The Internet economy has been built on the network effect (i.e. the effect that one user of a good or service has on the value of that product for other people). Investors and entrepreneurs have treated this like Moore's Law. But just as Moore's Law hits physical constraints, network effects have a limit in many types of online communities. Indeed, in some cases, a reverse network effect may exist: as new people join, others are motivated to leave. This dramatically affects the length of the competitive advantage enjoyed by these ventures. In this post, we'll look at which ventures suffer from reverse network effects, which don't, and which may suffer depending on the strategy they choose to adopt.

Signs of Reverse Network Effect

Real human social networks (as opposed to the tools that facilitate online communication between people in a social network) have an obvious reverse network effect. If I have too many people in my social network, I cannot pay enough attention to each of them, and without attention, relationships fade.

Back in September 2007, we looked at social motivation in the real world, specifically the two distinct types of motivation. One is, "I want to communicate better with the people I already know and trust". The other is, "I want to increase my visibility so that I can connect with more people."

Both have clear limits. The number of people I can really know and trust is limited, because knowing and trusting take time and attention. Increasing visibility, whether by blogging or tweeting or advertising or PR, is less limited. But when visibility goes beyond a certain number of people, it becomes no more social than broadcast media or spamming. The personal touch is gone. The real community spirit is gone.

As Groucho Marx remarked, "I refuse to join any club that would have me as a member." If it is an exclusive club, the membership has to be limited. If it is not exclusive, is it really "social"?

eBay is the classic example of a network that lost its community spirit after reaching a certain scale, as the comments on this post vehemently demonstrate. This loss opened the market for eBay alternatives, such as Bonanzle. It is unclear whether these alternatives will face the same issues when they get to scale.

Etsy will be an interesting case study. It looks like it may avoid the reverse network effect by being focused entirely on handmade goods. That focus may limit its eventual scale, as not everybody wants to make or buy handmade goods, but it will still be able to build a business more than big enough to satisfy even the most demanding dreams of avarice. That focus on handmade goods will act as its social glue. It will grow as eBay would have had it decided to remain focused on the global garage sale, not getting lured into selling mass consumer goods. I suspect if eBay had done that, its core business today would be smaller but ultimately more profitable and sustainable.

Networks Without Messy Human Interaction Are Immune

eBay the corporation owns two businesses that have almost perfect network effects and do not suffer from any reverse network effects: PayPal and Skype.

PayPal works for consumers because merchants use it, and merchants like it because consumers like it. The fact that everybody likes it won't make me like it any less.

Skype gets more useful with each new user, and each new user promotes Skype, consciously or unconsciously, for his or her own reasons. Even better, the cost of providing the service goes down with each new user, and that is really unusual (a function of Skype's P2P architecture). Google and PayPal also benefit from each new user, but they still have to service that user, and that costs money. In the case of a video service such as YouTube, the servicing cost is significant. So Skype really is in a league of its own when it comes to network effects, and that is why it may become the world's largest telephone company and the biggest economic success story of the Web 2.0 era. (Google Voice, having just thrown its hat in the ring to battle Skype, will be interesting to watch. My bet is on Skype.)

The difference, though, is that we do not look at these services as communities. They are simply enablers of commodity transactions: payments and phone calls.

What About Social Networks?

Does Facebook still feel special now that it has 175 million users and is growing at a rate of 600,000 per day? That is a sincere question for Facebook users. I am one of the few people who do not use Facebook. For whatever reason, I never got into the habit. And I have no compelling reason to start now. This has made me a bear on Facebook for a long time. As a non-user, I miss what makes Facebook special and why it grows so fast.

The network that I use is LinkedIn. I use it to connect to people who my contacts know. As far as I am concerned, it is a utility in the same way that the phone or my email/CRM system is a utility. Do I care that LinkedIn has 7.7 million users now? Does that make it more valuable to me? No. Is Facebook 22 times more valuable to me than LinkedIn because it has 22 times more users? No.

Here is the theory:

In a social network, the value for existing users of a new user joining the network plateaus once users have most of their own contacts in that network.

Of course, the social network grows in value as a business as more users come into it because the network then has more eyeballs to sell to advertisers. But that is different from network effects. As a LinkedIn user, I do not benefit when more strangers join. I already have about 90% of my contacts in there, and the remaining 10% may remain hold-outs; and I don't really care anyway, because I can always reach them outside of LinkedIn. When I meet somebody in business, I look them up on LinkedIn and connect if they are on it (I cannot recall anybody recently who was not on it). I do that to keep my Rolodex updated, which is a very valuable utility for me. But my actions are not growing the network. If I joined a new network, I would spam all my contacts asking them to join, but I have zero motivation to do that.

As a long-time Facebook user, do you care when more strangers join? Does it make any difference to the value you get from Facebook?

So, it is possible that the network effect has a natural plateau.

A Plateau Is Okay, But a Steep Downward Slope?

Whether that plateau turns into a downward slope depends on the monetization strategy adopted by the "owners" of the social network. And in the social networking business, a downward slope can very quickly become a steep slope or even a cliff. Trust tends to be binary; viral can work both ways; and switching costs are minimal. That is why I put "owners" in quotation marks. You can never really own a social network or community. You can provide services and extract rent for only as long as the community wants those services and is willing to pay the rent.

This is where the incredibly high valuations of businesses such as Facebook and LinkedIn may become problematic. If these businesses get too eager to monetize to justify those valuations, they may create the reverse network effect.

There are only two ways to monetize: ask for revenue from users who regard the service as a utility (like paying for the phone company), or ask for revenue from vendors that want to sell something to the people using the service. Thus far, all the major social networks have taken the latter fork in the road. They don't want to become utilities because that wouldn't justify their lofty valuations. So they have to sell more to those who use the service. At that point, the reverse network effects may kick in.

Craigslist

Craigslist chose a different path by not taking on external investors. It has no valuation to justify. It can leave masses of money on the table without any worries. So Craigslist won't suffer the reverse network effects that come from over-eager monetization. Its model of allowing lots of free listings will sustain high growth and is clearly impacting eBay's business.

Twitter an Interesting Case Study

Once again, Twitter is the interesting one to watch. The ease with which one can add and delete who one follows makes its size self-regulating. As a real-time search tool, its value goes up with each new user. As a communication tool, it goes up as new people join who might be interesting to follow. Its openness may prevent the reverse network effect.

The other reason Twitter is an interesting case study is that it has not yet disclosed its revenue model. If the revenue model it does adopt involves selling to its users, the reverse network effect may kick in. Twitter would become classic MBA case study material, a fact of which management must be well aware!