Forrester Analyst Suggests Second Life Stymied by "Iron Law of Oligarchy" - He's Half Right

At the last Second Life Community Convention, Chief Product Officer T. Linden gave a milestone presentation on the new SL viewer, aimed at making the Second Life experience more intuitive and user-friendly. Created by veterans of Adobe, TiVO, and other companies renowned for great UI design, it was an impressive and exciting unveiling.

And then T. opened his talk for audience questions. First one had nothing to do with the viewer. Instead, a gentleman demanded angrily, "When are you going to make it possible for us to belong to more than twenty groups in Second Life?" (Or words to that effect.) The rest of the questions for T. from SLCC attendees were largely in that vein, highly detailed concerns about outstanding bugs and features, which while important and valid, had little or nothing to do with the new Second Life viewer.

Tom Grant, analyst with the esteemed Forester Research firm, has a thought-provoking post which provides a name for this phenomenon, and suggests it's holding Second Life back:

[T]he Iron Law of Oligarchy, a small and often unrepresentative sample of users [who] wield disproportionate influence over the vendor's thinking about products and services... Second Life is an extreme case of how you can develop a very happy group of customers, and still fail miserably at reaching a wider audience. Some businesses are comfortable with that outcome, as long as the customer base stays loyal, and the business stays profitable. Most would be terrified to discover that their best customers are, in subtle ways, holding them back. I can't say for sure that the Second Life notables are the reason why the UI is still klunky, and the useful content is hard to find, but I definitely have my suspicions.

As the incident at SLCC suggests, there is some truth to that observation. I've seen a similar phenomena on this blog, in many of the responses to my editorial series on making Second Life more mass market:

To a significant degree, the distinct tenor was active and almost angry resistance to the features that might make Second Life easier to learn, and more accessible to the tens of million who regularly play virtual worlds, and the 10+ million who downloaded the SL software, but found it too intimidating and confounding. As one reader put it succinctly, "Don't dumb down things for the riff raff."

But Grant is only half right. For every Second Life user who could scarcely care less if SL fails to go mainstream (even if that ultimately leads to SL's decline), there's a plethora of content creators, educators, enterprise users, and many more who want Second Life to get big, and indeed, have a deep personal and professional stake in seeing that happen. (Including, well, me.) Consequently, Second Life's userbase also devotes more energy and effort to growing and improving, and defending itself from outside critics, than just about any other Internet product. That's a rare and precious thing.

So the gentleman at SLCC complaining about poor support for SL groups isn't wrong. He's got a legitimate complaint, one that should be addressed one way or another. But as I've tried to argue, the best way improve Second Life on the granular level is to first improve it on the macro level -- make SL relevant not only to the hundreds of thousands who've already embraced it, but the millions who also would, given the right nudging. Then it's not just the Lindens trying to improve Second Life, but all the many more companies and individuals interested in thriving within it. Then that oligarchy will grow large enough to seem more like a republic.

Comments

[Ann Otoole]

I'd say the problem resides entirely within Linden Lab myself. If you want to grow you need to figure out who within likes the current status quo and does not wish to lose the sense of power and control they have. LL's behavior resembles a fiefdom more than an oligarchy anyway. You have to study the behavior over time. Like how if you want something taken care of you have to become a courtier and then once you attain status enough to be heard then you can attend court (office hour) and plead your grievance to the nobility (Lindens).

The symptom of developing a new UI without actual customer involvement is pretty interesting and meshes in with this article nicely. Where is that new client anyway?

But nothing much is going to change.

Interesting article though.

[Toxic Menges]

I think Mr grant has hit the nail on the head. Linden Lab find it difficult to talk to anyone but their super users, because only their super users take the time to read not only the Second Life blogs, but any blogs or forums regarding this or any other virtual world.

Super Users find it difficult to fathom why other users have an issue with a learning curve, or they don't see it as an impediment - because they have forgotten their own first tentative steps in using the Second Life client.

As far as having a disproportionate voice, yes they do, if you take into account my first point, any language barrier, against the +ve financial investment, and length they have been users. They know where to shout, they are not afraid to, and they know the product inside out.

The more mainstream something becomes, the more simplified it needs to become to enable enable mass use.

The more mainstream something becomes, the more cliquish those who were around in the "old days" become. Of course there is resistance, whilst common sense would dictate that business owners and content creators should be welcoming the opening of the floodgates, there is always a hostility of some kind towards those who will bring about change.

Second Life, and in fact other virtual worlds go against the grain of the western world in as much as age is revered here. An oldby is respected, quite apart from real life where the old are regarded as stupid, and not to be listened to. Whereas by contrast. the young are ridiculed and laughed and and are not made as welcome as they could be.

The problem is inherent in almost every community there is, online or not, but certainly much more visible in old communities on the internet.

One of the first things any leader of a community is taught is that you will never please all of the people all of the time, this is common sense of course, what they should also be teaching them is that the voice of the top 10% is very rarely the voice of the rest of the community - they merely understand how to manipulate their voice best.

Linden Lab need to be strong enough to press on for the good of all their users and look to engage those who aren't super users to convert them to that status, and also those who also those who have already fallen by the wayside or who have yet to visit.

Those with a voice will adapt, and stay loyal - but they will always chatter the loudest - the trick is to make it possible to hear all the voices, not just the loudest.

OnRez used to have a Welcome box with info and events - a simple web page that popped up .. this is the perfect way to start to get communication started with everyone.

[Ordinal Malaprop]

The idea that the Lab are vastly influenced by a resident oligarchy I don't think is true - obviously there are influences, but if the Lab is making decisions based on the goals of old and valued residents, well, these must be an entirely different set of old and valued residents to the ones that *I* know.

Mr Grant is I believe far from the mark in connecting this to the User Interface in any case. As we all know there was a policy decision after open-sourcing the viewer to basically abandon any changes to UI in the Client (or simply changes to the Client generally, apart from those required to use new grid-based functions) I assume with the expectation that third-party viewers would fill the gap.

Unfortunately, third-party viewers have not done so, concentrating instead on adding _functionality_ rather than simplifying the interface (to be fair, a far more significant challenge than radars and object exporters). They are generally clients for "power users". After all, who writes them? Mostly, experienced individuals who know the existing client but wish it would do more. What personal motivation would those people have to code a viewer with a simplified and cleaner UI for new users, when they are already familiar with it and so are those they know? If they had a special business purpose in mind they might (e.g. the CSI viewer) but that would not result in a generic client for new residents, but one tailored for a specific project or area.

[Arcadia Codesmith]

Development is always a balancing act between "fix and refine what you've got" and "give them something new". To be a continuing success you've got to do both, do them well, and do them at a pace that will stretch your resources (however vast or limited they are) to the breaking point.

The shiniest lure in the tackle box won't catch any fish if your pole is broken.

[Riven Homewood]

Despite all the complaints about SL's "learning cliff" and "clunky interface" the existing interface does an amazing amount of complex things in a fairly straightforward manner.

Yes it takes a lot of hours inworld to learn how to use it effectively and yes it seems quite overwhelming at first. But so does mastering any piece of powerful software.

I've been dabbling with Photoshop for years, and I still can't do much more than crop a photo. I know people who are true Photoshop experts, and they got that way by spending a lot of hours playing with it and learning how to use it effectively.

One result of all this effort is that once you begin to master the SL interface you feel a lot of buy-in. I'm convinced this is one reason SecondLife users are so fanatically loyal to SL, despite constant problems with the software and policy reversals by LindenLabs. After scaling that learning cliff, you feel like this is your world. And chances are you are the kind of person who makes a good neighbor in a virtual world.

SecondLife is all about community. It's all about the wonderful people you encounter there and the marvelously-creative things they do. LL appears to ignore community-building features like a reasonable limit on how many groups you can belong to and a reliable means of communicate among group members, focusing instead on flashy graphics and "the new-user experience". Regular users encounter this over and over during the course of year, so it's quite natural that they would ask about it when given the opportunity.

[Maggie Darwin (@MaggieL)]

The established residents don't have any problem with "macro level" improvements as long as sufficient attention is still given to fixing the "granular level" defects. And right now they aren't.

Marketing isn't enough, you must execute as well, and LL is failing to execute on a grand scale. Linden Research must fix what's broken. The twenty group limit has been complained long enough to now be properly seen as a design defect. There are multi-year-old significant JIRA items (group chat lag and failures, for example) that still remain unfixed.

Many of us will consider the Linden Research product to be a failure to the extent that known defects remain unfixed, no matter how many big changes that some consider "improvements" are made.

By all means, show prospective users lush machinima of happy shiny avatars sailing an infinite sea in their newly-purchased yacht. But what will their "user experience" be when they finally get here, earn enough L$ to buy that yacht, rez it, sail up to a sim boundary and then get orbited, not by a griefer, but by LL themselves?

Why wonder why retention figures are lousy? And then blame it on the volunteer mentors?

Recently Linden Research appeared on somebody's list of top-20 SillyValley start-ups. Hey, guess what, Lindenpeeps...IT'S 2010: YOU'RE NOT A STARTUP ANYMORE.

Time to start executing.

[Maria Korolov]

Riven --

I have to agree with the Forrester analyst. In order to get my employees, colleagues, and casual friends into a virtual meeting the entry process has to be super simple, Web-based, and fast. They don't need building tools. They don't need to change their clothes. They don't need to edit scripts.

If they decide they like the world, and want to do all of these things, then they can download a full browser and learn how to do it. But they should be able to attend a meeting without having to spend two hours learning the interface.

However, there might be another reason as well. Casual users -- ones who don't create things -- aren't likely to rent land or buy stuff, thus contributing to Linden Lab's bottom line. If there was an easy-to-use Web-based viewer for Second Life, we might see a flood of people coming to Second Life for casual get-togethers and ad-hoc meetings, putting a significant strain on SL servers without a corresponding increase in revenues.

As a result, we might see more progress on the Web end with OpenSim, where companies own their own grids -- and are much more anxious to make the platform more user-friendly and accessible. Virtual meetings improve training and collaboration and cut travel costs, making for a clear ROI -- if only the meetings were easy to attend.

- Maria Korolov
Editor, Hypergrid Business

[Ordinal Malaprop]

I might add a paragraph that I thought of when replying on the Forester blog:

> If there is a problem it is that SL throws new users into an open-ended "sandbox" world with barely any training, and expects them to work out what they want to do next. This makes the UI seem far more confusing because you don't even know what you're supposed to be doing with it.

[Ananda Sandgrain]

While I would concede the notion that as a long-time user I don't understand the issue with the viewer ("hard"? this thing is waaay easier to learn than, say, Word or Skype); I don't think the rest makes any sense.

If Linden Lab wanted to behave like an oligarchy they could turn around the whole issue of the first hour experience quite rapidly. Picking favorites would allow them to get new users immediately into an item shop of the best and brightest stuff on the grid, pre-assembled into easy-to-use package on a standardly-shaped avatar. If the first hour operated more like the avatar-customization features in mmo's or games like The Sims and then dropped people into sponsored venues it would improve the first hour immensely.

But you see, to do that, the Lab would have to play *actual* favorites and form oligarchic partnerships with their biggest and best content providers. From where I am it looks like just the opposite, and that the problem is all new users are dumped into the middle of a free-market anarchy.

[Metacam Oh]

If the lab managed to get the users, they'd be dumped into an area with 20 avatars and fail to be able to move around.

[k.d.]

It was amusing to read the quote "Don't dumb it down for the riff-raff." That's sounds a lot like the comments by computer professionals in the 80s when office PCs started to take over from mainframes. Sounds like someone is total denial.

[Tessa Kinney-Johnson/Harrington in Virtual]

SOME POINTS OF INTEREST:
1. The number of groups allowed IS INDEED a client and server side issue, so that would definitely be a valid question in a presentation about said say.

2. The reason these type presentations by LL get hijacked by the users attending is because so many major design flaws have not been addressed in over seven years of development - a serious sign of a development team that is completely and utterly lost touch with its audience. If old issue were either resolved, OR if valid reasons were given as to WHY they could not be resolved, then they'd not keep cropping up inappropriately. Its all about communication, which LL still hasn't grasped.

3. When the majority of your staff has rarely if ever 'played' and/or interacted with the community they are designing for, how can they possibly understand the importance of resolving the problems the community cries out about?

4. Its been said that if someone was to draw a cartoon of LL and its loyal users, it would depict a room full of coders bowing down to the head of LL as if he were a God, while its user base picketed out the in the parking lot, saying "What About US?!?!"

The crazy thing is, LL has tried to change, listen & grow but I'm afraid its too little too late. The masses simply have no faith in LL doing the 'right thing' by them, be they business, casual users, loyalist or creators. Such failure is a HARD thing to accomplish. Takes a very concerted effort to create that level of distrust, which doesn't exactly bode well to the future of the Second Life we've all grown to view as more than 'just a game'.

[Melissa Yeuxdoux]

I chatted this morning with a new friend and asked him what he thought about SL's user interface. He said it stank. Despite that he's very enthusiastic about SL. I have to wonder whether currently there has to be something about SL that motivates the new user so much that it's worth the hassle of the user interface.

I'd hope that the SL client is designed such that it's easy to rip out one user interface and replace it with another, so that a new UI can be created without denying us old fogies the interface we know and, uh, know.

One very good point my friend made: these days, nobody would use a text editor or word processor without an undo feature--why do we have to do without it when editing objects in SL?

[Tessa Kinney-Johnson/Harrington in Virtual]

YOU do have an edit undo feature when editing prims, Melissa. You use go up to EDIT/UNDO up by FILE on the top toolbar. Its always been there and probably always will be, unless they've taken it off with the new client, which I really doubt. Hope this helps make your SL experience better!

((Hugs)) Tessa

[Opensource Obscure]

Suggested reading:
Hard But Smart Decisions: Blogger Tackles The .5% Problem
by Louis Gray, a smart and insightful web commenter and analyst.

His piece is about Blogger, the Google's blogging platform, and how they're managing the support for a legacy feature - a feature that if removed could possibly negatively effect a loyal, but very small, population of users.