Competition to Second Life®

At the last Town Hall meeting, where Philip Linden explained to us how Linden Lab will “facilitate” the selling of L$ for US$ in Second Life®, he mentioned how worried he is about competitors catching up with SL.

As a matter of fact, he mentioned the word “competitors” several times. This is the first time I attended a Town Hall meeting where Philip justified this action with […]this is the sort of things that if we didn’t do, a competitor of SL would.

For those that don’t understand the mechanism, it works like this: Linden Lab is quite open to having their users (“residents”) doing all “sort of things” that other companies strictly forbid. So, the copyright to your content in Second Life is yours, guaranteed by the Terms of Service — unlike other companies, whose worlds retain all copyrights (you are basically creating for them). Second Life also allows third-party sites to integrate with their world — actually, it encourages them. Instead of filing lawsuits against people who are creating “external” content and bring it in into Second Life, just like every other MMORPG (they consider that they have an exclusive monopoly on content), Linden Lab actually promotes these third-party sites actively. And finally, instead of suing people who are selling Linden dollars at eBay, Linden Lab has worked closely with some companies to set up a system where you can exchange money in a free market.

Currently, at least three sites offer these facilities: IGE, Anshe Chung, and the e-Commerce site SL Exchange, which also sells L$ for US$.

Gaming Open Market sadly is shutting down their two-year kingdom as the undisputed ruler of the money exchanges. It had a different approach — it was an exchange market, where users would set the price — thus, no wonder people felt that it has the most “reasonable” ratio and the one more correctly representing the “real value” of the L$. It also showed the trend of the market and reflects the fears and the optimism of the users — like any other exchange, it reflected the psychological market. We’ll have to see if the Linden-operated in-world exchange will step in GOM’s shoes and is able to replicate its role in SL’s economy.

But this is apparently just the tip of the iceberg. There were always people exchanging money and accounts at eBay for all sorts of MMORPGs. And even before eBay existed, people had their own ways of selling/reselling money and accounts. It’s not this aspect that makes “competition” threatening.

The point is, there are many, many aspects which may threaten Linden Lab’s continuing success — and the virtual economy is just one of them.

What are SL’s threats?

Let’s do a bit of SWOT analysis here. For the ones unfamiliar with MBA-gobbledygook, SWOT analysis is just a term to describe an operation’s Strengths/Weaknesses (internal), Opportunities/Threats (external). This is usually a fun and creative way to approach the way a company/entity is being run.

Far from me to claim to be an expert on these areas, but let’s see what we can come up with:

Strengths

• Motivated team — Linden Lab employees work extra hours, do billions of things at the same time.
• Expert team — from their resumés/CVs you can see that you have top-of-the-line experts working at Linden Lab right now.
• Willingness to experiment — unlike other companies, LL often introduces radical ideas, even if their benefits are not immediately visible. They try to draw a balance between “features” and “bug fixing”, even if some features seem to be worthless. But the drive to innovate is present.
• Despite many claims to the contrary, it has an amazing technical support team.
• Also despite many claims to the contrary, LL is probably the only platform so far that was designed from scratch with an emphasis on scalability (MMORPGs tend to add further “shards” or “new worlds”; SL has a single “world” which can, in theory, accommodate an infinite number of users and was always planned that way. Some competitors claim better scalability, but that is unproven so far).
• It’s the only virtual world with collaborative, dynamic content, with such a large user base — since the modelling tools are designed for amateurs (other worlds are either dynamic, or simply static but collaborative, but in the “static” case, you need to work with 3rd party, very complex modelling tools).
• Copyrights are held by the users creating content, not by Linden Lab. The latest versions of the Terms of Service allow LL to regulate this aspect further. LL also fully complies with the DMCA to provide further protection and a means to deal with copyright violations.
• Virtual economy — it’s stable, it works, it has checks and controls, it is successful, it allows people to make some (real) money with it. This is encouraged by LL in total contrast to most other platforms, where only “official” items can be created/bought/sold.
• Social aspects are even encouraged and promoted by a group of people at Linden Lab (known as the “Community Group”) that are employed just to support the emerging society. Overall, most Linden employees are often seen in-world, and they are active participants in conversations with the users (or host events for them), encouraging a stronger and closer relationship between LL and their users.
• Separate grids for “adult” and “teenager” content (as well as private islands which only you can use).
• Growth has increased to about 500 new users per day (thus, probably Philip’s claim of “1 million users by the end of 2007” will not be met, but it won’t be too far from that).
• Board and staff are generally open-minded to open source projects (internally they rely mostly on open-source tools) and integration APIs. They view their platform as “user-expandable” and not necessarily as a monopoly on the technology they wish to secure for as long as possible. The platform supports “minority” operating systems as well (Mac OS X, soon Linux).
• LL is partially VC-funded by Pierre Omydar, the founder of e-Bay (and definitely another great business visionary of the Internet!).
• Overall, the strongest points of LL seem to be the strong links to their community (beyond technical support), and, naturally, R&D.

Opportunities

• Firstcomers to the Metaverse model — Nobody else is really building a similar model yet. So, Second Life is the reference that others will need to copy.
• Grasped quickly the attention of the academic world (7 college classes this semester!), the education/teaching/training world (c-Learning or cyberlearning is the next buzzword to watch) and the Big Corps (like Wells Fargo).
• Draws interesting comments and remarks from the media, which cannot “fit” them very well on the usual MMORPG model and thus most often than not make positive reviews on their platform.
• For a tiny company with slow growth, they have a “global” vision, incorporating telecommuters all over the world in their staff (ie. they embraced quickly enough the need for German, Japanese and Korean content and tech support).
• Broadband bandwidth and sufficiently fast computers to run SL are starting to be the norm, worldwide.
• The 3D content market has hit a ceiling (ie. professional 3D designers that sell their 3D models and creations through websites like Renderosity) in terms of potential users and needs a new market (so, more talent is pouring in into SL).
• SL is free nowadays, an obstacle that many pointed out to be a barrier for new users (the number of new users went up five times per day since that announcement).
• SL is now attracting less “niche” users (ie. programmers and designers, who are still the large majority) and more “mainstream” users. Many of them were never interested in “games” (ie. MMORPGs) and put SL to completely different uses than it was originally planned. This is probably the fastest-growing market for LL in the near future.
• Artists worldwide are getting funding, grants, and sponsoring to embrace “digital art”, and more specifically, “3D digital art” — thus encouraging the artistic milieu to explore SL.

So the future looks definitely promising! To be honest, let’s see some of the major drawbacks:

Weaknesses

• LL is not making much profit (if at all). The business model simply does not allow for many benefits due to growth (more users mean more servers and more bandwidth and more technical support — this means more income, not more profit, since the costs rise in linear proportion to the growth).
• Complete lack of priorities (no roadmaps, no project management, no quality assurance — the vacancy for the Quality Assurance Manager job offer, posted last year, is still open).
• LL relies too much on users to tell them what to do (worse than that, it relies on a surprisingly small number of users to have their saying, without any concern on their background and credentials to emit proper opinions).
• Nobody at LL knows what they’re actually selling (see my previous article on this subject).
• They are still figuring out the best way to communicate with their users (ie. too much relying on the forums, the voting tools…).
• The decision, in October 2004, to stop all new feature development and to concentrate on bug fixing definitely gave us a more stable platform, but it also means that SL did not get much “innovation” in terms of the platform (ie. the competition is developing much more fascinating features these days).
• Too many abandoned/incomplete projects:
  • XML-RPC was introduced in June 2004 and never completed;
  • The implementation of Havok 2 in SL is now a running joke since nobody believes any more that it will ever be implemented (Andrew Linden recently said that they were “halfway through with the development”);
  • Even the LSL interface to Havok 1 was never fully completed, many functions are still missing;
  • The interface is still not customizable (nor are the keys on the keyboard), despite many “configuration files” which show that something in that direction has been planned;
  • No new permission system (and the current one seems always to have problems), although it was discussed/announced last year;
  • Groups never worked as they should, and fixes have been always postponed;
  • At this moment, the status on in-world HTML is “shaky” at least — despite demos shown publicly months ago, it seems that it won’t make it into 1.7;
  • Some bugs are still around since beta days;
  • LL continues to stubbornly defend their “own”, primitive Instant Messaging system, instead of adopting a permanent solution (like Jabber).
• Time to introduce new features/bug fixes is rarely correctly estimated (examples: 6 months for Havok 2 — waiting time, 2 years and counting; 6 weeks for HTML — waiting time, 6 months and no results…).
• Certain implementations are way out of control in terms of man-hours allocated — for instance, PayPal integration, which also allows payment by credit card (these days you don’t need a PayPal account to use the system…), is something that just takes an afternoon to implement (assuming you have a billing system in place, which LL does) — LL took over 3 months.
• The Abuse Reports system gets… abused. It’s hard to get “fairness” and “justice” in Second Life since LL has some difficulty in understanding “what is the truth” based upon the few facts they can gather electronically.
• Overall, the weakest spots at LL seem to be Marketing and Project Management.

Well, you get the idea…

Threats

• A small user base compared to MMORPGs (although somewhat near the “social” virtual worlds like There and The Sims Online) makes SL be “virtually unknown” among the 5 million estimated users for 3D entertainment platforms.
• Funding is not enough for really big teams and projects, and since profitability is low, nothing prevents a Big Player to enter the very same market with enough money to surpass LL’s struggling achievements in just a few months.
• The most successful 3D environments are right now done on MMORPGs with static content, providing a much richer and fuller experience than SL can ever aim to provide (3 years ago, the difference was not so visible). MMORPGs can always make their engines better and faster, and deploy DVD-ROMs with more and better static content. SL, in contrast, will always be limited by the amount of bandwidth you’ll need to get “better” dynamic content, and certain aspects of the rendering techniques used simply cannot be applied in the near future to improve SL’s rendering pipeline.
• Big worldwide marketing campaigns by the top entertainment companies are able to reach a few million prospective customers in few months. LL takes years to get to the same level just by relying upon “word of mouth”.
• LL, not having a specific market to target their platform, tries to please all, and ends up giving each niche market crippled functionality — i.e., it’s a platform for creating, playing, making money and socializing, but in all those aspects, the interface is always limited (3D modelling has an amateurish tool; combat and vehicles are not up to the same level of development than any MMORPG, even the slowest and oldest ones; models like There or IMVU are totally oriented towards making money; and, sincerely, there isn’t a worse IM system currently available).
• In contrast, other MMORPGs/3D Chat Rooms have learned from Second Life’s emphasis on “create your own world” and are now offering much better social interaction tools, all forms of content development (to the degree that static content allows it) with better integration with professional 3D modelling tools (thus appealing to the 3D pro designers searching for new markets) and slowly starting to “open up” their games/platforms to allow users to make money “legally” out of their gaming experience (instead of suing them, as has been the standard practice so far).
• SL’s cost structure is way too high! To support around 60,000 users, LL needs around 2,000 servers at its co-location facility. World of Warcraft, the biggest success these days, is able to support 4 million users (1 million of them simultaneously!) with just a hundred servers. LL already runs one of the 500 largest grid-based parallel computers in the world. If it grows to 1 million users in the future, it will highly likely be one of the 50 largest! Thus, around 2010, when eventually LL is able to have the same amount of users as WoW has today, they’ll have to employ an operations staff of perhaps a hundred or so highly trained professionals (sysadmins are expensive to hire and need years and years of training), while WoW can probably survive with just one or two right now! (A good rule of thumb is that one sysadmin with 10 years of experience can successfully run 500 servers by themselves, especially if they are all alike in terms of configuration, which is the case).
• Also take into account that since SL is now free, LL has to rely on people buying land to get some extra income. So, they have immediately increased their user base, but not their revenues, while at the same time demanding the need for more tech support people (another good rule of thumb said that you should have one tech support operator for every 1000 new users — these are the ones that require tech support most, at the beginning. LL partially offsets this by encouraging volunteer tech support in the form of the Mentor/Greeter/Instructor/Live Help groups, around 500 unpaid users who contribute their time to do tech support). In contrast, WoW users, paying a fixed amount every month, constitute a highly profitable venture, even if WoW “fades out” in another half a year or so — to be replaced by WoW II or something like that.
• The proliferation of so many good engines for the development of 3D virtual worlds (the most notable — although by far the one that impressed me less — being OpenCroquet) show that there is a growing academic and open-source network of excellent developers working in this area. Free from the stress of having to run a successful business, academia and open-source volunteers are slowly growing “curious” experiments into fully-fledged engines for running virtual worlds. Again, to the best of my knowledge, they are still far from SL’s current development level. But the sheer mass of volunteer developers eager to jump into the 3D VR bandwagon is growing. Since they cannot improve (yet) upon LL’s code, they start from scratch with a 3D engine and add all the features they need. Things like the Open Source Metaverse Project, with just a handful of active developers, are already at the stage LL was in, say, 2001 or so, with meshed objects, more complex prim types, a 3D modelling tool, a better physics engine, and Lua as the de facto scripting language. And it will probably be “Second Life compatible”. Again, this is just a tiny project, but an example of how fast the technology can advance “outside” Linden Lab.

Conclusions

To the best of my analysis, Linden Lab’s biggest crisis came with version 1.5. It was improperly tested, and, despite some eager optimists (I seemed to be one of the very few exceptions that actually got a much better performance!), it rewrote completely Linden Lab’s development plans. Instead of concentrating on earth-shattering features, or fixing very old-time “bugs” and “nagging details”, all the efforts of Linden Lab were focused on stamping out the more outstanding bugs — like all the issues with the asset server (inventory-related, teleport-related, and otherwise) and the login servers. Only with the latest versions of 1.6 were these problems overcome. In the meantime, all other “projects” or at least “projected features” were delayed sine die. We now can’t expect a Linux version soon, or integration with Havok 2 (which, in the meantime, is already outdated), a new interface under control of the users, or even HTML-on-a-prim, not to mention some Mono support (faster scripts). The most radical change in the past year was video streaming (never had the success of audio streaming) and some better building tools (nothing too spectacular, although they are definitely better). All the rest were minor tweaks. The upcoming 1.7 version, already 2 months late, will just have 3 new LSL functions, an improved scheduler for the scripts, and HUD attachments. Although it has several hundreds of bugs fixed (yes, really!) it does not bring anything radically new to Second Life. It can only be considered as a “maintenance release” — like the jump from Windows 98 R2 to Windows 2000. LL was very conservative in their development. For the users, this means looking at competing platforms and weeping. Things like the end of GOM also meant overtime development of a new in-world money exchange (something that an amateur programmer would be able to do over two weekends, but that LL needs, of course, about 6 weeks to do). So, the only hope is that some of the promised developments will make it into 1.8, but some pessimists say that 1.8 is simply going to be “skipped”, and we’ll just have SL 2.0 instead, with a new renderer (which we already know that will be crippled in so many respects that most people won’t be much impressed by it, except perhaps for its speed). This means waiting for sometime next summer for some exciting new features, at the current development speed — although, to be honest, LL released 38 (!) versions of the 1.7 beta in just 2 months (compare that to 14 versions of 1.6 in about 6 months or so). Estimating another 6 months for 1.8 (again, just to wait for HTML, and eventually some integration with Havok 2) and a further 6 months for 2.0, this would mean that by the end of 2006, with perhaps some 200-300,000 users, Second Life will look much like what we know today — just busier in terms of population and events.

It doesn’t surprise me at all that LL is emphasizing “community” (with the first Second Life Community Convention coming next week) instead of “technology”. However, my feeling is that this is not enough. The most eager users of SL are the ones currently relying upon the community aspects, but there is quite a large user base of programmers that get more and more discouraged as each version is rolled out without anything “interesting”. Perhaps Mitch Kapor, President of the Mozilla Foundation and currently a LL board member, is applying the “Mozilla experience” to SL: if you have an outstanding technology (Netscape’s rendering engine), people will wait for it to be deployed (7 years in the case of Mozilla), no matter how long it takes. Does that mean that “outstanding features” of SL will only be available in 2012? Only time will tell, but I’m pretty sure that LL’s competitors won’t be idle that long. The only reason they haven’t been more aggressive is simply, as said, because the business model is not very interesting.

Still — Philip now thinks about the competition. That, in my eyes at least, is a Very Good Thing 🙂