Social Networks have been around since the inception of bulletin board systems in the mid 1980s, and each one of them seems, for a time at least, to be the radical new paradigm that establishes how people will interact with one another over the web. Certainly, this seems to be the case to those investors (whether individual or corporate) who pay surprisingly stiff premiums in order to be a part of the next big wave, yet in truth social networking sites have a surprisingly consistent "life-cycle" that seems to play out regardless of the "angle" that the sites have.
I've seen (and used) the analogy of club-hopping to describe such sites, but a better metaphor perhaps is the development of urban centers. As with urban centers, the driving factors are less "social" - who the cool people are - as it is resource based, where the resources in question are money, opportunity and signal strength (how strong is an individual's signal in the noise of the community).
This life cycle almost invariably follows a clearly defined arc:
1) Seeding - people, dissatisfied with a given networking site, create a new "space" designed for targeting a specific niche community.
2) The pioneers arrive into the new space, often brought in by a friend of a friend.
3) This process continues until resource pressures reach a critical point, at which point the space has to start acting more like a business (which in turn imposes an overhead which requires more people to feed the mill).
4) The site experiences rapid growth, but at the expense of losing the sense of community that had brought the early adopters in. At this stage, seeders and pioneers begin to defect to other sites.
5) The site can continue as in #4 for some time, but as it does, it begins to seek ever more intrusive forms of renumeration (taxes or advertising) on the members. This attracts outside money, which of course is seeking its own replication.
5a) Many social networking companies are typically bought out at this point, often by companies that have SN sites that are at #6 and are hoping to "revitalize" their offerings. The fresh infusion of capital usually sustains an SN site for a while, but it won't make it any younger.
7) The site eventually dies, suffocated in its own waste. Its brand name becomes toxic. The company built around the site in turn usually morphs into some form of services company, often of a fairly dubious nature (such as creating virids).
Sites are generally most "interesting" as they transition from #2 to #3 - the focus is usually fairly tight, the people that are part of the community are knowledgeable and helpful, and the opportunity to build reputation in that community is fairly high. At some point, however, the benefits brought about by an increasing community size are outweighed by the resource costs associated with that size, which in turn forces the managers of the site to think not like salon hosts but as business managers.
Indeed, another analogy that might work here is that most such communities in stages #1 to #3 are generally like private equity companies - they are typically smaller, but far more focuses. Once you open up the share structure of a company to become public, the imperative to make money just to support the rising cost of the company's bureaucracy and infrastructure can often outweigh the benefits to the initial participants in that company.
The irony in this, of course, is that by the time a social networking site needs significant infusions of cash, they've already passed the point where they are attractive to the early adopters - the risk takers. In essence, what investors are paying for there is ad revenue that comes from a rising customer base. Note that in very few cases does the actual focus of the social networking site come into play. This is why it's so easy to start social networking sites (and why its incredibly risky to invest at #1 or #2, precisely because of that very low barrier to entry).
The virid factor plays a big part in the demise of social networking sites. A virid is a virtual entity, typically some form of mindless spam bot, though AI bots are becoming more prominent. You also have "hooker" virids, which are actually flesh and blood people, but who are nonetheless paid a certain amount to entice unsuspecting customers to commercial porn, gambling, or high pressure sales sites. All of these are essentially parasitical in nature - they serve primarily to leech away the ad potential of a site's members.
When the parasite load reaches a certain critical mass, the site will implode - membership will drop precipitously, what ad revenues existed dry up, and the only people visiting such sites are those following older links or the very late adopters. In many cases, this is fatal to the company doing the hosting of the site as well, unless they shift their focus into becoming virid producers themselves.
One of the critical differences between an urban environment and a social networking site, however, is that prime real estate is a limited resource. Gentrification occurs in flesh and blood cities because the cost involved in tearing down a building and putting up something new in its place is generally lower than trying to develop a community far from a cosmopolitan center - something that commuters paying top dollar for gas prices to commute from far flung exurbias are now coming to realize.
However, in a virtual world, the cost of building a new brand is usually considerably lower than the cost of trying to revitalize an existing one - in the long term. Revitalization (gentrification) in the real world may move you back to step #1 or #2, but online it is likely at best to hold you where you currently are for a little while longer.
Thus, social network companies should be seen at best as short term investments - they may pay nice dividends over the course of a decade, but they will eventually play out, and in time can even become a liability, as companies that move to #7 may find themselves under intense regulatory scrutiny. This doesn't mean that they aren't good short term investments - many can be quite profitable for several years (both in terms of money and in terms of the value received back from the community) - only that they should be seen as transitory enterprises at best.
Kurt Cagle is an Online Editor for O'Reilly News