On the same day, I got hold of two bits of news from the metaverse front: first, Blue Mars announced their pricing. Oh, all right, I know I’m behind the news… the point is that this relaunched the Second Life® vs. Blue Mars debate once again. I started following the SLED (Educators in SL) mailing list and how the academic community was not impressed — neither with the pricing, but much less about the requirements to run Blue Mars.
Then, several hours later, another bomb dropped in my mailbox. Raph Koster’s brainchild, Metaplace, went out of business, terminating operations on January 1, 2010.
If you remember late 2007/early 2008, you’ll remember what every “virtual world expert” was saying: the Metaverse would happen inside the Web browser. The future was supposed to be embeddable 3D VW applications, hopefully inside Facebook. Kids would be using virtual worlds like crazy, and a huge growth in that area was predicted. While nobody was really claiming that Second Life would disappear overnight, they were expecting a big shift towards completely different audiences and technologies, namely, Flash.
And people believed them. The Electric Sheep Company moved out of Second Life and started doing their own Flash-based virtual world — and made a few sales (yes, one was for a company wishing to embed it on Facebook). Raph Koster gathered venture capital and made a big splash about his notion of a virtual world with 16-pixel-high avatars on isometric landscapes which would become the next revolution in user-generated content — specially because you could use SketchUp for building things, or even just copy 3D objects from Google Warehouse and immediately upload them to Metaplace and start having fun. And of course they were not the only ones. A bit all over the place, companies and opinion-makers started telling us how everything we thought about the “metaverse” was wrong, and that concepts like Second Life/OpenSim were simply outdated and had to evolve or disappear.
At this point I did only gently cough and said, “follow the money”.
Metaplace’s closing down does not come as a surprise to me, but for totally different reasons. While I recognise that the funding method through venture capital is a good way to develop innovative technology, I still maintain that there is a huge gap between “academic research” and “successful business”, and not everybody is able to leap that gap.
Academic research brought us things like… OpenCroquet (now called OpenCobalt). Which is not dead yet, but it’s a telltale sign that one of their major developers, Duke University, has fully embraced Second Life and OpenSim — although they still allow Julian Lombardi to place a nice link to it under his personal homepage. It’s a nice technology, when you can get it working at all, of course. But it’s hard to put a valid business model behind it. It’s great for the research lab. But as a business model?…
There are thousands of “virtual world construction kits” out there on SourceForge and similar websites. Most are tiny one-person-projects. Many are research tools with solid development behind it. None have impressed us much 🙂 … except occasionally for some nice graphics here and there or a paper or two showing how cool an obscure renderer was implemented.
Koster’s Metaplace had the advantage of at least having a solid technological platform — for the expectations generated in late 2007/early 2008. As for its business model… it was not clear. Everything was to be given away for free, except for some obscure licensing schemes, which were never the strong points. This is not surprising — almost all hi-tech virtual world start-ups keep giving everything away for free, and then suddenly realise they don’t have a working business model. VastPark, who had a cool 3D rendering engine, was perhaps one of the latest massive failures (they now released all the code as open source and closed shop). I’m sure that Metaplace’s ultimate failure happened for the same reason: once the venture capital was exhausted, the question was — how many sales did Metaplace actually make?
If you find it so harsh to judge technology by its failure to sustain a working business model, think again. We are supposed to be cleverer after the dot-com bubble burst! But we’re not. We’re still selling “ideas” and believing they’re valuable. They’re not; anyone can have good ideas. A few can implement them. But only a very, very tiny group of people can actually make them profitable. And it’s just those ideas that actually will survive on the tough marketplace.
It’s not about the “financial crisis”. If I were in a very sarcastic mood, I’d say that a solid business model, that is not based on sheer marketing but of actually fulfilling a real need, will survive any financial crisis, so long as the company behind it is not wasting money on unnecessary things. But that’s oversimplifying the issue, of course. Nevertheless, a good hint on whether a technology will survive or not is following the money.
Gwyn’s Golden Rule of Metaverse Survival: Who Makes Gold, Will Survive
My first instinct is always checking if I can understand the business model of a technological company. If it’s clearly stated on their site, and they can explain how they attract customers and sell them technology, there are good reasons to believe they’re going to succeed. If they have diminishing costs as the number of customers increase, they’ll be very successful at some stage (see eBay and Amazon). If their costs rise linearly over time as they get more customers, they will manage to survive (see Linden Lab) and might even self-sustain themselves when reaching a plateau (There.com is a good example). In fact, during a financial crisis, an “elastic” company that can survive without exponential growth is surely going to be around for a long time (Blizzard might be one of the best examples; they don’t need to duplicate or triplicate their number of World of Warcraft players to continue to shell out a billion US$ per year of profits).
But if you can’t figure out how they make money — they’re doomed to disappear.
Consider Twitter vs. Facebook. Twitter’s costs in infrastructure just grow and grow, as more people send more and more tweets through their gateways. Over half of the tweets never reach their sites — they just use the many (free) APIs to deliver messages among users on third-party sites. Twitter has no ads. They don’t visibly sell any service. They don’t report income. We all “know” that they sell profiling data, but… to whom? And for how much? We don’t know. So why are they still around? How long will their venture capital last? How will they turn Twitter into a profitable service? If you have any clue, please leave a comment — I’d love to know the answer 🙂
Facebook is an interesting example. They started just like Twitter — everything was for free. But although Facebook has several APIs as well, it’s far more complex to manage your account without ever going to their website. So this means that at least they had something solid: a webpage with lots of traffic. What do you do with lots of traffic? You sell ads (and, of course, profiling data; thus their draconian terms of service demanding “real people with real names”). This is what Microsoft told them when they bought a share, in exchange for using Microsoft’s ad-serving software on Facebook. This managed Facebook to actually start making some money out of their technology. We all understand web-based ads: we just need to look at Google to see how you create a huge technology giant out of that (while ironically, during the dot-com days, people thought that web-based ads were doomed to disappear… Google just laughed and bought the competition and established themselves as a quasi-monopoly on ad serving 🙂 ).
But Facebook is also changing… once touted as the social networking site that would replace all other social networking sites, thanks to their innovative technology — embedded applications from third parties! — what are people doing now? Facebook’s timelines are too complex to follow (like Twitter). You need hours and hours to keep up with your timeline. Unless you’ve got a dream job with nothing to do (and I was shocked to see how many of my close friends actually do not do anything at all the whole day and just spend their time at work tweeting and sharing links on Facebook!), it’s impossible to keep up with everything! So what do people do on an information-saturated Facebook? They play games. More precisely, they play Farmville. Farmville is saving Facebook from becoming irrelevant. But it’s not exactly a “new” idea; Flash-based gaming is a few years old, and it has always been a successful business.
My question is only if Facebook will become irrelevant for “serious business” once the enterprise world discovers that people don’t read any links any more, they don’t watch ads… they just play Farmville. Very aggressively. And perhaps they will start buying more ads on Farmville instead. I wonder what that will do to Facebook’s income. As a platform to disseminate Flash-based games, Facebook couldn’t have worked better — but was that really its ultimate purpose? Or perhaps I ought to ask if it matters at all? If they can make a profit somewhere, they ought to survive, even if they rename the company “Gamebook”.
Perhaps my bias against all these social networking thingies is just rising to the surface. I sigh as I log in to Google Wave for the last time. I had such high expectations! I wanted a social networking technology that allowed me to get good information, filtered from hundreds of sources, and trustable. I don’t need gazillions of messages coming in from hundreds of services telling me to click on links to read blogs, to watch movies, or to play games. I have no time for them!
Instead, I just wish to talk to people. Not any people, but the kind of people that know what I like and that share common ideas and tastes with me. At some point in my life this used to be a favourite IRC channel; later, perhaps a forum; in some extreme cases, blogs with comments. Twitter and Plurk were fun when I had 20-30 “followers” whom I knew well. Now I get IncognitoJoe telling me how catching butterflies in Costa Rica is good for his health. How could that possibly even remotely interest me?
No, ironically, the best “social networking tool” I have right now is to log in to SL, jump in the middle of a conversation, and ask my friends about the news. Even the clumsy, limited, ugly, and unreliable Group IM Chat provides me with more information in half an hour than Facebook with a million messages with shared links and comments on links, and “likes” on comments to links…
So, well, back to virtual worlds, where people talk to people, and although spam exists and is growing, it “feels” manageable in some odd way. Here the question is, what kind of virtual world is good for me?
When you get so used to a hammer as a tool to hit on nails, everything will look like a nail, even a screw… and SL has that feeling. It’s hard to imagine a “different SL”, so it immediately becomes the frame of reference for everything. I’ve long since explained that its success comes from user-generated content, persistence, and visual contiguity. Getting all three is not easy, and most technological platforms drop at least one of the three. In general, from the users’ perspective, user-generated content seems to be the more interesting one. VastPark, Metaplace, Multiverse, Blue Mars, even IMVU to an extent, all put a strong focus on a certain degree of user-generated content.
User-generated content is a bliss and a nightmare. There are two extremes that are easy to implement. The first is where you simply upload content to it and don’t care what happens next. The virtual world platform simply becomes a repository for public domain content. All content is free; any content can be grabbed, tweaked, changed, replaced. Notions of “intellectual property” are irrelevant.
This is an approach driven by two very strong groups. The first group — by far the largest and the most reputable — comes from the academic research network. In almost all cases, universities and research labs are underfunded, so the more free content there is, the happier they are with a particular technology. This was something that was not clear to me until very recently — I have to admit being quite naive sometimes. At a recent conference that I attended about virtual worlds, one speaker was showing a quite interesting 3D engine based on Flash (or was it Java? I forget…), and one of the key features of it was the ability to upload standard .OBJ files. The speaker deliberately showed how “his” platform was totally different from Second Life because it allowed any .OBJ to be easily uploaded (he didn’t realise that meshes are around the corner for SL, too), and this was great news for academic researchers, since it meant they would have access to a ton of content very easily.
I was a bit baffled about those comments. I argued that SL had 2.3 billion items, and 250,000 new ones were added every day; surely that was plenty to chose from? Also, gluing prims in SL might not be easy for an absolute beginner, but certainly it was way easier than learning Maya or Blender?
The speaker patiently explained that this might be right, but not all those items are free content. Whereas all over the net you can find free 3D content in standard formats, ready to be uploaded. 3D modelers are constantly dumping their meshes on Google Warehouse and similar sites, with millions and millions of objects, where you pretty much can choose whatever you need. Thus, you don’t really need to “model” anything: just browse the web and upload your meshes. Simple and cheap!
Now I was not baffled any longer; I was utterly flabbergasted! What this person was saying is that the Web is full of cheap content, most of it pirated (yes, I’ve checked some of those “3D model repositories”!), and that a “great” virtual world would be one allowing these meshes to simply be grabbed and uploaded. Researchers, he claimed, have no time to learn complex tools (true) and have to rely on public domain content for their work. Well… I understood the argument, and was at a loss to discuss it. What can beat free content, specially if you don’t have any qualms about the origin of that content? Those repositories are so huge that it’s impossible to file DMCA claims against them, even if you managed to find one of your meshes uploaded there “by accident”. Also, people like Stroker Serpentine were bold enough to file lawsuits against Linden Lab — but would they do the same against the Google giant? I guess not…
So, for a moment, I was scared. These people, with decades of experience in their field of studies, were actually telling me that “meshes are good because they’re free”. And suddenly the whole drama of “prims vs. meshes” was turned into a discussion about “paid content vs. free content”, which totally caught me unprepared!
The second group are ideologists, often programmers born with a Richard Stallman poster by their crib. They strongly believe that information should be free, and since content is nothing more than information, it should be free as well. In a world where everything is for free, everybody is happily sharing, and there is no “need” to bother with paying for anything. This group is smaller than the first, but they’re deeply interconnected; many universities, due to lack of funding, encourage their students to grab whatever free technologies are available to cut the costs (one day I should address this common fallacy; universities are effectively trading licensing costs for labour costs, which is cheap and readily available on the campus; but I’ll leave that for another day 🙂 ). This shapes their mindset to “believe” that “all good things out there are free (or ought to be free)”. While to a certain degree this is actually not far from the truth — I’m using one of the best blogging software in the world, which is free, running on one of the best database engines, also free, on top of one of the best operating systems, definitely free — it’s too easy to overlook that all these “best things” are available for “free” (in the sense of “not paying for a license”) because the companies behind it actually make money from services, not from licenses. A typical example is tech support calls or maintenance agreements, which cost orders of magnitudes more than a “software license” 🙂 Alas, this is too often overlooked by the ideologists…
Then there is the other extreme. Some companies actually understand the problem (legal and otherwise) of allowing anyone to grab pirated meshes out of a repository and upload it to their virtual world, so the alternative is screening content. This is what There.com (and IMVU) do, and Blue Mars is going the same route. The idea is that you just trust a limited number of designers and carefully analyse what they’re actually uploading. Now what this means is that there is a manual, human-based process involved in the content creation: you have to submit to censorship, and although the concept is encouraging for the ones worried about content theft, it also means a lot of hassle.
But we’ve seen what the first model lead to: Metaplace, relying on “free unrestricted mesh uploads” (you could upload directly from Google Warehouse!), and no solid business model, ceased to exist. I hope that there is a lesson to be learned from that.
Second Life, of course, lies in the middle of both extremes. There is no “hassle” in uploading content; content is actually built in-world using the permission system, and even though we all know the risks of illegitimate content copy, the truth is that intellectual propriety rights are preserved to a degree — and a new innovation by Linden Lab will soon allow a step further. So while the concept has known flaws, the spirit behind the intellectual property protection in SL is actually correct. Protected content leads to a content economy — something that possibly only Lessig might have foreseen in 2003, but few people still understand how vital that is. And Second Life’s model doesn’t require a priori “screening” of content.
It is said that the best kind of business is the one where both parties feel they have made a good deal. My best example is, of course, eBay. Ebayers are happy customers of the service, because it allows them to make money, too. They’re happy if eBay is making cartloads of money — because they’re making money too. The happier eBayers are, the more they will buy and sell, and the more commissions will eBay charge. That model is fundamentally solid, because eBayers will continue to come back to eBay to make business.
Second Life is the eBay of virtual worlds. In a sense, while SL at some point was sort of used as a “reference” for virtual worlds, which would try to emulate or copy some of its technological features, now Second Life seems to be too odd compared to all the rest. Nobody seems to understand what the business model is (or why it is so fundamentally important)! Every other week I get someone blaming LL’s prices, calling them “greedy” or “extortionists”. In their mind, LL should be just providing free server space and free content for everyone to enjoy — like, say, Metaplace. They don’t understand that LL provides a service, and charges for that service: the ability to participate in a somewhat regulated marketplace. This is what we get when we pay tier: the ability to be in a virtual world where property has value; where you can buy and sell content; where you get a reasonable amount of protection of your intellectual property; where you get little interference in what you can do and build in the virtual world. It’s obvious you can get some of those things elsewhere; but there is nowhere else to go to get it all.
If I had written this in, say, 2002, one might have raised several objections to this model. While in 2002 we had eBay as a model showing how a protected (but not too regulated) marketplace is a good and solid business idea, nobody really knew if the same would work with 3D digital content. After all, for 2D digital content, things are dubious (because the Web doesn’t implement intellectual property!). Why should 3D digital content be different? Well, it would be a huge shot in the dark “believing” that these concepts could actually be translated into a solid business model. But at least you could argue around the business model: what does Linden Lab sell? Server space. Why do people lease server space from LL? Because they want persistence of their content. Why do they need it? Persistent content with intellectual property rights attached to it can be transacted — bought, sold, given away. The more persistent content you have, the more likely you’ll need to lease more space (think shops!). So there is a direct tie between the economy growth and the growth of Second Life. And we’re just talking about one single aspect of SL: its content marketplace. What about the entertainment value? That adds a completely new (and unexpected, at least for 2002!) business dimension to the whole concept of the “virtual marketplace”… where an even more abstract concept, “entertainment”, is bought and sold.
Although I’m not a big fan of Snowcrash at least I will admit that Neal Stephenson did predict some things quite correctly. He thought that the two biggest sources of transactions on the Metaverse would be “programming” (in his book, creating 3D models, either of buildings or even avatars, was labelled as “programming”, so it encompasses the whole range of digital content creation in SL) and “entertainment” (namely — guess what? — live music!). He got that very, very right from the start. It might have been pure coincidence or just a very good analysis of what virtual worlds would be used for.
So when looking at all other virtual world platforms, I ask myself what their business model is, or, at least, what they think their business model is going to be. The socio-economic dynamics of Second Life are very complex. It’s not easy to understand that the next newbie logging in to SL might be an amateur DJ, entertaining friends on webcam chats (a popular pastime where music sometimes is “performed” in front of a tiny audience). They come in to SL and see there is an opportunity not only to entertain their friends here, but also get some money from tips. But then they figure out they need a streaming server; a lot of in-world operators provide that service for a few L$. Next comes some cool animations and a trendy, fashionable outfit. Then you have to hire an agent. Establish a group. Go shopping for devices, clothes, props. Start going to venues — which in turn will hire builders, programmers, and lease land. Communities start to look for entertainment, and hire DJs and musicians. RL companies do the same. But it’s all interconnected; SL is not “just about” music performances, as it isn’t “just about” building pretty things, or designing fashionable avatar clothing. It’s about a million things happening at the same time, people with a wide range of interests, all cross-connecting, all somehow contributing towards the economy (even if they never spend a single L$ in SL and give away all their content for free) and the society (even if they never join group chats or specifically go to a venue to attend an event).
This grew organically over the years, but it’s important to note that it didn’t happen in a vacuum. Its foundations were a solid business model of the company maintaining Second Life. Oh, sure, I know that a lot happened by coincidence and very little by design — nevertheless, the ability to sustain this complex virtual ecosystem is Second Life’s triumph, and it’s next-to-impossible to replicate.
Blue Mars has little of the above ideas in mind when it does their announcements. You can discuss on their forums and see the strong focus on the technology. I’ve read the announcement of “Caledonia”, Desmond’s ‘city’ in Blue Mars. It’s even supposed to have a larger area than Caledon in SL. While I’m sure that a lot of people with state-of-the-art PCs will enjoy moving over to Blue Mars to see what it has to offer, my only question was — who will build all that content? How much will it cost? How will Desmond make a return on his millionaire investment on Blue Mars?
Now I have at least one partial answer to that: most of the content will simply come from the plethora of free 3D mesh repositories out there and tweaked to look “Victorian” enough to please Desmond’s future customers.
One can only wonder what will happen when LL introduces their own mesh uploading features in SL. One thing is for certain: unlike all other virtual worlds allowing any pirated mesh to be copied and uploaded, LL will provide actually quite reasonable intellectual property protection on meshes as well. And then the question will remain: between a digital content marketplace with a working business model, where content transactions have been proved to work reasonably well, and a new “content-is-free-because-we-can-copy-it” model, what will ultimately survive? Looking at Metaplace’s failure, and OpenSim’s difficulty in growing much more in spite of all the free content there, I still think that Second Life is our best choice to establish the foundations of the upcoming Metaverse.