Microsoft Life?

Ironically, just a few days after writing my other post, imagining that after a rough summer, with lots of media drama around Second Life and Linden Lab’s ever-annoying attitudes to frustrate their userbase more and more, we would enter a period of relative calm until Christmas (when LL is supposed to start announcing the grid-wide implementation of meshes and Display Names), the SLogosphere was suddenly shaken up with the interesting rumour that Microsoft might be buying Linden Lab (some sources even think that this might already have happened, and that both companies are just stalling their joint announcement). Both Microsoft and Linden Lab have declined to confirm or deny the rumour.

The fun bit about the rumour was how it was spread. It was started by the notorious hacker and Woodbury griefer Tizzers on Twitter, who didn’t even post any sources for the rumour, nor even any follow-up messages. For some unfathomable reason, this simple message gathered the attention of some blogs and e-zines, some of them actually quite reputable. As soon as it was picked up by them, others started to cross-link to the original message. Tizzers was promoted as being “closely in contact” with Linden Lab; how that bit of misinformation popped in I have no idea.

Some of the more serious e-zines, part of reasonable large publishing houses, actually tried to get in touch with Microsoft to comment. ZDNet created an interesting article noting that Microsoft hasn’t done any acquisitions this year, so this might be their first one. All of a sudden, ex-Linden employees and even some actual employees have somehow “confirmed” that there was some truth in the rumour; somehow, a few were aware that there had been conversations among Microsoft and Linden Lab.

Now where did all this come from? Tateru Nino, who usually gets her facts right, has a more reasonable explanation. And this requires a bit of delving into corporate tactics…

I must say I love to spread rumours too 🙂 It’s a fun exercise. For my April Fool’s articles, which have often unpredictable consequences, I use always the same trick: start with a few verifiable facts, re-interpret them in a new way, and add a simple “white lie” which could be true (i.e. it has to be plausible). People will often check on the facts and find them to be true and correct, and thus be lead to accept the whole article as genuine — a typical fallacy. For instance, an announcement that Google would have a research division on the Moon is utterly unbelievable because nothing can be verified. A more clever announcement would be that Google’s Page & Brin would partner with Richard Branson to supply Virgin Galactic‘s future orbital data centre — the media knows that Branson is insane, but he does own a company that does sub-orbital space flights, and together with Google, they would have money enough to build a space station (if ordered to the Russians, it would cost merely US$ 1 billion — cheaper than what Google paid for YouTube in 2006). So that kind of “news” would be plausible and parts of it could be validated.

This rumour about Linden Lab’s sale to Microsoft has undergone similar strategies. Journalists reporting on this have correctly claimed that LL was closing offices in the UK (which is not “news”; it was announced a couple of months ago, but it’s an useful factoid to “revive” for the unsuspecting audience); they have connected Woodbury’s Tizzers to Linden Lab (quite a feat, but there are certainly a few articles about LL’s reaction to the Woodbury griefing that can be located on the net, and this would establish “a connection”). They assume that the leak originated from Linden employees, or, more believably, by disgruntled ex-Linden employees (people can validate that a lot of employees were fired, and immediately “believe” they would be disgruntled). And, of course, Linden Lab is VC-funded like pretty much every technological company in the US, and a good journalist is aware that, sooner or later, shares held by VC partners will be sold out (more on that later). Then, fortunately for the journalists, both Linden Lab and Microsoft have neither denied nor confirmed the rumour, which adds to its credibility — a straight denial by both companies would bury the rumour immediately.

So put all this together in a package — some journalists added a few more verifiable factoids to conjure up more credibility; I address ZDnet’s case below — and you have a “believable” rumour, because parts of it can actually be verified and validated and found to be true, even if they are irrelevant. And the whole rumour in itself wasn’t explicitly denied — so it means it could be true. Get enough media to crosspost, let journalists watch the blogs and the social media networks finding hundreds or thousands of people writing about the very same issue, and the rumour spreads like wildfire. I have to take my virtual hat off to Tizzers; I never managed a fraction of the success with my own story that Google had bought Linden Lab, and, in my not-so-humble opinion, I had far better arguments back in 2006… I guess that to spread a good rumour you have to be a great hacker and known griefer, and I’m neither 🙂

Anyway… first of all, let’s assume, for a moment, that the rumour is actually true (I’ll address later the issues that I personally have with it). If so, no matter how much a journalist might pester Linden Lab and/or Microsoft, they wouldn’t answer to them. A message neither confirming nor denying is usually the standard routine. Let me explain why.

Last year, Business Insider valued Linden Lab at US$ 800 million. Obviously this is just a number based on observation of estimated profits and the userbase, as well as positive media exposure. Linden Lab is a private corporation, so they’re not required to publish their numbers. I would even venture so far as to admit that the people at Business Insider might not have thoroughly analysed Second Life’s numbers — which are published every quarter — and, based on those, you can roughly estimate how much LL makes. It’s a popular pastime of many residents, trying to figure out how much profitable LL actually is. Since the numbers have the same source (the SL economy) they are usually around the same range. Based on that, we can estimate the following: Linden Lab’s owners might be willing to either sell a share of it (like Facebook did) and get a new partner, or sell everything and cash in.

LL’s funding partners have invested, until 2006, around US$ 11 million (these are the public numbers we got). Based on that, we know that typical venture capital companies want to recover at least ten times of what they have invested, after a decade. the timing is about right. If we assume they own 10% of LL, they would wish at least to get US$ 100 million back, and that would value LL at around US$ 1 billion. Typically, VC companies and business angels expect to invest in ten companies, have seven of them utterly fail, two doing well enough not to lose money, and strike gold on the last one — selling their share of it for the amount they’ve invested in all ten. It’s like gambling, but legal: you bet on a dozen companies, hoping that one pays off. LL is probably one of the very few of those that might “pay off” — if they find a willing buyer.

Now, LL’s VC funding partners have not sold their share of LL yet — we can see that from their corporate website — so it’s quite reasonable to expect them to say: “We have been with you for over a decade. You guys are one of the rarest things on Earth: a profitable Internet company with a solid business model. We’re very happy. Now we want to recoup our investment and put the money elsewhere”. This is standard routine for a VC company; all work pretty much the same way.

Since LL is not a public company, everything is handled discreetly. As soon as some media start hinting at the value of LL, it’s reasonable to admit that the VC partners of LL might have started to get in touch with potential buyers. They have great arguments. They could say they had a share worth US$ 100 million and tell potential buyers that, thanks to LL’s reasonable profits (estimated at around US$ 50 million annually — some people, including Hamlet Au, have come up with this number based on tier prices and commissions on the LindeX; and we also know how many employees LL has, and how many of those are developers, so we can also estimate labour costs. Office lease costs is also easy to calculate based on what real estate websites show in the area around LL’s offices), a potential investor would get their money back quickly enough. After at least half a decade of profitability, LL can make a good point of having a solid business model, and some investors will look less at the media exposure and more on the real numbers.

So after a while, the VC companies might bring up a selection of potential candidates. In some cases, they might require the approval from LL’s Board to sell their share (say, LL’s Board might not be interested in partnering with a direct competitor). In other cases — and nobody knows how the share distribution is at LL; the VC companies might own the majority of the company’s shares, although this is rare (10-30% is usual, more than that would be very strange indeed) — they might just inform who the new owners of their slice of shares will be.

Let’s assume that they have a short list of 3-5 potential buyers (Tateru and others consider even that they might have had a dozen offers or more). What happens next is that this list gets ordered by priority, and a period of negotiation starts: both LL and the potential buyer enter an agreement of exclusivity during the negotiation phase. What this means is that the potential buyer will now be able to perform due diligence on LL (i.e. look at their accounting…) without fear that another potential buyer might grab the deal under their feet. During this period, both companies shut themselves off from the media and sign mutual NDAs. Nobody is allowed to talk about the negotiation. Other potential buyers are told to wait; LL will not reveal to anyone who is on that list, not even to the potential partner currently under negotiation. Usually, this phase lasts a few weeks: the potential buyer has to understand the nature of the business from inside (not from wishful thinking based on media reports!), but LL doesn’t want them to take too much time, because they might drop the offer, while others on the list might have better ones.

In some cases, an initial bid is requested, and LL might order the list according to the highest bidder, and negotiate the price afterwards. In other cases, LL might request a payment to move up to the list, if there are a lot of potential buyers. There are some negotiation variants at this stage, but the important thing to remember is that it will all be secret. That’s one of the huge advantage of never going public, you don’t need to reveal anything to the general public at all, and this also includes the media: journalists can beg, speculate, and be creative about what they write, but when a privately held company is being sold, nobody has a clue until the process has been completed.

And all employees — even the ones leaving the company — are under NDAs as well. They won’t “leak” anything (and most of the time they will be kept in the dark anyway) for fear of getting heavily sued!

This negotiation might actually have started early in 2009. Like Tateru reports, it’s highly likely that nobody wanted to pay a price acceptable to what LL was expecting. LL is actually in a complex situation: investors love tiny, cheap, and very popular Internet companies, grabbing them when they’re young and still inexpensive but already rocking the media. When they grow to a certain size, like Facebook for example (or YouTube when Google bought them), they become harder to estimate. There are plenty of ways to estimate companies, and each arrives to completely different numbers. Interestingly enough, the dot-com era brought a new way to estimate them: based on media hype, which is the most unrealistic and senseless way to estimate a company…

That’s why the bubble burst. Ten years later, companies are still getting evaluated that way! Most investors never learned the lesson. And as a side-effect, journalists, as the providers of speculative hype about a company, have a much stronger influence in the negotiation process, which is completely insane: instead of trusting an accountant or financial consultant to estimate how much a company is worth, investors believe more in journalists… and rumours spread by hackers and griefers!

It’s a strange world.

But that doesn’t mean that the rumour might not have some grounds for actually being true…

Why Microsoft might be interested in Linden Lab and Second Life

Microsoft is not exactly a “newbie” corporation in terms of understanding the importance of something like Second Life: they have been around SL for a few years. They used to have a permanent presence, and did some developer conferences in SL; these days, they have moved to OpenSim (namely ReactionGrid) and have done some great job integrating Live ID logging in to the SL Viewer, but a few Microsoft-sponsored groups still remain in Second Life and meet there or do product releases. Compared to all the other companies doing some work in SL-compatible virtual worlds, and although IBM and Intel lead the development and interest, Microsoft is not exactly crossing their arms and waiting. So, a good, strong point in favour is that they know about SL, they know about LL, and they’re not strangers to virtual worlds, but active developers.

A few journalists think that Microsoft might be looking to launch their own equivalent of Sony Home for the XBox. With the upcoming Natal controller-free interface, it’s plausible to imagine that having a generic social virtual world for the XBox. Imagine that the building tools would simply be gestures to “grab” prims in the air, resize them, move them around…! Second Life has a lot of possibilities to become a technology for Microsoft to “show off” some incredible things.

While Microsoft might not always have bought the best companies out there, they aren’t too bad, either — things like Office or Navision come to mind as having been rather good acquisitions; I think that in both cases the companies behind them were profitable, or at least, close to that. MS’s Office division is definitely still very profitable! So Microsoft might be looking for a profitable company for a change, and caring little about the media hype.

A few things come to mind… Linden Lab sort of started avoiding IBM and Intel, with whom they were co-developing the intergrid teleporting protocols. This was a bit baffling, since immediately afterwards, IBM and Intel increased their work on OpenSim even more. The “romance” with Microsoft might explain why LL dropped interop and estranged IBM and Intel.

Microsoft could also have influenced LL’s decision of merging the two grids: after all, their games division targets content for the teen market. The bad news is that all adult content might have to go…

LL also announced the end of Avatars United. This would naturally make sense if Microsoft was thinking of buying LL. Microsoft already has Live, and shares in Friendster… and, of course, in Facebook, which would explain why LL was so eager to develop closer integration with Facebook.

With Zune, Microsoft has shown they’re also interested in sales of digital content. Second Life is, for all purposes, the largest marketplace for virtual goods in existence, in terms of number of available products (270 terabytes of items is the number quoted by LL a year or so ago) and overall sales, being worthUS$ 0.6 billion annually. Microsoft may think they could leverage the SL Marketplace to their advantage and announce themselves as being bigger than their “new old rival” Apple — something that Linden Lab keeps forgetting to announce: the overall SL economy sells more than Apple’s iTunes & App Stores…

And imagine that Microsoft would persuade LL to get rid of their Google Appliance Server… and replace it with Bing instead. Also, Microsoft has their own ad technology (the one that used to power Facebook ads!) so this might replace in-world classifieds.

An interesting possibility would be relaunching Second Life as an enterprise product, using Microsoft’s world-wide sales network to reach a far wider market than Linden Lab ever managed to get. LL’s enterprise division could thus be revived from the dust and given a second chance.

And of course they might use Second Life to show off their Azure cloud platform (replacing what Linden Lab has been doing with Amazon’s own service to push textures and assets via HTTP).

Finally, this would also explain why M Linden had kicked out so many employees, namely, developers… Microsoft has plenty of programmers! A leaner LL would be more interesting for Microsoft to buy, and of course, they could bring most of the viewer code  to their internal teams to develop further.

So these are good, valid arguments for the rumour to be true…

Why Microsoft actually might not be the buyer…

From a technological perspective, Linden Lab is the worst possible example for the kind of company that Microsoft usually buys. LL is too keen on open source code, use it extensively in their own products, and they have released the viewer as open source. They support multiple platforms on the viewer; while Microsoft also develops for the Mac, it’s quite rare for them to support Linux as well. Interop would be one of the things that Microsoft would want least to develop!

The servers are completely the opposite of what Microsoft likes: they are developed natively for Debian Linux and the backend is MySQL, and they’re un-portable to Microsoft’s technology. By contrast, Microsoft is active in developing for OpenSim… because it’s a Mono application, close enough to Microsoft .NET and fully compatible with it. It doesn’t cost anything to launch their own OpenSim grid, and if they put the Microsoft marketing power behind that, they won’t need Second Life’s brand awareness at all.

Second Life also has no technology worth to be patented. The most original technology it used was 3D content streaming, but, as you all know, Linden Lab is phasing it out in favour of HTTP asset downloading from the cloud. The rest, individually taken, is not sooooo original — it’s the ensemble that is special. So Linden Lab is not worth much as a source of patents… even though the brand name, “Second Life”, might be worth something, since it has a relatively good brand awareness — so long as Microsoft can clean its association with adult content!

Sony Home was, for all purposes, a flop. Would Microsoft emulate Sony and fail as well? It’s not likely. This year, Microsoft is announcing a lot of remakes of their ‘classic’ PC games. Some analysts believe that Microsoft wants to push gaming back to PCs (although I fail to understand why they’ll launch Natal this Christmas). Buying SL to put it in the Xbox, thus, would make no sense at all… and it would mean turning SL into a PG world. Even if Microsoft would be so keen to try this out, they could do it way more easier with OpenSim… and compete with IBM, Intel, and Linden Lab with the only SL-compatible virtual world running inside a Xbox.

Also, although Microsoft’s game division focuses on the residential market, Microsoft is usually more interested in business applications. Second Life Enterprise could potentially be one, but Linden Lab, for all purposes, dumped their enterprise division. During due diligence I’m sure that Microsoft would find out it didn’t sell that well… and Microsoft already has plenty of useless, unsellable products on their portfolio.

Why would Microsoft be interested in getting just a tiny slice of the US$ 0.6 billion content economy? The commission fees on the LindeX are too tiny to make a difference. They could, in theory, attempt to limit free content upload and demand from content creators that they pay larger upload fees or sign-up as a “registered Microsoft virtual world content creator” (not unlike what Apple charges to allow developers to sell their applications on the App Store). With a US$100 fee, 100,000 professional content developers, and a million amateur ones, there might be some money to be made out of those fees. But… would content creators be that willing to “sign up” and pay Microsoft for the privilege of doing something in their favourite virtual world which has been, for all purposes, free until now? It’s far more likely that this would simply push every content creator into the OpenSim grids — where Microsoft already has a relatively strong presence and is quite aware of the growth of OpenSim.

Similarly, why would a company already making half a billion US$ annually from ads served by their adCenter technology be interested in a handful of classifieds on the Second Life Search? Even if Microsoft forced aggressive ad spamming inside the viewer, why bother…? There are just a million and a half active users! Granted, Microsoft lost the revenue from placing Facebook ads, but placing ads in Second Life would hardly be significant…

As for branding… “Second Life” and “Microsoft Live” might be two brands too easily confused with. Would Second Life survive a name change after almost a decade? Perhaps, if the focus changes radically. But if it does, why buy Linden Lab? Again, running an OpenSim grid would be far cheaper and make more sense… and it could be simply rebranded as Microsoft Virtual World, Microsoft Metaverse, or something like that, to avoid confusing it with Live.


All in all, when considering the arguments for Microsoft to enter the social virtual world business, the key issues against it are:

  • Second Life’s market, compared to Microsoft’s, is far, far too small
  • Linden Lab uses too much Microsoft-unfriendly technologies which are impossible to port
  • Brand awareness is good, but usually associated (by the media) with adult content; changing the brand would be pointless, since it’s probably one of LL’s biggest assets; the current brand is confusing with an existing Microsoft brand
  • There is little that Microsoft cannot do by using OpenSim instead of Second Life, if they’re serious about the virtual world business: with the advantage that OpenSim is completely free, Microsoft-friendly, and compatible with Microsoft technologies (unlike LL’s core servers) — and, of course, Microsoft is already tinkering with it. And Microsoft already partners with ReactionGrid for establishing a presence for them and their own partners. Why should they buy Linden Lab?

The only argument for buying a share (or even all) of Linden Lab is for its profitability. US$50 million annually might not be much for the software giant, but, in this era of economic downturn, it might not be too bad, either. Leveraging on Microsoft’s huge and successful marketing network, they might potentially reach a wider audience, so far as they understand that Second Life is really a niche market (and I’m sure that after 3 years or so of intense use, Microsoft is well aware of the market limitations). You might have noticed how strong IMVU’s web advertising is — some sites show three or four ads for IMVU on the same page! Now imagine that Microsoft puts the Second Life viewer into Microsoft Update, and starts pushing ads through adCenter as aggressively as IMVU does via Google AdSense! It’s reasonable to admit that they might get 2-3 times the number of residents that way, and push the profitability into US$100 or so million dollars. Again, not much for Microsoft, but possibly better than what some divisions are making — most run at a loss.

Microsoft might also be starting a change in their attitude. I call this the “friendly IBMisation”: when IBM finally understood, by the turn of the century, that they weren’t going to lead the PC market, they completely redefined themselves in a relatively short time. All of a sudden, they embraced open source technology very aggressively. Their trademark grey suits and ties were replaced by cool techie guys in casual wear and ponytails. IBM, all of  a sudden, switched from being “corporate dull” to “geeky and trendy”, and, thanks to that, still remains as an industry leader.

With Apple surpassing Microsoft in size — and definitely in coolness! — Microsoft under Ballmer might be looking for ways to become the “cool company” they were around 1985, when they pushed IBM out of the PC business. I have no clue if Ballmer can pull that off, but they might try. One clever change would be to look at buying profitable companies instead of media-hyped ones. IBM, by the turn of the millennium, started focusing on low-cost, high-quality open source architectures to sell to their clients, because they were far cheaper to maintain and further develop internally — by reducing costs, they enhanced profitability. Microsoft might instead adopt a new attitude: buying successful, profitable companies with a proven business model instead of relying on media-hyped one-person companies. If they bought Linden Lab and pushed that message out, it would certainly put in question the whole attitude towards “cool technology” of the past decade…

I’m not sure if Microsoft was one of the bidders or not — and if Linden Lab rejected Microsoft’s offer, we will never know, no matter how hard journalists push and how much information gets “leaked”, through reliable or unreliable sources — but, like Tateru, I think that Linden Lab, or at least their VC partners, have been entertaining bids for a long time now: at least a year, but possibly more. It would be very surprising if they didn’t get a single bid, since at least the VC partners, after a decade, would almost certainly be looking around for buyers of their own share, if not of the whole company, no matter what LL plans — unless, of course, Philip is interested in buying back the shares held by the VC partners (which could be a possibility; I would, if I trusted the company I ran; in fact, I did exactly that, a decade ago). I also agree with Tateru that the “long wait” for selling out a share of LL is because LL (or their VC partners) are waiting for better offers. The more the wait, the better. The economic downturn means that fewer companies will be willing to invest; however, Linden Lab remains profitable and is increasing profits by downsizing and dumping all “exploratory” projects which didn’t have any success. They might be one of the few good examples on survival through a financial crisis, where they have actually grown in size and profits, while everywhere else companies are shrinking. Microsoft, or any other clever company, might start looking at companies like Linden Lab as far better choices than media-hyped techie companies without valid business models, and this might not only change the tide in the technological industry, where all that is geeky and has “lots of registered users” attracts attention, no matter how much money they’re losing. No, we didn’t learn anything about the dot-com bubble 🙁

Still, it’s really a question of time until LL’s VC partners sell their shares. There are three possible alternatives: either they sell them back to Philip; they sell them independently; or they agree with the Board to sell the whole company. When you have run on VC funding, these are the only choices: no VC company, no matter how good the relationship between them and the board, will want to hold on their shares indefinitely. They’re in the business to invest and get a huge return on the investment after a few years. They waited almost a decade. Now it’s time for them to sever their ties with Linden Lab and go away.

[UPDATE: Thanks for the many commenters that pointed out that the SL viewer code is written in C++ and not C#. To be honest, I have just looked at the Mac version, and that was quite a long time ago… I’ve deleted the erroneous reference from this article]

Print Friendly, PDF & Email
%d bloggers like this: