Archives

Is Amazon too big?

This video is sponsored by Brilliant! The first 200 to use the link in the descriptionget 20% off the annual subscription. Elizabeth Warren wants to break up Amazon,Facebook, Google, and Apple. The argument goes, roughly, like this: Techcompanies have gotten really big, really fast. And they’ve abused that power by favoringtheir own products over their competitors. Therefore, once a company reaches a certainsize, it shouldn’t be allowed to own both the products and the platform on which they’resold. In English, Amazon can’t sell the Kindleon its own website, and Apple has to pick between owning Music and iMovie and News orthe App Store itself.

Now, whether Warren is 100% serious or mostlyjust generating publicity for her presidential campaign, she isn’t alone. This is only the latest part of a much biggermovement. We’re at an inflection point in history,where new, fast, Silicon Valley is crashing in to slow, old, government. Everywhere, all at once, the power and influenceof big tech companies are being questioned: Google was recently fined $1.7 billion inEurope for being anticompetitive. Sprint is trying to merge with T-Mobile, despitesignificant push back. Facebook has, well, continued to be Facebook. And Spotify has launched an all-out attackagainst Apple for what it claims is unfair treatment. So, who’s right – Apple or Spotify? Google, or the EU? Facebook, or, literally everyone else? First, we need to understand why these companieshave such a huge advantage. And how, in just a few years, Amazon wentfrom being synonymous with “cheap, convenient shopping”, to a scary, political, nebulousWalmart-like mega-corporation. Let’s say you wanna start a grocery store. Maybe you know a little bit about merchandising. Maybe you come from a long line of grocersso selling produce is just in your blood. The other kids were playing with fire trucksand trains but you, you were daydreaming about the retail implications of The Engel Curve. Anyway, the bad news is that the grocery businesssucks. Like, famously so.

If you’re lucky, you might manage a profitmargin of 3%. Unless you have some revolutionary way ofarranging bananas on the shelf, you’re just one of a thousand stores, which customershave no special loyalty towards. You don’t see a lot of “Proud mother ofa Safeway shopper” bumper stickers. But – there is money to be made at the very,very top. If you can become a Kroger, or a Whole Foods,or Trader Joe’s, well, that’s a different story. The trick is surviving long enough that yousell lots of things, so you can, (a), turn around to the companies making those thingsand say “Hey, we’d like to buy 3,000 stores worth of your bananas, can you make us a deal?”, and, (b), cut out the middle man by creatingyour own generic brand. With size comes leverage, which lets you buycheaper, and, ultimately, make more money. But, again, the problem is getting there. Competing with established companies in anyindustry usually means losing a lot of money for a long time with only the hope of makingit back in the future. But don’t give up on your dream quite yet. Here’s an idea: Forget groceries for now,let’s just find some way of making money. Like, I don’t know, selling cloud storageto enterprise customers! It’s a good business, no one else is doingit very well, and, in a few years, you’ll have so much money, you can come back to yourdream of starting a grocery store. What does cloud storage have to do with sellinggrapes? Is that really the most exciting thing youcould be doing? Pretty much nothing and probably not. But who cares! Money is money, and as long as it makes morethan the grocery store loses, you can afford to slowly grow it into an empire, even whileit isn’t yet profitable. This, if you haven’t noticed, is my verycrude way of describing Amazon. It started as a book company, but Bezos hadno special love for books. That was always just a good way of generatingcapital for his real dream. Today, books are a footnote. The new distraction is called Amazon Web Services- AWS. If you’re already familiar, bear with mefor a sec. A few videos back I said: > A thousand downloadsdon’t cost any more than one. Scale is (nearly) unlimited. My point was: it’s a whole lot easier tosell a thousand note-taking apps than it is a thousand actual notebooks because softwareis made of bits, and bits don’t cost money. Which, is mostly true in that context, but,not totally accurate in practice. Consider the scale at which some companiesoperate: Out of all the bandwidth, from every phoneand every computer, in every country, 15% is just people watching Netflix.

15%! Even Uber, which, in theory just connectsthe nearest driver to the nearest rider, stores over 100 petabytes of data – or 100,000,000GB. It also fluctuates dramatically. Every startup dreams of hitting the frontpage of Reddit, Unless you’re the engineer, in which case you have a heart attack tryingto keep up with such a huge spike in views. Companies, and, especially, startups withlimited budgets, have a tough choice: Either buy too much capacity, Or save moneyand hope they don’t get too popular. At least, until AWS. Amazon realized it could solve this problemwith a service: Only pay for what you actually use. If your business suddenly explodes in popularity,no problem, just pay a little more. Turn the handle for more data, as you wouldwater or electricity. Amazon takes care of the rest, the same waywe outsource building windmills to electric companies. Now, if we look at its total revenue, andthen divide it by source, it’s pretty much what you’d expect: Amazon is mostly an onlinestore and you’re probably wondering why we’re talking so much about AWS. But what about its income? Where is it actually making a profit? This is where it gets interesting. Now, Amazon looks like a cloud storage companywith an online retail business on the side. In the 4th quarter of 2018, AWS accountedfor 58% of the company’s operating income. It alone made more money than McDonalds. So, yeah, it sells lots of USB cables andbananas, but that’s not where the money is. AWS is camouflage. It makes the company look good overall andconceals how much money it loses. One business subsidizes another. And this is where it gets tricky. Because, if you’re one of the other grocerystores, you’re thinking “This isn’t really fair – how can we compete with someonewho doesn’t even need to make a profit?” Safeway and Publix don’t have a $25 billiona year cloud storage businesses. When they sell bananas, they have to, like,ya know, make money.

This is how, one after another, Amazon entersand dominates a new industry. It plays by a fundamentally different setof rules. Turns out it’s a whole lot easier when you’renot super worried about the whole profit thing. Of course, predatory pricing, when a companylowers its prices to starve out the competition, isn’t a new idea. But once they’ve done so, companies usuallyraise their prices again – that’s the whole point. Amazon, on the other hand, has always keptits prices low. It’s not playing the long game, it’s playingthe looooong game. Here’s it’s revenue, and here’s it’sprofit. The company touches more money than ever – itjust doesn’t keep it. Profit has stayed around 0 because it’smore interested in growth. That’s the loophole. In this essay, researcher Lina Khan explainshow, since the ‘70s, antitrust law has used short-term prices to determine whether a companyis being anticompetitive. In other words, sure, Amazon is big, it’sdominant, and it’s killing lots of competitors. But it’s prices are low, so it flies underthe radar. You might be thinking – so what? If a company uses its size to save you andme money, isn’t that a good thing? But when products are subsidized, either byanother profitable business like AWS, or, Venture Capitalists burning money for thesake of growth, they don’t have to compete on their own. Products win not because they’re the bestbut because they’re funded by someone, or something, unrelated. For example, on iPhone, Apple has the PlatformAdvantage. It controls which apps are allowed on theApp Store, and doesn’t have to give up 30% of its revenue or follow the same rules, aseveryone else. If you’re Clash of Clans, this may seemlike a relatively small price to pay for access to 1.3 billion users. For someone like Spotify, it’s a very differentstory. Music streaming is the digital equivalentof a grocery store – Spotify has such tiny margins that giving Apple 30% breaks the entirebusiness model. And even if Apple didn’t make a dime fromits Music or News apps, it might still offer them just to attract users to the iPhone. Spotify needs to make money, but Apple Musicjust doesn’t. Apple’s apps, therefore, almost certainlyhave more users than they “should”. Which is not saying they’re good or bad,but that some number of people, maybe 1, maybe 1 million, use the service only because Applehad an unfair advantage in putting it in front of them. Likewise, Amazon has the Platform Advantageon its website. At some point it realized, hey, wait a second,if we have all the data, and we control what people see, why on earth are we sending customersto someone else’s product? So now it competes with its own sellers, oneverything from batteries to backpacks and keyboards. But wait, how is that different than any othergeneric brand? Target has Up&Up, and Walmart, Great Value,but no one’s complaining they have an unfair advantage. The difference is lock-in. It’s much easier to switch grocery storesthan it is between iPhone and Android. Companies like Facebook will always say “Look,you chose to use our service, you chose to give us your information, didn’t you quityour job, become a lawyer and read our 3,000-page terms of service?” But that’s not really true. We made one, unrelated choice, like buyingan iPhone or Android, which required that we make a bunch of other choices later on. Nobody knows what they’re getting into. And this will only happen more as companiesget even bigger. Amazon is an extreme example because AWS isreally profitable and groceries are really not, but entering new categories with theresources you already have is kind of what a company is. You might start by making smartphones butthen use that money to sell refrigerators. Fast-forward a few years and now you selllife insurance and container ships. Samsung is less a brand and more a Buy-N-Large,E-Corp conglomerate. Like Amazon, it barely makes sense to thinkof it as a single, unified company. One division sells parts for the iPhone. Another fiercely competes against that verysame device. Even Apple is moving in this direction. It may not be in the business of refrigerators,but you now buy your iPhone with an Apple credit card, download Apple apps, back themup on Apple’s cloud, and watch Apple-branded TV-shows. But there’s also a benefit to this integration. One of the bests feature of the iPhone isthat it’s all designed by one company, as one, coherent product. Because Apple owns both Music and iOS, they’reeasier to use, more convenient, and more powerful together. And because you have no choice but to useApple’s App Store, your phone is more secure and your data more private. Now, of course, you may disagree. For some, having more freedom might be worththe trade-off for privacy and security. But that doesn’t diminish its value forthe rest of us. In other words, breaking up some of thesecompanies would actually mean a worse experience for you and me. And if your “consumer protection” proposalmakes our lives worse, it’s probably a bad one. The EU has shown – time and time and timeagain – that governments can make technology worse simply because they don’t understandit – even with the best intentions. So, what’s the solution? I don’t know. If anything, we’ve learned you should beskeptical of any simple solution to a problem this big. Instead, here are some ideas: First, we need to expand the scope of whatqualifies as anticompetitive behavior.

Low prices don’t mean the customer isn’tbeing harmed. On the other hand, closed, locked-down marketslike the App Store aren’t necessarily always a bad thing. Second, we need to reexamine some mergersand acquisitions. It happens all the time: An exciting youngstartup gains some traction only to just be bought by a Google or an Amazon. Sometimes we never heard from it again. Founders are incentivized to sell their companies- to the tune of billions of dollars. But society at large would be better off withmore competition. The key is balancing the benefit we all getfrom the scale of companies like Amazon, with the drawbacks of their immense political andeconomic power. Many of these ideas to break-up tech companiesare designed only to get headlines. But a real solution requires a deeper, mathematicalunderstanding of the problem, of the type you can learn on Brilliant.org. Their pitch is simple: Sure, you can spendhours reading textbooks, but that’s not how most of us learn. We need examples, challenges, and pictures. Their courses on Science, Math, and ComputerScience are interactive and you can do the problems a few minutes at a time. Even if you’re just waiting in line, youcan quickly do one of their Daily Challenges. I just had to take a 17-hour flight, so Iwas glad to see you can now download courses to use offline. Use the link in the description to sign upfor free. The first 200 people will get 20% off theannual premium subscription so you can view all the Daily Challenges and take all theirproblem-solving courses.