of unsold product in the supermarket supply chain to only the products sold on any given day. This also lets use less shelf space for a product and store a greater variety of products in the same space.

Customers may also have the option of having specialty products sent to their nearest Whole Foods for when they do their next grocery trip. For example, my favorite honey, Tasmanian Leatherwood honey is only available on Amazon. I could choose to have that available at my local Whole Foods the following day. In fact, Amazon could detect that I have it in my cart, send it to the store proactively and notify me that I can have it at my local whole foods, prompting me to go in and buy more than planned.

That's exactly how it works. They've already cost-cutted and streamlined it all about as much as it will get. Yet, it all reduces down to people pulling pallets from manufacturers off trucks, putting them somewhere, other people making custom pallets for stores from those pallets, shipping them, and then breaking those down at store for sale to customers. The only time I saw real innovation in that was Costco and its predecessor's brilliant decision to eliminate the store to just let people shop off the pallets and such in the warehouse.

I think Amazon did try some stuff like having storage containers robotically come toward workers. They might have some tricks up their sleeves. The fundamental model is pretty basic and optimized, though, with competition driving margins down making it harder to optimize over time.

I think Amazon did try some stuff like having storage containers robotically come toward workers.

They bought that technology from Kiva, then bought the company, and all their current-generation distribution centers use it. (Amazon has distribution center generations; they don't upgrade existing buildings while in operation, but build new ones with newer technology. Some pre-Kiva centers are still running.)

Automatic picking is still an R&D problem. Amazon has put some effort into that; they have an annual competition where people can compete to win a few thousand dollars solving a billion dollar problem.

Automatic mixed-case picking and palletizing can be automated. It isn't quite fast enough, quite general enough, and quite cheap enough to take over.[1] Except inside frozen food warehouses.[2] Humans are not very productive at -23C, but special-purpose automation machinery is. Amazon has the scale to push that gear into being cost-effective.

[1] https://www.youtube.com/watch?v=XO7fvrdTCgs [2] https://www.youtube.com/watch?v=LuvrP1QKsdk

They're really neat tech videos. The first confirms how slow and expensive I'd expect it to be. The second looks really complicated compared to humans picking off pallets with PITS guiding other pallets or rollers. First vid I found on YouTube:


So, they have pallets all over the place. The order pickers are driving something automated to grab the stuff. They get a good amount at once. They can stack it on new pallets or ship it as shown. Pretty straight-forward. I could see some addition speeding up dealing with the vertical pallets. Those self-driving forklifts from show Silicon Valley could be nice. Yet, the automated versions are so expensive that some (all?) DC's in both Walmart and Kroger got rid of the pallet-wrapping machines per conversations with employees. It wasn't cost effective to maintain vs just having low-paid people do it they said.

So, it will be interesting to see what happens with all this. I like the idea of doing frozen first as people might be moving slower. I do also recall those having boxes that are more similar to each other compared to other loads. I'd start with pallet or roller optimization software that tells people to get the stuff in the order that stacks or ships easiest. That might exist already. I only have second-hand info from the DC's so don't know. Maybe have robots that move things automatically dropping pallets or adding them as in the video with the humans order picking off them. Maybe some lights or something on the dolly to show where to drop the next box for ideal stacking. There's possibilities..

They weren't comparing AWS to the grocery business. They were comparing the Amazon Book Store to the grocery business and positing that, like AWS, Amazon would use the grocery business to prove their just-in-time supply-chain-as-a-service product.

He wasn't explicitly comparing the two, but saying that the three (books, AWS and grocery) were all part of the same formula/strategy.

"the grocery business is huge ($900b/year), Amazon wants to be a part of that channel."

I think that fits with his analysis. His take on Amazon is they want a small piece of every single economic transaction, literally. So he is predicting Amazon's end goal isn't to make Whole Foods into the dominant grocery chain. But to insinuate themselves so they have a cut of every grocery transaction, whether it's Whole Foods or any other grocery store making the final sale to the customer.

The grocery business has brutally low margins (~1%). Bezos famously said "your margin is my opportunity", but there's no margin == no opportunity... at least as far as the traditional grocery business goes.

There's some far deeper strategy going on here, lots of possibilities.

I think they're going for grocery logistics, not groceries themselves.

There's still plenty of margin for guaranteeing timely, correct delivery of perishable goods.

Think distributors, not retail stores. At least that's what I got from the article

Agreed - definitely something bigger going on - although Whole Foods has net margins in the 3% range.

Oh I agree - but I think his attempts to link that idea back to a grand unifying theory of Amazon, for me, just seemed forced.

Thompson is good, but he gets blinders on for some things. Last year, Uber was going to be blowing our noses for us.

I'd love to hear a devils advocate POV about Amazon from Ben.

With your hints Thompson can come up with a strategy of a Global Nose Blowing Service. It can explain with neat diagrams how previous attempts failed and why in mid to long term most people on planet are going to use a modularized GNBS instead of doing it in old fashioned way.

I think this article hits the nail on the head. When I use a grocery app, I'm usually shopping at Whole Foods. By modifying the structure so that delivery is a feature (Much like Prime), instead of a separate good, they'll be able to take most of the current players in the space to the cleaners.

> The cost to build AWS was justified because the first and best customer is Amazon’s e-commerce business

I've seen statements that AWS wasn't actually used for Amazon own needs (in the beginning) and was developed by a different team at least a couple of times, here on HN and elsewhere. Were these rumors false?

History is always a little murkier than stated – it probably started out as something and then found its 'product/market fit' serving Amazon.

Steve Yegge's famous essay covers it quite well, I think: https://plus.google.com/+RipRowan/posts/eVeouesvaVX

> So one day Jeff Bezos issued a mandate. He's doing that all the time, of course, and people scramble like ants being pounded with a rubber mallet whenever it happens. But on one occasion -- back around 2002 I think, plus or minus a year -- he issued a mandate that was so out there, so huge and eye-bulgingly ponderous, that it made all of his other mandates look like unsolicited peer bonuses.

His Big Mandate went something along these lines:

  1) All teams will henceforth expose their data and functionality through service interfaces.

  2) Teams must communicate with each other through these interfaces.

  3) There will be no other form of interprocess communication allowed:  no direct linking, no direct reads of another team's data store, no shared-memory model, no back-doors whatsoever.  The only communication allowed is via service interface calls over the network.

  4) It doesn't matter what technology they use.  HTTP, Corba, Pubsub, custom protocols -- doesn't matter.  Bezos doesn't care.

  5) All service interfaces, without exception, must be designed from the ground up to be externalizable.  That is to say, the team must plan and design to be able to expose the interface to developers in the outside world.  No exceptions.

  6) Anyone who doesn't do this will be fired.

  7) Thank you; have a nice day!

You missed the best part, where he says something along the lines of 'haha, just kidding about #7. Jeff Bezos definitely does not care about your day'.

Thanks for sharing that. It had some nice gems in it.

AWS was not built for internal-first use. For quite some time (often years) after S3/EC2/etc launched, they were not available internally.

From what I hear, they are now used pretty extensively internally (and that may have been an end goal)

I've seen similar statements.

Btw, even without a perliminary big customer at start, when it started, AWS offered a phenomenal and unique capability for developers, that price didn't matter(to an extent), and hence scale probably didn't matter much.

So i'm not sure if this "big customer" theory is correct.

I'm not from the US so i have never been to whole foods, but isn't the point of whole foods is that you get a totally different selection than other retailers ? and if so, how could it be a customer for regular grocery products ?

Many of the brands that are dominant in Whole Foods are also available, to some degree or another, in other grocery stores.

But Whole Foods has a more targeted sourcing strategy for fresh veggies, fruit and meats. They carry more of those organic brands than the usual grocery store.

It's not entirely a "whole 'nother thing" but it is a pretty deep curation of food, food types, brands, and some of the more eclectic, local and smaller food brands.

You get used to Whole Foods but when I visit a Safeway or Raley's, I'm pretty shocked what a more traditional, mainstream grocery store looks like.

That said, and on an unrelated point... though I buy more from Whole Foods than any other grocery chain, the degree to which all the stores have started to be more like each other and less local-driven has been a little disappointing. The Whole Foods in Flagstaff is much the same as the Whole Foods in Roseville.

The local-produce movement has some surprisingly harsh downsides, and I'd guess that Whole Foods reassessed its strategy after battling them for years.

The first big problem is that the macroeconomics of local tend to suck. The opportunity cost associated with keeping supply constrained in a small market grows in proportion with our ability to reduce transportation costs and integrate national and international markets. Meanwhile, a fragmented patchwork of smaller farms and companies serving local markets tends toward waste: farms invest in the high-end produce they think will maximize yield per acre in the local market. Unit costs per item are higher, and nobody has incentives to produce the essential, but low-margin items that the market still needs.

The second big problem is that consumers don't actually know what local means. Not really, at any rate. Enjoying blueberries in Southern California in the summer? Those aren't local. Oh, but we still want them? And we want them year round? So...I guess we're back at some combination of local + (inter)national, in which case we're still supporting factory farms, giant monocultures, and seed conglomerates -- we've just conveniently pushed these things out of our minds. But they still exist, and we haven't really solved the problems we've set out to fix. The point is, consumers want their goddamn blueberries. And their bananas, which, don't even get me started how much actual blood has been spilled over the international banana trade. (Seriously. Look up Chiquita on Wikipedia. You'll be tempted to use the phrase 'conflict bananas' after reading up. Don't use that phrase. But, like, still. Wow.)

There is probably some sort of ideal market, combining the economic efficiencies of concentration with the variety and quality of specialization. I think we'd all love to get there. But this is the kind of problem Jeff Bezos seems precisely better equipped to solve than John Mackey did.

> I'm not from the US so i have never been to whole foods, but isn't the point of whole foods is that you get a totally different selection than other retailers ?

In that WF's selection is limited based on certain criteria linked to virtue signalling in some portion of the market; much of the grocery selection at WF is also available at other chain retailers, but much of what is available at other chain retailers is categorically excluded by WF.

> and if so, how could it be a customer for regular grocery products ?

Maybe Amazon is interested in targeting the same market segment as WF, and plans on tailoring their grocery operations appropriately.

It isn't totally different - I can still purchase all my staples, meat, and whatnot there. They do appeal more to more niche dietary habits, such as vegetarian/vegan/no gluten.

Health food is their focus and a lot of the difference in practice is what they don't carry. There is a spectrum of stores between stores that don't carry any organic food and local cooperatives that tend to be exclusively organic, whole grains, etc. food (there used to be more independent health food stores but Whole Foods bought them or drove them out of business). Whole Foods is a mass market compromise that was able to dramatically lower prices on some key items vs. cooperatives and has about half organic food (and for the other half tries to find at least somewhat unusual products with some, often dubious, claim to being healthier in some way or higher quality or more environmentally friendly). One thing Whole Foods does that is unusual (other than cooperatives) is to label where the perishable food comes from. I don't know if they actually source more locally than others, but because it is labeled it is possible to tell if a particular item is local or not.

These days many large chains have organics/natural foods sections; they might have 10-20% organic products, but being much larger there isn't that much that is only found at Whole Foods. Those other stores often have organic sections because people who buy organic food often are not interested in most of what those stores sell. Some chains focus on cost and have very little organic food; the lowest cost organic food is frequently double the price.

Long winded response, but none of what Whole Foods sells is really unique. Also, Whole Foods has limited reach and there may be quite a bit of demand for health food stores that is not currently met, particularly outside large cities. The problem with independent stores is the cost; there are a lot more people who will pay 2x for healthier food than people who will pay 4x or 8x, and enough items need to be 2x to make the store a regular stop. For all the reputation Whole Foods has as being expensive (and it is for many things), part of the key to their success is relatively low cost key health food items. Cooperatives have been able to lower their prices but often still have areas where everything they sell is quite a bit more expensive.

The article has a lot of obvious inaccuracies but might be on to something with the overall theory. If so, hopefully it wouldn't just be regular grocery products but also health foods. It makes more sense than anything else I've heard that Amazon might want to expand the distribution operation while leaving the stores themselves relatively untouched.

> Health food is their focus

More accurately, health and environmental fashion trend / virtue signalling food is their focus; actual health is very much beside the point.

While a lot of what they sell is as you say, I do think they have a core health food focus. Many actual health food items were not nearly as widely available before Whole Foods (and still aren't in many areas).

Whole foods is more about "organic" foods and brands than necessarily a different selection.

Which makes it even a better (i.e. more hard-to-please) customer from Amazon's perspective.

I was surprised Stratechery didn't talk more about the warehousing side of the grocery industry, which is a huge business unto itself.

Given the huge scale of Amazon's fulfillment centers, those are the businesses I would think should be the most afraid of this news.

My biggest take away is that Amazon will use these stores to take on the restaurant supply business. I'm willing to bet the CEO of Gordon Food Service, which covers eastern US, is clueless on how his company's fortunes are imperiled.

I'm wondering if Amazon will attempt to buy something like Costco or Bestbuy. Then they'd have both ends of the market and also have the ability to connect the big box to the relatively smaller box.

interesting quote regarding amazon's updated mission statement to be earth's most consumer-centric company where folks can discover things to buy. I wonder what buying a house/home on Amazon would be like.

IMO, any blog or news article that was written AFTER Amazon has been in the news are simply trying to ride the press wave.

I especially find it tickling when most of the authors have no prior work experience or affiliation with any business of Amazon.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact