59 Comments

Summary:

Clay Shirky is a brilliant analyst of the digital era, and if there’s anybody who could implode the media vortex currently surrounding Second Life, it would have to be someone of his caliber. He attempted that last week with “Second Life: A Story Too Good to […]

Clay Shirky is a brilliant analyst of the digital era, and if there’s anybody who could implode the media vortex currently surrounding Second Life, it would have to be someone of his caliber. He attempted that last week with “Second Life: A Story Too Good to Check”, a Valleywag post that was subsequently Boing Boinged. As someone who contracted with Linden Lab as their “embedded reporter” for near three years, and still has a substantial interest in the world they created— consider that both full disclosure, and plug — I actually found his post something of a relief. Attention around SL has been growing at such a heart-throttling pace (total registered accounts were under 100,000 only a year ago, and just blew past 2 million), you begin to hope for something, anything, which will slow growth to a more manageable clip.

But is Clay’s analysis correct? Yes and no. The very topic came up at a GigaOM staff meeting recently, and when a couple staffers suggested Second Life is over-hyped, I blurted out what is, I think, a far more accurate assessment: “It’s too hyped– and it’s not hyped enough.” There is, to be sure, far too much attention on raw numbers and corporate-funded promotions in-world, and to be equally sure, not enough coverage of the internal economy and the grassroots prototype content creation which make the place truly revolutionary. After the break, my attempt to separate hope and hype, as reflected through Shirky’s polemic.

What Clay Gets Right

Where Shirky scores undisputedly is against the breathless media reports that proclaim “two million users in Second Life!” This number repeats the total registered accounts listed somewhat confusingly as “Total Residents” on the SL homepage, which are not truly reflective of users who actually spend time in-world on a regular basis. “Resident” has always been Linden’s official name for any person with a valid SL account, but it’s easy for the uninitiated to also infer the word’s literal meaning, of “living in a place for some length of time.” Which two million account holders are, plainly, not.

As he notes, that number is not even the nearly 800,000 or so who’ve “Logged In Last 60 Days” (another SL homepage metric that confounds more than enlightens), since most of that includes those who’ve tried Second Life within that time—but never returned. He estimates the churn rate of One Try Then Bye Bye at 85%, and by Linden’s own measure, that turns out to be, to his enormous credit, pretty much on target. When last month my blog’s demographitrix Tateru Nino wrote a story discussing Second Life’s churn rate, Linden CEO Philip Rosedale suggested retention was 10%— a percentage so low, it shocked me. (“[A]bout 10% of newly created residents are still logging into Second Life weekly, 3 months later.”) When I checked with Linden Lab last week, Philip and Marketing Director Catherine Smith reported back a slightly higher percentage, this one gauged by returning users from over the last 30 days (but not those who created an account within that period)— “12-15% and has remained steady over the last year… we know churn will be high, but the difference is a network effect and constantly changing content that people do come back to see. ”

So this, as it turns out, is where to set the Second Life bar: of the 2,000,000+ registered accounts now, roughly 240,000-300,000 are regular users, residents in both the colloquial and literal sense. Clay is right to call for the media to stop reporting that very top number without caveat. Shirky is further correct to wonder if all the big companies recently promoting their brands in Second Life (NBC! American Apparel! Adidas and Toyota!, etc.) count as news, since it’s not clear if this is just a gimmick, or if they’re actually getting any measurable return from their promotion dollars.

The truth is, no one quite knows if that’s the case, and similar to the Web’s dot com boom, the metrics for judging a real organization’s success in Second Life are still controversial and contentious. While the mainstream media largely hasn’t asked this question yet, it’s been roiling through the metaverse blogosphere (as here and here from me, and here and here from SL-based marketers and virtual world development studios.) A reckoning is surely due soon (if not overdue already), and good on Shirky to call for it.

What Clay Misses

But Shirky’s wholly valid points go right off the rails, in my view, when he suggests that the hype around Second Life is merely due to the spectacular (but largely spectral) growth of created accounts. Worse still are his speculative explanations (i.e., virtual worlds are easy for reporters to understand, young reporters don’t know about LambdaMOO, etc.) for why the press has gone so wacky over SL.

As someone who began with feet in both worlds (I first caught a demo of Second Life in 2003 as a freelancer for Salon.com and Wired Magazine), I can offer some perspective.

User-Created News Content

Even throughout 2004-2005, when the world had less than 100,000 users, media attention for SL was already fairly voracious. I can’t speak for Linden Lab’s direct PR efforts, but I can name the numerous times when big media outlets e-mailed me on their own, without Linden Lab’s prompting at all, not because they were particularly interested in Second Life, or considered it the Net’s next big thing, but because they kept stumbling across fascinating stories about it on SL-centric blogs, including mine.

I’m not vain enough to think they came to me because they liked my writing; these stories were compelling on their own. (Rarely did they even know my own blog was funded by Linden Lab, when they came across it; and often, their interest was in the controversial stories I wrote which didn’t depict the world in a utopian light.) Big media reporters read about SL’s private detectives, who stung unfaithful virtual lovers, for example, and reported on them; they read about a formerly homeless musician who made a living singing live as a frog in Second Life, and did so too; Residents banding together to help Katrina’s victims, and several stories followed.

That pattern continues today, and while it is true Linden Lab’s publicists shepherd some media toward some SL stories, at least as much attention emerges from the grassroots of the blogosphere. (And even when the big media turns its spotlight on the world, as Time Magazine just did when they made Second Life part of their “Person of the Year” profile, unfettered hype is not always the result: I doubt Linden planned for Time to complain about how it “takes forever to download… [and] sucks up hours just to design your character”, and mostly focus on “Meaningless, multipartnered, degrading sex” and detachable penises.)

What’s more, we’ve reached the point where the world’s content creators now bypass Linden Lab entirely, to hold their own press conferences. Clay mentions recent news of Anshe Chung, the Second Life avatar putatively worth a million dollars, and he’s smart to be skeptical about that figure, as was I. (When asked about it, Philip Rosedale pointed out to me that Anshe’s assets are not as “illiquid” as Clay seems to think, since she can put up her virtual land holdings on the auction market immediately.)

In his skepticism, however, Shirky misses an even more crucial point: the press release that the media picked up was put out not by Linden Lab, but by the avatar herself, since she now owns a company which employs a staff of skilled 3D developers (not gold farmers) in China, all paid for by her virtual world commerce.

This should not be surprising: user-created content inevitably leads to user-generated news. And when there’s a monetary reward for creating that content, it’s also going to involve user-generated hype. 58 Second Life Residents make over $5,000 a month from their in-world activity, while nearly 3000 of them earn $50-2000 a month. The world’s top content creators, in other words, have even more incentive to promote Second Life than the employees of the actual company which owns it. (And that’s not even mentioning the five or six “metaverse developers” which create Second Life experiences for real world clients, nor the advertising, marketing, and PR firms which have set up shop in-world, another engine of SL enthusiasm.)

Why Experience Matters

Throughout Shirky’s essay, I kept thinking, “This is someone who’s never really explored Second Life to any significant extent, if at all,” and it wasn’t just because of the factual errors which emerge through that gap.

As it turns out, Clay Shirky has very limited personal experience with Second Life. (See his e-mail to me below*, and his argument for why that lack shouldn’t matter.) The larger problem isn’t the errors, however, but an experiential absence that leads him to reason, with a fairly sly leap in logic, from “Second Life is much smaller than usually reported”, to, “Therefore, it’s not the Net’s next big thing.” (Or as he charmingly puts it, not the “Immanent Shift in the Way We Live®”.)

But in between both statements, several stepping stones are missing.

First, to the basic factual goofs:

“If we think of a user as someone who has returned to a site after trying it once, I doubt that the number of simultaneous Second Life users breaks 10,000 regularly. If we raise the bar to people who come back for a second month, I wonder if the site breaks 10,000 simultaneous return visitors outside highly promoted events.”

This simply isn’t true. This month, as last, in-world concurrency has regularly been 15,000-18,000 during peak hours (barring downtimes, a quick glance at SL’s homepage between Noon-8pm PST will show that), and given the new account creation rate (about 12-14K total throughout the entire 24 hour period), far less than a thousand of them are new users at any given hour, on average. Concurrent users at peak have regularly been exceeding 10,000 since September; three months later, it’s fast approaching 20,000 at prime time. (See the screen capture above, taken at about 2:00pm last Sunday.)

What’s more, that number has very little to do with “highly promoted events”, as Clay suggests, because most events are architecturally limited to 120-140 people maximum. (Anyone who’s been to such events will tell you about the monumental lag it takes to reach even that figure.) For another, most SL activity isn’t based around one-time events, it’s based around communities and established popular sites.

That’s two instances where Clay’s lack of ground level experience with Second Life undermines his analysis. Another comes up with his mention of CopyBot, the Resident-created external program which seemed to enable instant theft of user-created content. Because its existence provoked such FUD among the Second Life community last month, Shirky offers it as evidence that user-created content in SL is a sandcastle easily demolished by the right hack. What he does not mention (likely because his in-world visitations are so brief) is that overall, the CopyBot panic receded almost the moment it began. After a bit of refereeing from Linden Lab, content creators largely shrugged and went back to business, and many of those I interviewed were actually more irked by the protest, than the CopyBot itself. (And longtime Residents know that collective fits of agida are regular, melodramatic– and quickly forgotten.)

Inexperience crops up again in Clay’s dismissal of 3D interactivity as a powerful medium superior in many instances to a 2D web. He runs with Rosedale’s admittedly inapt Amazon-in-Second Life example (“you could… browse the shelves, buy books”) to declare useful 3D applications as a lost cause in general. Once again, were Shirky to explore the world more, he’d regularly come across nascent or prototype experiences which already hint at how 3D interaction could indeed become an invaluable resource to numerous real world fields. He’d see applications in, for example, retail shopping (as here), online gaming and entertainment (as here and here), data visualization (as here), national security (as here), international relations (as here), non-profit fundraising (as here), architecture (as here), scientific simulation (as here), education (as here and here), and therapy (as here); just ten industries worth billions of dollars, which could potentially impact hundreds of millions of Internet users, quickly culled from my bookmark cache– and that’s not even mentioning the as-yet-unproven applications which have already gained traction, like in-world celebrity appearances (as here), political activism (as here and here), and marketing/brand promotion (as here.)

All that in mind, it’s hard to comprehend Clay’s analogy of Second Life to LambdaMOO. I have to ask: in what sense except the ancestral one does a text-only online world have any meaningful relation to an immersive 3D world with an internal building and scripting system, in which users can stream audio and video, import and export data from the Web, retain IP rights over the content they create, and easily exchange the internal currency for real cash in an economy with total transactions already in the several millions per month?

For Four Hype Busters, Four Facts

In his essay, Clay Shirky offers four hypotheses for why there’s so much attention to Second Life, so let me offer four actual news items, to suggest a counter-narrative:

- Pierre Omidyar and Jeff Bezos separately invested in it, while Mitch Kapor was a founding investor.

- When he was Microsoft’s technology evangelist, Robert Scoble declared “Second Life is an operating system” and that “Microsoft needs to pay deep attention to it”.

- IBM is investing $10 million to develop a technology lab within it. (Note that they are not investors in Linden Lab the company, but just like Anshe Chung or any other successful Resident, are buying land virtual land and finance in-world development they plan to do there.)

- For the last two years, total user hours and number of land owners, along with economic activity, have been growing at geometric rates.

Second Life is not a YouTube-level phenomenon, and I personally suspect it’ll be quite awhile before it reaches that stratosphere (say three years), let alone become the Net’s next generation (say ten years). From my vantage, however, I’d say Linden Lab’s hardest hurdles toward those milestones are not waning hype, as Clay evidently thinks, but the inherent limitations of their own architecture.

Can they really build a fully streamed world comprised of tens of thousands of servers? That’s way above my paygrade, but I’ll guess that challenge fits under the rubric of Fricking Hard. Can they fix a profoundly unfriendly user interface and thoroughly disorienting first hour user experience, which are aggressively, almost intentionally unwelcoming to the vast majority of interested users? Both shortcomings are at the heart of Second Life’s poor rentention rates, but neither have significantly changed in the three years since its commercial release. You have to wonder, whatever their stated intentions, if Linden’s tech-centric corporate culture simply puts their improvement at a low priority. (I have a dream, and in that dream, Linden Lab developers are locked inside 1100 Sansome with their own mothers, and not let out until they’ve improved the interface sufficiently enough so that their dear moms can easily use it.)


Still, the world keeps growing, and shows no sign of plateau. I mentioned SL demographic expert Tateru Nino, so I should close this out with a graph she crafted for me last night. Tateru’s chart is the growth rate of regular users, based on current account growth and 15% retention rates, and it shows how large Second Life will meaningfully be, same time next year:


Will Clay Shirky count 1,175,000 active Second Life Residents as hype, or something closer to his “Immanent Shift in the Way We Live®”?

That will, I suppose, have to wait until next December. Whatever the case, I do hope Clay gets his way, and reporters only cite that figure of active users– and not the seven million accounts it’ll take to reach that number.

*Clay Shirky’s first-hand experience with Second Life (e-mail to the author, re-posted with his kind permission):

I’ve tried Second Life three times, in different incarnations. I got a pre-launch walkthrough of Second Life from Mitch and Philip several years ago at PC Forum (in 2002, I think) which was more of a conceptual demo than anything. I used the service in 2004 for a bit, to see the state of play in comparison with There Inc.– there were not many users then, but the tech was obviously getting good.

Most recently, I logged in once over the summer and once a couple of months ago with a new avatar each time, but after a short orientation period, I realized that my use of SL wasn’t actually what I was interested in. In the same way that I am too married and too employed to have much use for Friendster et fils, my own reaction to SL is irrelevant.

And this, I think, is the key point of that piece — I am not criticizing the in-world experience, or wondering why anyone would spend their time in Second Life doing X. I’m too old a Usenet hand for that; after pouring two years of my life down the sink of alt.folklore.urban, I’m not one to pass judgment on the experiences other people find engaging. But the target of the piece isn’t the users, it’s the press, mainly, and Linden Lab as enablers. Authenticity of individual user experience is beyond external criticism, but social cues about engagement or utility are not. In a network with no gatekeeper, social judgment is our first-order filter, and perversion of that judgment is therefore a serious risk. My question is not “What is exciting the passionate users of Second Life?” or even “Where are those users taking the platform?” Those are both interesting questions, but irrelevant to my current concern. My question is a lot simpler: “How many passionate users are there, both in absolute numbers and as a percentage of the whole?” A related question is “Why is that number so hard to get to, and why is the press (wilfully? cluelessly?) reporting logins as a metric for those users?”

-clay

By Wagner James Au

You're subscribed! If you like, you can update your settings

Related stories

  1. There’s a new key they added to the keyboard: It’s called “enter” and is used to add paragraphs to articles.

    Share
  2. How about not-bold text, proper paragraph usage and only excerpts on the home page?
    Sorry Om, but I think you’ll need to spend some more time on this project ;)

    Share
  3. I think you’ve misread my essay. You spend quite a lot of time noting that I’m not qualified to critique the Second Life experience. This is true, and an observation I made to you myself before you published this. It is not, however, an observation related to my original piece, because I did not then and would not ever critique the Second Life experience. Nothing I could say about the user experience, pro or con, would matter much; I simply don’t trust myself to get it right.

    A corollary is that I don’t trust you to get it right either. The only people I actually trust are the people with the most at stake, which is to say the people who have tried it. I want to know what they think, at least in aggregate.

    I am not a Second Life skeptic, in other words, I am a Linden Labs skeptic. I might be a Secodn Life skeptic if I could see the adoption figures, but I can’t, so instead I want to know why LL is publishing bogus numbers and hiding real ones.

    Your focus on what you like about SL, irrelevant to the question of adoption numbers, is the source of a number of misreadings of my original question. It doesn’t matter that users are making their own press releases — that’s the blog revolution, not anything particular about SL. I’m not willing to regard access to auction markets as a proxy for liquidity, because liquidity is about supply and demand, and I’m not sure about the demand for those goods. My inclusion of Copybot was unrelated to user reactions, because I don’t care whether people were panicked or delighted; the observation about Copybot was about the press, and not the users.

    So let me say, once and for all, ignore me if I ever start talking about the in-world experience — I haven’t before, I’m not now, and I won’t in the future.

    Having gotten that out of the way, let me also say how disappointed I am that you added nothing new to the key question: how many users come back? I’m perfectly capable of reading the published version of the simultaneously logged in numbers, as is everyone, but those numbers are are useless, as they mix real users with ‘try it and bail’ users. My question was and is “What is the number of simultaneous return users?” I don’t have those numbers. Do you?

    Similarly, I am capable of compounding 15% interest — real estate calculators are a dime a dozen — but you don’t say anything about how that 15% figure was derived. What, in your definition, is an active user, and how are you measuring them? Because either you don’t have any different numbers than I do, at which point you’d do well to be more of an LL skeptic than you are, or you have numbers I don’t but aren’t sharing, which makes me wonder why. Given the effort you put into this post, I wish you’d taken the original question more seriously, instead of trying to change the subject.

    Share
  4. Sorry guys, the original post accidentally got deleted. i am working with wagner and fixing it.

    Share
  5. Wagner James Au Monday, December 18, 2006

    My apologies for the WordPress burp, but the post should all be restored now.

    Clay, I did address the “What is the number of simultaneous return users?” question:

    This month, as last, in-world concurrency has regularly been 15,000-18,000 during peak hours (barring downtimes, a quick glance at SL’s homepage between Noon-8pm PST will show that), and given the new account creation rate (about 12-14K total throughout the entire 24 hour period), far less than a thousand of them are new users at any given hour, on average.

    Share
  6. Hi Clay,

    I know you are a very busy person so here are some highlights from THIS article that contains some answers to your questions as I see them.

    “What is the number of simultaneous return users?”

    The answer, according to this article, is roughly 15,000 conservatively. I believe that.

    If you had asked “What is the number of regular return users?”

    The answer is, according to this article, roughly 250,000. I am less sure of this number. In a comment on your article, I said that I think the number is probably closer to 100,000.

    I agree with your desire to get down to the bottom of this.

    Here is a modest proposal. Why don’t we collectively come up with an acceptable definition of what we consider to be a “regular SL resident”. Once that definition is agreed upon, we present it to LL and ask them what that number, according to our definition, is. They’re under no obligation to provide that information, but I think it would mean a lot to a few people if they did.

    What do you think?

    Best regards,
    Eric

    Share
  7. Bravo on the point that the wrong things are being hyped too much and the right things aren’t being hyped enough.

    About time someone put it out there in a direct and non-acidic manner. :-)

    Share
  8. Among this “hype or not” discussion don’t you get the sense that the people who matter most (the folks who use SL) are too busy to comment because, well, they’re actually enjoying Second Life?

    some assorted thoughts: http://elapsedtime.blogspot.com/2006/12/worthy-of-debate.html

    Share
  9. I’m skeptic about second life first why aren’t there any negitive comments looks like large censorship. Also most people that plays it is large corporatation trying to sell in a virtual made up economy.

    For a normal gamer, person, I think first life is complicated enough. This seem like a fad, prymid scheme, lure people into thinking they are making money. I mean why whould any living person live in a virtual world and dieing in the real world?? people have to realize you aren’t going to make money off a system where the producures are the ones laughing on their way to the bank.

    Share
  10. As a pretty reliable rule of thumb, in a large multiuser application the number of total users is about 10 times the number logged in simultaneously at peak periods (but not exceptional peaks). What this means is that, unless you are dealing with accountants at closing, when nearly everyone is on the system at once, you can generally multiply the simultaneous sessions by 10 to get your user population. In SL’s case that gives us about 150,000 to 200,000 regular users, which falls between your two estimates, and well below the journalistic/hyped figures of 800,000, 2 million, etc.

    May Supreme Avatar Smile Upon you!
    Br. Taras
    Prelate, The Avatars of Change
    Great Avatar Hall
    Rieul, SL

    Share

Comments have been disabled for this post