Digital networks have two significant properties: they can expand very rapidly, given the right conditions, and, if they do grow, powerful network effects kick in very quickly. A network effect is the phenomenon whereby the value of a network increases according to the number of people using it. In the 1980s, Bob Metcalfe tried to quantify the multiplier effect of additional users with what became known as Metcalfe’s law, which says that the value of a network is proportional to the square of the number of users. A network with 10 users has a value of 100, but one with a thousand users has a value of a million (1,000 squared).
Ever since the web arrived, the overriding mantra of online entrepreneurs has been “get big fast”: recruit users or customers as quickly as possible to get to the point where the network effect kicks in and discourages potential competitors. For most of the successful firms, the way to achieve this was to offer free services. This was the route chosen by Google and, later, Facebook and YouTube. For Amazon, growth was achieved by low prices, efficient dispatch, good customer service, prompt delivery and a colossal inventory.
But because capitalism never knows when to stop, eventually even the exponential growth rate promised by Metcalfe’s law wasn’t good enough for Google and co. They noticed Reed’s law, which said that if a network can have sub-groups then its utility can grow at an even more colossal rate. And so they began to turn themselves into platforms on which entrepreneurs could build services. A platform in this sense is a set of tools provided by the owner to enable third-party developers to create applications that interact with core features of the platform (thereby keeping users within the network to leave monetisable data trails).
The two companies that made the most of this idea were Amazon and Facebook. In 2000, Jeff Bezos launched Amazon Marketplace, a platform that enabled third-party sellers to sell goods on the company’s platform. In 2007, Facebook unveiled its “Facebook Platform for developers of social applications”.
Both of these initiatives prospered mightily. More than half of Amazon’s annual sales now come from third-party sellers. And because it’s easy to become a seller, huge numbers do. One report reckons that more than a million joined in 2017 alone. Facebook had only 20 million users when it launched its platform in 2007. It proved a turning point for the company; the resulting surge in third-party services (online games, quizzes, dating apps, etc) turbocharged user growth and took it to where it is today – with 2.4 billion users.
Three cheers for platforms, then? Er, not quite. For years, there have been reports of how scammers thrive on Amazon Marketplace. The formula was simple: sign on as a seller; list various items at bargain-basement prices; take the orders and claim that you have shipped the goods; receive payment from Amazon (always a prompt and efficient payer) within two weeks; then scamper before the angry feedback arrives. For years, Amazon has played whack-a-mole with these crooks, with limited success because it’s a truly sisyphean task.
But last week, a different challenge emerged. An investigation by the Wall Street Journal claimed that Amazon had “ceded control of its site” with the result that it was carrying thousands of banned, unsafe or mislabelled products. “Just like tech companies that have struggled to tackle misinformation on their platforms,” said the Journal, “Amazon has proven unable or unwilling to effectively police third-party sellers on its site.”
Sound familiar? It was Cambridge Analytica’s exploitation of Facebook’s platform, remember, that led to the current crises assailing the Zuckerberg empire. The company’s initial response to the Observer’s exposure of the scandal was that CA was simply a bad actor exploiting an innocent platform. Amazon has responded more maturely to the WSJ revelations saying that it’s doing its best, which I’m sure it is.
But in both cases the moral is the same. As the tech analyst Benedict Evans summarised it last week in his newsletter: when you let anyone on Earth on to your platform, that includes bad people. “Conceptually,” Evans says, “this is no different to what’s been happening on Facebook or YouTube, except now it’s lead in children’s toys instead of Jihadi videos”. He’s right: tech platforms are double-edged swords. The problem is that the companies that live by them don’t die by them.
What I’m reading
The science of sleaze
Why were so many scientists apparently charmed by Jeffrey Epstein? Katha Pollitt has an interesting essay in the Nation.
Jonathan Mayer and Arvind Narayanan do a terrific demolition job on Google’s excuses on tracking protection on the Freedom to Tinker blog .
Palladium has a scarifying critique of Yale by Natalia Dashan. But I suspect it could apply to most Ivy League colleges, all of which are hedge funds with nice universities attached.