Eureka! We've discovered network effects.

Photographer: PASCAL GUYOT/AFP/Getty Images

Facebook Is Big, But Big Networks Can Fall

Megan McArdle is a Bloomberg View columnist. She wrote for the Daily Beast, Newsweek, the Atlantic and the Economist and founded the blog Asymmetrical Information. She is the author of "“The Up Side of Down: Why Failing Well Is the Key to Success.”
Read More.
a | A

My column this week on Facebook and privacy triggered the inevitable query: “But what about network effects?”

Few buzzwords are hotter in tech circles than “network effects.” This was so 15 years ago, when I was an MBA candidate grinding through job interviews; it is so today. Probably, when the heat death of the universe is imminent, and our nine-tailed descendants are trying to figure out what to do, some bright Johnny will suggest we can keep things going if we can just add another 2 billion stars to our user base.

Don’t get me wrong: Network effects are important, and I frequently talk about them in relation to everything from media companies to neighborhoods to choices about motherhood. But when I hear the term, the hairs rise on the back of my neck, because it's often used imprecisely. People say “network effects” when they are really talking about switching costs, or regulatory coordination, or spillover effects, or any number of other things that are at best tangentially related to what the network effect model was built to describe.

Worse, far too many people seem to use the term the way college sophomores deploy the names of philosophers they have just read, in the mistaken belief that a piece of jargon can magically banish disagreement. Your firm doesn’t seem to have a viable revenue model? You’re just saying that because you don’t understand network effects! Someone seems insufficiently worried about the market power of some technology behemoth? It must be because that benighted fool has never heard about network effects!

Network effects are a useful concept, but not when deployed in this slipshod way. Worse, such careless routine deployment actually threatens the concept's usefulness in conversations where it does offer real insight.

So just what is a network effect? The term describes a product that gets more valuable as more people adopt it, a system that becomes stronger as more nodes are added to the network. The classic example of network effects is a fax machine. The first proud owner of a fax machine has a very expensive paperweight. The second owner can transmit documents to the guy with the pricey paperweight. The thousandth owner has a useful, but limited, piece of equipment. The millionth owner has a pretty handy little gadget.

We have to be a little careful about this, because we are not simply saying “more users are better.” When more people buy Ford automobiles, the company can amortize the fixed cost of its capital plant over more units, making the cars cheaper to produce, and allowing the company to offer them to me at a more attractive price -- but this is a phenomenon that is properly described as “increasing returns to scale,” not “network effects.”

People also often describe things as network effects that have nothing at all to do with the number of nodes in the network. For instance, people will say leaving Facebook is hard because all your photos and data are in its system. But this would be true even if you were the only user left on Facebook; it is what economists call a “switching cost,” not a network effect. In fact, switching costs matter a lot: Network-driven markets are more likely to converge to monopoly when switching costs are high than when they are low. But switching costs are just as important in non-network markets, and they should not be described as if they are a feature of network-driven markets, rather than a basic fact about many sorts of economic activity.

Now that we’ve gotten all that out of the way, onto the obvious question: Why is everyone so excited about network effects? Most usually, because they are thought to give companies great power. Consider Google, which uses data gathered from user clicks (among other things) to decide which search results are most valuable for a given query. The more people, the more clicks, the more data. Starting up another search firm is hard, because you need a lot of users to compete with Google … only it’s hard to get a lot of users when your results aren’t as good as Google’s.  In markets with strong network effects, the natural equilibrium number of firms may be one, because a single network provides the most value to users. Of course, in such a situation, the owner of the network can generate monopoly-like rents.

The prospect of a quasi-monopoly is very exciting to venture capitalists and stock market investors, and by the same token, very dismaying for folks who want to start a competitor business, or those who worry about the various ways that the network owner might use all that market power. So you tend to hear a lot about network effects in two sorts of conversations: investment pitches, and demands for more regulation.

These claims about network effects are not wrong, but they are often overblown. Network effects are a strategic advantage, but they are not a license to print money. The VHS tape was a beneficiary of network effects until it was knocked out by a superior format. Microsoft still dominates the market for desktop operating systems, but desktop operating systems no longer dominate our technology usage. History is littered with the remains of technologies and companies that used to enjoy the luxury of network effects -- and then found that what network effects giveth, they can also taketh away.

You can think of network effects as a little bit like leverage. Debt magnifies your fortunes in both directions. When the housing market is rising, you can put down a modest sum as a down payment, and enjoy 100 percent of any equity gain in the house. During the boom, people could double or triple their initial down payment in a relatively short period of time, sell the house, and walk away with all that cash. This wouldn’t have been possible without leverage, in the form of a mortgage. 

But when the market tanked, those same people were left with an asset that couldn’t be sold if they got into trouble, and many of those people ended up in foreclosure, or bankruptcy. If they’d paid cash for the house, that wouldn’t have happened.

So with network effects. When your network is growing rapidly, things are splendid! Every new user increases the value of your network and encourages even more people to join. But there’s a small catch: What happens if your network starts shrinking? Suddenly, it’s getting less valuable, which means more people are likely to leave, which makes it even less valuable. Rinse and repeat all the way to the court-appointed receiver’s office. More conventional businesses don’t have this vulnerability. If people start eating fewer potatoes, farmers can just plant fewer potatoes. They don’t have to worry that everyone’s going to stop eating potatoes just because the neighbors decided they don’t care for them.

Looking at the most recent Pew study on Internet usage among young people,  I see that 71 percent of teens use Facebook, with the median user having slightly less than 150 friends. Forty-one percent of them report that they use Facebook most often. But when I look at a similar Pew study from 2013, it looks to me as if 76 percent of teens were using Facebook, with a median number of 300 friends, and 81 percent of social media users reported that they used Facebook most often. If I were Facebook, those numbers would keep me awake at night -- not because Facebook can’t survive with only 70 percent of the market, but because a network that is getting smaller and less valuable to its users is a network that is very vulnerable to disruption.

What’s actually astonishing is just how evanescent such strategic advantages have proven. Fifteen years ago, people worried that Microsoft’s network-effect advantages made it unstoppable; now it’s an also-ran in everything new-market except gaming consoles. The rotting corpses of old social media sites litter the landscape. And of course, finding a place to send Aunt Maisie that birthday telegram is getting darned hard.

That’s because network effects are, for all their power, surprisingly narrow. Even companies that dominate a niche for a long time, as Microsoft did, can be undone when some other niche becomes more important. Facebook may be the most powerful social media network of its kind, and people may find it very convenient to use Facebook because so many other people are on it. But that does not mean, as one chap said in my comments, that people do not have a choice about using Facebook. And as history has shown, over and over, it certainly doesn’t mean that people won’t choose to use something else in the future.

  1. The reports are not quite directly comparable, so here’s how I derived this number. According to Pew, 94 percent of “teen Twitter or SNS [social networking site] users” reported having a Facebook account in 2013. Eighty-one percent of teens surveyed reported using a social networking service like Facebook. Ninety-four percent of 81 percent is 76.1 percent. Now, in 2013, 24 percent of teens also reported using Twitter. I assume that most of them also had SNS accounts. But if I’m incorrect, that would only mean that the decline from 2013 to now had been even larger, so I am using the smaller figure as the most conservative estimate.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Megan McArdle at mmcardle3@bloomberg.net

To contact the editor responsible for this story:
James Gibney at jgibney5@bloomberg.net