A Design for Life: How Platforms Are Weaving Their Way Into the Fabric of Capitalism
by James Meadway
29 March 2019
News this week that Apple’s Event – once the favoured platform for latterday guru Steve Jobs to parade shiny new trinkets – was devoted solely to foregrounding its subscription services tells us something critical about the direction of travel for capitalism.
As Grace Blakeley has written for the New Statesman, the global economy is not in the best of health. For companies supplying consumer hardware, the problems appear immediate. Rising costs were already leading to squeezed margins, but slowing demand ramps up the pressure on manufacturers. In January Apple reported its first profits warning since 2002, citing slower-than-expected iPhone sales. (Tellingly, the company will no longer be reporting how many iPhones it actually sells.) Samsung, meanwhile, also reported a “surprise” profits warning this week, blaming a slump in memory chip prices and demand for its handheld displays.
It’s easy enough for manufacturers to blame their problems on relatively short-run demand conditions. The slowdown in China is real enough and, even without the dampening effects of the government shutdown and ongoing trade disputes, the US economy is starting to look decidedly peaky. But the longer-term challenges for the platform companies – those companies (in Nick Srnicek’s definition) using data technologies to bring different groups together, like Facebook, Amazon, or Google – are posed by threats to their core business models.
Fundamentally, platform companies make their money from some combination of two things: first, by offering a service that (usually) becomes more valuable the more people use it; second, by extracting the data this generated by their users and marketing it in some form. The first creates a major drive towards the monopolisation of a particular service (“competition is for losers” as serial tech investor Peter Thiel put it). There is one Facebook, and for much of the world, really only one viable service offering whatever it is Facebook now does.
But this creates a problem down the line: if your company’s profits depend on adding more and more users, there are only so many eyeballs in the world you can grab. With well over 2bn users, Facebook is approaching saturation point. Accounts of a slowdown in its user growth last year led to the company suffering the biggest single-day fall in its market value of any company in US history, and reports of younger users switching off in droves are hardly a sign of a healthy future.
Meanwhile, Facebook’s attempts to market that data have run into a few hiccups. A combination of a slowdown in user growth, challenges in turning their data into money, and steadily rising costs of infrastructure are all bearing down on the platforms. Even many of those offering paid-for services, like Uber or Netflix, are burning through cash and have yet to report a profit. And whilst artificial intelligence (heavily dependent, in its current form, on access to huge troves of data) offers some new prospects through the creation of new markets, it also requires heavy investment and is likely to reinforce the winner-takes-all economics of platform capitalism, leaving little space for the losers. Adaptation is needed, even at vast expense.
‘All your banks are belong to us’.
Until now, Apple has depended (very successfully) on a business model in which owning the glossy electronic product, from iPods through to iPhones, in turn gave you access to content that either worked best (or only worked) on Apple products. Everything integrated with everything else, as far as possible, locking users into sticking with Apple products. But with its latest incarnation of this business model, Apple TV, failing to take hold in the way the company would have liked, Apple has confirmed its TV content will now be available on non-Apple products. In other words, the company is making a clear break with what it did in the past.
But the most important news in Apple’s announcements last week wasn’t the giddying prospect of another season of Carpool Karaoke. It was the launch of Apple’s credit card, in cahoots with Goldman Sachs. Previously, its Apple Pay service provided only the bread-and-butter of facilitating transactions. But now Apple will be offering loans to its customers. Facebook has been exploring payments services on its platforms for at least five years, and won an Irish banking licence two years ago, with some false starts to its name. Google has also just acquired an Irish banking licence – Irish licences allowing access to banking markets in any other EU member state. Amazon has been offering not only payments services but small business loans for the last few years. But far ahead of them all is Ant Financial, the $150bn online financial services provider that grew out of Alipay, itself originally the payments service for Alibaba, China’s (more profitable) Amazon equivalent.
This lurch sideways into money-dealing by the Western platforms makes a lot of sense. A bank specialises in information, of a particular kind: if a bank wants to make money, it has to make judgements, over time, on the ability of its borrowers to repay their loans, with interest. The more it can know about its borrowers, the more loans it can make safely, and the more profits it will make. And, so far, if the tech giants want to make money, the more they can know about their customers, the better: either to slice them up into neat market segments, for sale to advertisers (licitly or otherwise), or to better target them with selected products from their own marketplaces. In both cases, companies will make more money the more efficiently they can scoop up that information.
Interestingly, there is good evidence to suggest that banks have not, certainly during the 2000s boom, been particularly efficient at doing this. Like the tech companies, banks and financial institutions saw operating at scale as the route to rising profits, rather than necessarily meaningful innovation or cost-savings. And one way to think about the crash that ended the boom in 2007-8 is that it was a spectacular failure of information: the banks thought they knew enough about the subprime mortgages hiding in their complex financial products to be able to make safe predictions about the behaviour of those mortgages, and thus make money forever more. (Spoiler: they didn’t.) The thought that existing banks may be fairly inefficient users of information is one of the factors pushing regulators into supporting fintech, for instance, in the belief that spur of competition will sharpen the industry up.
But, equally, if you are a platform company sitting on a vast trove of user data, and running out of quick ways to monetise it, you might sniff a few opportunities here, too. The vanilla version of this is to creep into every customer’s financial interactions by offering payments services. This means quietly squatting between the buyer and the seller, and transferring money from one to the other when requested – think PayPal or Apple Pay or Google Pay. But whilst, ideally, you can take a slice out of every transaction, it’s only going to be a small slice (reports suggest Apple Pay takes 0.15% of each transaction), and liable to decline over time. What you really like is the lovely high-quality data that payments services produce. It’s high quality because payments are (of necessity) accurately recorded, and conducted with a high degree of consumer trust. But as always with data, you need a lot of it to extract the value, and so you need to drive hard for scale – quietly sliding yourself into every single transaction every possible customer will ever make. Apple Pay currently sees over 1bn transactions every three months – and that number is rising.
However, much of the market remains sewn up between existing financial institutions and, as Facebook’s payments services problems have shown, it’s harder to crack than it might at first appear. A far juicier prospect for newcomer tech companies, once their payment systems are in place, is to begin to perform finance’s magic trick and exploit the bigger margins to be made in offering credit and loans – particularly if you can mobilise your vast data reserves to enable more informed (if not necessarily better) decisions to be made about who receives which loans.
Beyond that point, on the further shores of what might currently be attempted, Facebook is spending significant sums in developing its own cryptocurrency, for use via WhatsApp. Linked to the dollar, and reportedly intend for launch later this year, Facebook initially intends to focus its cryptocoin on India’s $69bn remittances market – the world’s largest.
The future of capitalism.
There’s a line of march here. You might even call it progress. Take Apple as the example. You start by selling a product – computers and then, from the 2000s, MP3 players, in the form of the iPod. Then you start to sell services for the product – Apple Music. Then you sell services, regardless of product – Apple TV. Then you sell the ability to buy services – Apple credit card. Every step on the road gets further away from the messy and uncertain business of charging money for physical goods or delivering a service, and closer to the ideal form of capitalism – which is, of course, to hand over money in exchange for nothing but more money.
What matters above all else is something central to the operation of capitalism today. As profits become harder to find through the textbook route of investment and the creation of new markets (whether for goods or services), capitalism in general is turning towards various forms of rent-seeking. It’s significantly easier to enforce property rights on what has been created elsewhere, and demand tribute for access, than it is go through the costly and risky business of creating new value yourself.
This rent-seeking becomes easier still if – as is the case with various platform companies – something close to a monopoly can be maintained in a market. The lines between rent-seeking and value creation are usually blurred: Uber’s software and data undoubtedly produce some additional value, but surely not the 25% cut it takes from its drivers. From the point of view of generating profits, though, it really doesn’t matter where the balance lies. There’s no moral case for pulling one way or the other. It’s all just money.
What does all this tell us, and those of us on the left in particular? Three things.
First, that the struggle for control and ownership of data will intensify. The social issues are huge and, as Cambridge Analytica demonstrated, beginning to turn into an active public concern. As the push to monetise data intensifies, concerns will take on an increasingly economic colouring.
Second, as the pressure for profitability on digital platforms increases, so too will their drive to squeeze revenue out of new markets. The more they eat into areas of economic and social life previously outside their orbit, the more their regulation and control will become a defining economic policy issue. This includes finance, and the particular risks finance generates for the system as a whole: it is, for instance, interesting to speculate about how prepared our financial authorities are for a crisis emerging in and around the mutant combination of Big Data and Big Bank.
Third, we are going to run into a fight over payments systems and how the fundamental infrastructure of modern capitalism operates – and it is unlikely that the traditional socialist answer of demanding more control over a central bank will provide all the answers when money (and the institutions that handle it) are drifting steadily outside government control. We need to think more creatively about the issues of regulation and democracy as applied to markets, and how we can democratise payments and the money system itself.