User-generated content is a contentious topic that puts developers, publishers, investors and ultimately users at odds with each-other. Investors and publishers often react with frank disbelief at the idea of allowing users to create their own content, but setups like Facebook and Second Life enjoy considerable commercial success.
Facebook's 2008 revenue was estimated to be somewhat in excess of US$200 million for 2008, while Second Life is estimated to have achieved somewhere between 100-200 million USD in the same period. User-generated content (UGC) can be highly profitable, but whether getting into those profitable situations, and moreover sustaining them is a whole other issue.
Numerous developers have traditionally been sitting on the fence about UGC, waiting to see if those platforms that feature it can ultimately bring the issues under control, and stand by ready to emulate those solutions. Few of them are willing to be laughed out of an investor's office by suggesting UGC as strategy at this stage of the game.
Supporting UGC from a technical standpoint is complex, intricate and expensive. It also raises the bar for hardware and networking for the actual customers, reducing the pool of potential clientele. UGC is a lot harder on the end-user's system and network than pre-generated content.
And that's before we get into exactly what people are creating. Given the ability to create largely at will, people create all manner of things.
At this stage, nobody knows where a workable balance is between allowing users to create whatever they please, whenever they please, and supervised, editorially vetted content. Nobody has yet found a balance that appears to be both sustainable and persistent over the long term.
Linden lab has, for the last few years, been shifting its position on UGC (and related activities) in Second Life tightening up rules, and either explicitly disallowing, or segregating various kinds of UGC.
EA-Land instead introducted UGC as a part of its rebranding and replacing The Sims Online, in an editorially-controlled form, but within a few weeks of the change the replacement service itself shut down.
Unfortunately, UGC works to its own schedule, is unpredictable in direction, and – as yet – there isn't a sufficient body of data and experience to reliably generate successful business models that incorporate it as a primary feature. At this stage, successful models seem to be more of a matter of serendipity rather than of design. Success or failure appears to be largely decoupled from the actions of developers/operators.
What is quite clear so far, is that if you're going to be successful on a UGC-based model, that's going to take time. UGC is accretive, and models built on it will likely take years just to break-even, and that's a difficult position to find funding for.
