Internet Land Grab 2.0
“Give me a lever long enough and a fulcrum on which to place it, and I shall move the world.” - Archimedes
My former Wired colleague Cade Metz recently wrote a good article on WhatsApp’s 900 million user milestone, a milestone that the company reached with just 50 engineers. How did they do it? Following a presentation by one of the company’s engineers, Cade points to Erlang as the long lever that lets a handful of engineers move some 12 percent of the world’s population.
There’s no doubt that Erlang is a great language for messaging apps, and I don’t in any way want to detract from the WhatsApp software team’s obvious talent, but there are two problems with this framing: 1) Erlang is just the very tip top of a really tall software stack, and 2) It takes way more than 50 engineers to reach 900 million people.
Both of these points are related, and they both speak to the “are we in another bubble” question that currently vexes the valley.
The Rented Lever
WhatsApp’s unseen hordes of engineers and the massive, global compute infrastructure that those engineers are constantly growing and maintaining, are all rented in very small time slices by the full-time members of the WhatsApp engineering team. It’s the combination of on-demand engineers, the on-demand compute resources that those engineers are packaged with, and the proliferation of cheap, capable client devices at the network’s edge (i.e. smartphones and tablets) that makes up the real force multiplier–the lever, if you will–for WhatsApp’s core team, and not any particular technology.
When you think about the fact that this very same force multiplier can be rented by anyone with a credit card, a laptop, and an internet connection, then you realize that the breathtaking scale and speed of WhatsApp’s success isn’t any more of a “tech” story than, say, the similar success of BuzzFeed. Rather, it’s a story about the sudden, outsized rewards that can accrue to a relatively small amount of properly timed effort applied from just the right spot to a globe-spanning networked computing platform.
Because WhatsApp is renting a really long, world-moving lever that anyone can rent to move anywhere from a few dozen users to a few billion, then the difference between the 50 engineers at WhatsApp and any other group of talented engineers with access to the same really long lever of networked compute resources, is that someone at WhatsApp has pointed those engineers to a fulcrum that’s sitting in exactly the right place – in this case at the nexus of free SMS and no ads, no games, no gimmicks – so that when they grab that lever the world moves.
Thus the secret to success in the present moment – if we define “success” as “I have a billion users” – seems to lie in grabbing the big public lever and being the first to iterate into one of the primo fulcrum spots that’s characterized by minimal in-house engineering hours and rapid user growth.
The Fulcrum Lottery
I think the dynamic I’ve described here – a bunch of investor money scrambling to find exactly the right spot from which to leverage the same core technology stack to grab outsized user numbers and valuations – is why so much of the present startup environment has lottery feel to it, or like a scramble for location instead of a surge of innovation. WhatsApp is the group messaging app that scored, Instagram is the photo sharing app that scored, Uber is the ride sharing app that scored, Airbnb is the room sharing app that scored, Slack is the group chat app that scored – the lucky winners are the ones who are the first to wiggle their way into the right twist on a basic idea, with a precise combination of bells and whistles that can constitute a fulcrum spot from which they can move a mind-bending number of users.
If WhatsApp had offered users ads but no gimmicks or games, or no gimmicks and ads but games, or some other variant on free SMS, they probably wouldn’t be where they are, Erlang or no. If Airbnb hadn’t hit on the famous let’s take glossy photos of user apartments idea, would they have ever taken off? Here is the part where I put on my asbestos underpants say something heterodox: tweaking a product that sits at the very tippy top of an epically tall, rented software stack until you either start growing or run out of money isn’t innovation, it’s gambling.
Yeah, I know, I’m a jerk, because a startup is the hardest and most punishing thing you can do, and it’s considered rude to suggest that anyone’s success at it is largely attributable to luck and not “innovation”. And I’m sure you’re also thinking of Thomas Edison, and how he famously tried 1,000 different filaments before inventing the light bulb and then boom, innovation! Isn’t Airbnb just like that? No, it isn’t, because that story is bunk.
As Thomas Parke Hughes describes in his magisterial Networks of Power: Electrification in Western Society, 1880-1930, Edison and his team at Menlo Park built an entire, low-cost electrical distribution network, of which the light bulb was one of a number of critical new components. It wasn’t like the electrical grid was just sitting there with empty sockets, waiting for Edison to diddle his way through the solution space until he stumbled into just the right filament so that he could staff up and scale; he had to invent, finance, manufacture, and deploy the entire grid in order to take on the entrenched gaslight industry.
We’ve seen this sort of location scramble before in the dot-com bust, with the race for a desirable location within the space of possible domain names that could attract “eyeballs”. Just like the last time around, the cost of entry is low (i.e. the lever is cheap and the amount of effort required to operate it is low), the Bay Area is bursting at the seams with talent and money, the odds of individual success are astronomically slim, and the reward for hitting a sweet spot is a huge valuation that’s justified mainly by the size of the user base and not the size or sustainability of the revenue stream.
Context Switches and Confusion
Some of what’s driving the present mania around what I’ll call fulcrum plays–small team + magic product formula + long, rented lever = exploding user graph–is the suddenness of the qualitative change that the network has undergone when the growth in nodes went vertical post-iPhone. It’s like there has been a regime change in the system, and we still don’t have the mental context for dealing with it.
I remember back in the dot-com era, people didn’t have any experiential precedent or mental context for an open network that literally anyone could use to publish text and images to so many, so cheaply and frictionlessly. The scale of the thing blew society’s collective mind, so if your little team could put up a website and attract a million “eyeballs”, then people would throw money at you because clearly you were doing something miraculous. After all, previously the only properties that got in front of a million people were capital-intensive: TV shows, print publications, or sound recordings. But the million-user miracle wasn’t really yours – you were just leveraging the network in a pretty straightforward way. And because everyone could do it, everyone did, and eventually it wasn’t special and investable anymore.
Thanks to the smartphone revolution, which very quickly (as in the last 7 years) added a few billion more nodes to the network, we seem to be going through the same type of rapidly-changing-mental-context-driven freak-out yet again, except the numbers are all a few decimal places larger. People just can’t get over how large of a user graph you can assemble so quickly with so few programmer-hours, so they throw money at you for it because you must be doing something miraculous – look at that growth and those usage stats!
I think we can extrapolate this trend and predict that we’re not many years (months?) away from seeing a single programmer scale a piece of software to a billion users, because thanks to actual innovation in the lower parts of the stack it’s getting increasingly easy to scale to new heights – not that a billion-user app is currently trivial, but it will be trivial soon enough. And when it’s trivial, then “hey look at my billion users and how fast I got them with this tiny team” won’t be special anymore, and it also won’t be valuable. And when that moment arrives, if paper valuations are still tied to user growth instead of revenue growth, then we can expect history to repeat itself (again, with the decimal a few more places to the right). But until then, we’ll keep partying like it’s 1999.
Addendum: This Time Is Different?
One of the main objections to the bubble idea is the fact that the general public isn’t participating this time around. Private markets and venture capital and all that.
I have a theory, though, that the public is participating – they just haven’t gotten the bill, yet. The public is participating via the Fed’s zero interest rate policy and bouts of quantitative easing, which are driving the desperation for yield that is in turn funding the growth of the unicorn herd. When all of the institutional money that’s levered into the current bubble begins to run for the exits, who do you think will be on the hook for bailing out the losers that are too big to fail?