What's Amazon's market share? 35% or 5%? | Hacker News Reader

What's Amazon's market share? 35% or 5%?

ben-evans.com

186 pointsmooreds93 comments

I think most people recognize and understand that amazon neither has a relevant market share nor a monopoly in the traditional sense of the word. The author makes a good point that it is very much political as much economic.

The latter is important because the traditional understanding of a monopoly is somewhat antiquated with tech and has led to many people arguing to change its definition.

The most prominent example is Lina Kahn's wildy successful/read paper in the Yale Law Journal about amazon's antitrust paradox and how a company can operate in monopolistic ways without having a significant market share.

https://www.yalelawjournal.org/note/amazons-antitrust-parado...

A similar case is Apple and Spotify competing to grow their music streaming service.

It will be interesting to see how the current political framework is adjusted for this new economy dominated by big tech.

The modern view is the traditional view in the back to normal sense.

In the US Robert Bork's book "The Antitrust Paradox" started a period of time where the antitrust law was only about benefits to consumers and overall efficiency. The traditional and the modern viewpoint is that the role of antitrust laws plays role in controlling economic power in the public interests. Vertical agreements and price discrimination can be prohibited.

How monopolistic a company is its demand elasticity.

I've been reading The Economist's Hour by Binyamin Applebaum[1], which presents the dramatic changes of antitrust interpretation in some depth. Recommended read btw.

Bork credited Aaron Director, one of the first generation of the Chicago School and Friedman's brother in law, as his main influence. So making the consumer and efficiency the focus can be considered an early casualty of the wave of neoliberal deregulation we've been living with for decades.

Of course what is not considered is impact on workers, innovation, political power, market etc. Essentially it's hollowed out to benefit large business.

[1] https://www.economist.com/books-and-arts/2019/08/31/when-eco...

You might enjoy this podcast interview with Applebaum:

https://www.econtalk.org/binyamin-appelbaum-on-the-economist...

Thanks for that, a marvellous listen!

Really refreshing to hear two knowledgeable people who disagree on so many of the conclusions having such an engaging and positive conversation. Then find shared common ground so easily with both good at staying on point. Nice to hear he had a high opinion of the book too -- outside those conclusions drawn.

I'd have quite liked them to keep going... :)

>amazon's antitrust paradox

Haven't read the paper but isn't it very similar if not the same as Walmart's monopolistic behavior with suppliers, which has been going on before Amazon became big?

the easiest way to conceptualize monopoly is "the power to limit the possibilities and decisions of other companies" (other than via transparent buy/sell decisions in the marketplace of course).

the size of a company relative to some industry is an indicator of monopoly power, but not a necessary nor sufficient condition of it.

the idealization that markets only care about price signals (and particularly tries to ignore political concerns) doesn't represent any real market. antitrust (and similar laws) try to reconcile this discrepancy so that markets can function more closely to their idealized form in inherently political environments.

> the power to limit the possibilities and decisions of other companies

This is a very naive view of the world. In many non-consumer industries, companies limit the possibilities and decisions of other companies via contracts as a matter of business every day.

It seems fashionable these days to redefine terms to suit an agenda. I think that this should be pushed back against.

The term 'monopoly' is clear and well-defined.

If Amazon is not a monopoly "in the traditional sense of the word" then the correct term should be used to qualify what Amazon is instead of trying to redefine the word in order to be able to label Amazon with it.

Indeed. And what is weirder is that there seems to be an attitude that if it's not a monopoly then it must be fine and should be allowed.

Monopoly is used as a generic term to discuss everything under antitrust and competition law with layman, because people don't know what monopoly power, market power, barriers to entry or exit, restraint of trade, price discrimination, market failures etc. mean.

Everyone knows what pure monopoly is. This creates problems if people take it too literally.

Yeah, that's the whole problem. "Monopoly" has become a generic and unspecific term, but it's still very politically significant. So laypeople cry for X company to be regulated "because they're a monopoly" without having to have any sort of actual objective backing.

The point OP is making, and you seem to agree with him, is that we shouldn't overload the terms. "Monopoly" is a very specific enconomic term, a political term, and a subjective layterm used to portray large companies negatively. Bundling all those meanings together creates miscommunications.

My point is that just because word "monopoly" is used as a shorthand for more accurate term "monopoly power" it don't make calls for regulation any less important.

I don't think that's the definition of a monopoly that is problematic here, more the definition of a market, which is a more abstract group, composed of competitors and that companies are supposed to be able to enter somehow easily.

The debate there is whether online shops and physical shops are in the same market, and what geographical divisions make sense for online shops. It is also in the definition of tech markets, which are starting to clear up. We know for instance that it is not necessarily easy for a tech giant in ads/search (Google) to enter the social media arena, so these two things are separate. Hosting and cloud computing seem to be very similar markets, most players there having offers in both. I think the market of online banking is being defined right now with many unconventional players showing that ecommerce may actually belong to the same market as paypal.

Defining monopolies is much easier, if you start from the (rather political) use of the concept : deciding when a company has to be broken up to preserve competition. MArkets are just a tool to create definitions of monopolies. That debate is too often obscured by various ideologies on the wisdom of the crowds and invisible hands imbued with various mystical properties.

I often find people first assume something that isn't in US anti-trust law - that one must have a "monopoly" in order to be problematic - and then use a narrow definition of monopoly to conclude that no legal problem exists.

US law talks a lot about "market power", not monopoly. And...

> economy dominated by big tech

we both seem to think it is clear that there is market power.

Just looking at the numbers in the title, I assumed this article was going to be about AWS.

The numbers are roughly 35% of commercial cloud market share and something like 5% of all websites. (The actual numbers are 1-3% less)

But in light of the discussion here, I only see this as a good thing.

Amazon puts heavy pressure on the competition in the cloud market (and retail). Competitors like Microsoft and Google very quickly develop their own services to match those of Amazon, serverless being a great example. Or, take Walmart and Target’s foray into internet retail.

Amazon is a very interesting company. They constantly bring about new innovations and the market validates them often. And, competitors adapt to match their offerings.

I see Amazon as a company with quite a lot of growth potential left, but that’s not to say it will remain in the position it’s in. Just observing cloud market share, Microsoft is heating up, and taking a lot of those customers. As a consumer of both these products, cloud and internet retail, I respect what Amazon is doing. Even though I’m not a Prime member and only really buy books on Amazon, I constantly see their impact in pressuring other markets, and I benefit as a consumer.

> 5% of all websites (The actual numbers are 1-3% less)

Source?

It's not even clear whether this means

- 5% of servers running

- 5% of registered domains

- 5% of user traffic

Sorry, that was poorly presented on my part. I believe the most recent estimate was around 2-3% of all internet traffic was served on a web server being hosted by Amazon.

I can’t find the original source, but this Quora post[0] and included primary source[1] suggest the same.

[0] https://www.quora.com/What-is-the-estimated-percent-of-inter...

[1] https://www.netcraft.com/internet-data-mining/million-busies...

I'd be curious what type of insight they're getting from markets such as China, India, Russia, South America and SEA. I couldn't find much on their site. Their estimate could still be relevant for their target market, while being completely off globally.

Also, only looking at public internet traffic is a mistake since this isn't solely where AWS is trying to compete. You need to consider _all_ hosting, including government, intranet and other private players, which AWS does attempt to capture.

This is a really good point and really reinforces the take that Amazon isn’t as big as many of us believe.

Agreed. hard to be upset that amazon has improved my quality of life (+ millions of others') just because it improved bezos' more...

I spent a decade working in American Malls prior to the turn of the century and think Amazon has done a brilliant job of flipping the playbook and the playground of retail the past 20 years.

Instead of competing with 800 lb gorillas, Amazon turned many of the prior juggernauts into 2-tons of turd by making them move. Seeing anchor stores shutter and go bankrupt was a good indication of how slow businesses move, even post dot-com era.

In the "monkey see, monkey do" modern world (of commerce [too]), they keep on implementing things the monkey can't do (<48 hr shipping, airline fleets, AWS, space, etc) and provide the new curve to aim ahead of in commerce (analog, digital and hybrid) for the foreseeable future.

Amazon has to thoroughly annihilate a larger piece of the market(s) before they can be called a monopoly (its slash and burn farming in a way). We'll see how much longer they can keep the champagne flowing.

The 5% figure is out of _total US retail_, not online retail.

Just the fact that Amazon is now measured up against all retail is kind of frightening.

And yet it still pales in comparison to Walmart.

It doesn't pale. It's like a quarter or WalMart in terms of US sales, if I'm not mistaken.

It's not quite _all_ retail though. The author excluded market segments that Amazon doesn't participate in - cars, gasoline, etc

Retailers of that size - vs domestic retail share - are common globally among developed nations because the economy of scale is required to compete.

The alternative in a global economy would be to shut down access to your market to keep the foreign retail giants out. Otherwise they would come in and decimate your smaller, less competitive retailers.

It's why just two grocers (E.Leclerc, Carrefour) in France have 40% of their grocery market. Throw in three other retail companies like Auchan Holding and you have total retail dominance by just a few players.

It's why Schwarz Group and Aldi are so large in Germany and why Tesco has total sales equal to ~12% of all retail spending in the UK (domestically they're larger as a % than Amazon is in the US for example). If you scaled Tesco up, UK economy scaled to the US economy, they'd be larger than Walmart.

Interesting, thanks. I think market structure differences matter. As I understands it in the US Walmart can have an effective monopoly in an area by being the only big supermarket within reach. The area might only have a population that can support one big store and Walmart get to be it. That’s much harder to pull off in the UK due to our denser population.

This pits Amazon more directly against Walmart in particular in the US, because the package delivery model lets Amazon address the same markets as those big regional Walmart stores without having to build a big regional store. In the UK there would be very, very few areas where the main choices are Tesco and Amazon, but there are a lot of people in the US who’s main choices are Amazon or Walmart. Does that seem right?

Why’s that frightening? Over the last 20 years they’ve disrupted a bunch of incumbents through better customer experience and ruthless execution. I think Amazon’s lack of clear moral position on key issues is concerning (and potentially frightening) and I’d actually argue that Wal-Mart has been much more clear on where they stand on things, but I’m curious if you simply think a company disrupting a major industry is scary beyond simply change happening “quickly”.

Why is this frightening?

It seems to me only shows that online shopping is just a fraction of US's retail market, with the largest player only takes up 5%.

This places them a distant second to Walmart's 9% share of all consumer retail sales.

Is 5% a distant second from 9%? I guess but - what are the growth curves of these two businesses?

Is Amazon a monopoly? Does Amazon indulge in monopolistic practices? Rather than dance around an issue, state outright the real question that is being addressed or being glossed over.

Market share of what exactly? and is 5 or 35 even important as figures. For example the Walmart figure - does that include overseas assets. Walmart own Asda in the UK for example. Amazon act in mysterious ways in quite a few supply chains that say Walmart could not even dream of. I don't think the examples, figures etc are rigorous or useful.

So, what question is the blog really answering?

Not every article is even posing a question, let alone answering it. You could argue that's what a good article does. But some articles are just a series of things the writer finds interesting.

>Splitting out GMV is important because Amazon isn’t setting the price or choosing the selection for the third party Marketplace

Well. In some cases, Amazon will drop a third party's price for them and reimburse the difference - in other cases Amazon will simply unpublish the buybox/add to cart buttons.

Anyway, IMO the relevant market definition is online orders with 2 day shipping. I believe Amazon has well over 50% market share there.

This is relevant because if you're ordering something to receive within two days, buying from another site that only promises to get it to you within 4 days may not be an option for you.

That seems too narrow a definition for me. I order things expecting two day Prime shipping all the time but almost none are so urgent that waiting isn’t an option. For me it is simply a competitive advantage against other slower delivery options but also physical stores, because that requires the inconvenience and time expenditure of driving and walking through the store. That is, I feel the market they’re in (ignoring AWS etc) is general retailing.

There's a substantial portion of their customer base that would not consider other sites that took longer to ship. I'm not claiming everyone is like that but it's a relevant market.

There’s a substantial portion of Ford’s customer base who would never consider buying a Chevy (and vice versa).

Nevertheless, we consider GM as a competitor to Ford when determining whether or not Ford is a monopoly.

Those do roughly the same thing. I don't believe a subscription service that gives you no marginal cost 1-2 day shipping is comparable to anything offered by retailers without a two day shipping option.

That’s a position I understand but do not hold. I regularly cross-shop Amazon with EBay, NewEgg, B&H, random online shops I never heard of until Google found them, and even AliExpress.

With most things I order on Monday, whether they arrive Wednesday or Friday makes no practical difference to me. I either want it Monday or kind of don’t care typically.

Data point: do you pay for prime?

Yes. Even with Prime, I frequently buy away from Amazon.

I fail to see how third pary sellers, of which Amazon take a cut, are significantly different that suppliers if traditional retail outlets. The primary difference is less risk for Amazon, but otherwise they seem to have a very similar relationship of traditional retail <-> supplier. Not the same, no, but similar enough not to split hairs on the topic when discussing Amazon's market influence.

Agreed. Recently Amazon stopped showing addresses to sellers for FBA orders. If you don't even know who your customer is, it's hard to argue you are the retailer.

Emails and phones have been gone a while. Essentially all you get is a generic amazon hosted forwarding email address if you want to talk to your customers.

Yeah, it's the physical addresses that were removed recently.

Phones were removed about two years ago if I recall correctly.

Phone removal happened in waves. The most recent wave was about 6 months ago. Fba phone numbers went to amazon owned phone numbers with giant extensions

Iirc, FBA phones hadn't been available for two years, it's MF phone numbers that got proxied to Amazon phone numbers six months back. Then 1-2 months ago FBA addresses vanished.

why would an address be necessary to call yourself a retailer? if I paid them in cash, the seller wouldn't know my address either

If some other party did have your contact info and marketed to you based on your purchases, controlled the entire transaction, shipped the product to you, etc, each detail makes them more the retailer and the third party "seller" more of a supplier than a retailer.

According to Amazon, they are very different. A supplier supplies stuff that you sell. When convenient, Amazon asserts that they are mere conduits to the sellers’ inventory.

Add FBA, and suppliers are sort of like consigners, except Amazon claims to be at arms length, so as to avoid liability.

> According to Amazon, they are very different

So what? Thats a given. The question posed is how it is functionally different, not what legal relationships have been used to circumvent standard supply chain regulation.

the more apt comparison is Amazon as a mall for that, I'd say.

each store has its own suppliers like how each third party seller has their own suppliers, but they're selling from the same building, amazon

Except Amazon has a much more tightly ingrained relationship with 3rd part sellers than just collecting rent. They can directly influence pricing, decide which products can and cannot be sold, and take a margin of each purchase, not just a flat fee for space used (although they also take that for FBA items). I can't really think of an aspect of the retailer <-> supplier relationship that isn't in effect between Amazon and its sellers.

It feels like you're just defining the market as narrowly as needed to target Amazon. I mean you could add more arbitrary restricts to make it so Amazon appears to have a 100% market share. What's interesting is how much market power they have, not their share of randomly defined slices of the logistics and online retail market.

I think there's a substantial number of people who don't comparison shop because nobody else has a real prime competitor.

When Amazon has a hundred million subscribers to a service to ship in two days, it's reasonable to ask whether that's a monopoly.

> I think there's a substantial number of people who don't comparison shop because nobody else has a real prime competitor.

That's true of Costco as well. No one's trying to break them up.

Sam's Club is at least one competitor to Costco.

That's not what they were saying.

The comment is in response to one noting that people don't comparison shop because nobody else is a meaningful competitor for Prime. It doesn't make sense without the implication that Costco also has no meaningful competition.

Samsclub and BJs are two competitors with the same basic business model.

Walmart does 2 day shipping (and is a bigger retailer)

Yes, but Walmart's online sales are very low, and their selection is as well. If you offer 10k products for fast delivery there's an important sense in which you're not a competitor to someone selling millions of products.

Regardless, including Walmart and all other sales shipped in two days, in my estimation, Amazon still had a majority of that market.

So if Walmart with all of its resources, supply chain efficiency, and control over suppliers can’t compete against Amazon doesn’t that say more about Amazon’s ability to execute than them using “monopoly power”? If they win because they offer a better experience that means capitalism is winning.

Amazon isn't selling products, really. It's selling last mile logistics. Walmart has historically been great at the rest of the logistics chain but never really invested in last mile.

If nobody else can compete with Amazon, yes, Amazon executed better than anyone else. But they still have monopoly power. Typically a monopoly will have gotten to that position by executing better than everyone else, but that doesn't necessarily mean they should remain in the monopoly position unregulated.

Amazon has plenty of warehouses. Are you saying that if any company gets to be successful they should be broken up?

No, at no point in this conversation did I discuss what should be done in response to Amazon's monopoly power.

Isn't Amazon comparison shopping on its own though? As multiple sellers can sell the same item at different set prices? All able to offer FBA 2 day shipping?

It depends... When asked by regulators, 5% market share. When asked by customers, 35% market share.

Joking aside, Peter Thiel has great commentary on how companies avoid the monopoly label by claiming to be part of a larger community. A classic example is that Google search has at least 70%+ market share. However, when asked by regulators, they’re a “tech company” with many competitors and have a significantly lower market share.

We know how big a mess Marketplace is but I never realized how awfully big that was:

1. 58% of global Amazon ecommerce is actually made by third parties using its platform, not by Amazon itself

2. This 58%, the global third party sales were $160bn

3. Amazon charged a whopping $43bn in fees for this. That's 27%... and it is still worth sellers to use the Amazon platform.

How much of the 58% are fakes or worse, then we can start figuring out really how worthwhile it is for sellers

I didn't like the analysis because markets are mushed together. Amazon market share for cars and books is different. It is meaningless to analyze the monopoly impact where monopoly is defined as the whole market for all us goods. They have to be broken out or corrective action will be postponed for too long.

The author makes this exact point if you read past the 3 bullet points up top, such as the part about how Amazon has 50% of book sales, noting that to regulators it matters how you slice it.

Is Amazon "monopoly"? Yes, just like Google. How to verify it? Block every single address used by Amazon (or Google) and watch your internet and gadgets revert to the stone age and stop working completely. That is a de-facto "monopoly".

I wonder if there is a potential likely scenario that one or more tech companies infiltrate the political system and slowly transform it into an evil techocracy. Maybe we are in that progress already.

Pretty much the definition of lobbying...

I think the more important number is not market share but profit margin / cost per goods sold. That's the number stores have to beat or compete with.

Why are people so concerned about Amazon but not about Google?

I believe it's becouse Google is more abstract. Local stores are not shutting down due to Google. However I am equally concerned about Google.

Amazon's current market share (whether online or overall US retail) is not the point.

The point is how fast is that market share growing and where will it be in (say) 10 years time?

What's scary about Amazon for me isn't its future takeover of the US market. I'm more worried about it taking over all markets in the civilized world. Europeans are already using Amazon to purchase things online massively and I don't know what the situation is like in Australia but I'd bet a keg of beer that Amazon will have a substantial market share on all continents and in most civilized countries by 2025 if it doesn't already. There's very little one can do to stop this kind of juggernaut enterprise barring some local laws and sanctions. Amazon is also going to change the future of commerce, already has, really.

On the opposite. It is really easy to stop big corparations barring laws. Way easier than small.

If authorities would hold Amazon responsible for the scams and IP infrightment on their platform Amazon would have to kick out half of their sellers.

Reply on news.ycombinator.com