Virtual Goods Summit 2008
Moderator: Margaret Wallace (Rebel Monkey)
Panellist: Brian Balfour (Viximo)
Panellist: Lee Clancy (IMVU)
Panellist: Amy Jo Kim (Shufflebrain)
Panellist: Sean Ryan (Meez)
There's a a third type, "Branded UGC" - where users who create content start to become known-brands in their own right. This has already happened a lot in Second Life, it's not surprising to see it happen in other content-creator-heavy, environments.
It seems the big players / winners so far are still taking a very "experimental, unplanned" approach to the fundamental worrying parts that keep newcomers awake at night: what goods should I be selling? what pricing should I offer them at?
My commentary in [square brackets]. This is not a literal transcription, I've elided, ommitted, and expanded things for clarity.
RM - [after getting a show of hands] About half to two-thirds the audience have or are launching a goods-based game.
RM - In general, what processes do you use to decide what to put into your inventory?
Viximo - we focus on connecting content-creators with publishers. We leave it up to the creators, and give them support tools and hints/tips to do that, prompt them around holidays to create themed content for that holiday, etc.
Tools, e.g.: mostly the info we feedback about the buying activity; what types of people are buying what tupes of goods, whats popular, whats good.
Also, Scarcity: creators decide desired high/low inventory and price. Internally we decide what inventory level when we look at that.
IMVU - 120m items in the catalogue. We take the view that anybody at some point could become a digital-goods creator if you give them enough support from community perspective AND hands-on training/tutorials.
Meez - we start with the holiday theme. Then look at what the advertisers want to see. Then we look at the upcoming features, and make sure we have compelling items there. A lot of our prod dev driven by trends/fads in the userbase - pop culture influences etc.
RM - what challenges have you found from UGC?
Shufflebrain - "quality" and "appropriateness" are the biggies. At There.com we had both branded content and UGC. We thought the branded would be much bigger, but it turned out that UGC was much more exciting among users.
Copyright, porn, hate-speech etc - you have several different ways to deal with that. Pre-moderation, user-tagging of problem content - it's a management issue both legal and social/cultural for the product design. [basic explanation of trivial rating systems here]
When people come into the world for the first time, you want them to see great stuff, you don't want them to see crap.
Its not "branded vs UGC" in general - it's different for each game / prod design. Choose the right one for you.
It's about: who are your audience, what does "status" mean to them, and how much skill do they have at content creation?
IMVU - we also see a third type, "Branded, UGC". A lot of our users move from being CCs to being Branded CCs as they build up a significant brand purely within the IMVU world.
Meez - what is a VG? We all talk about it like they're just clothes, but that's just one third. Another third is world-features (can I levitate, can I glow like a lightbulb), and final third is privileges, access - "can I sit in a special seat in a public space?" etc.
In these worlds, often you're only about 2 inches high, so to differentiate, you end up going more extreme in clothing. And then there's not all that much you can make that varies that extremely.
[Adam: if you're making more extreme clothing, with more stuff sticking out the sides etc, specifically in order to allow differentiation, surely that leads to giving you far MORE options in terms of items to create?]
Ultimately all these items are about generating increased status.
Viximo - have to think of CCs as different from normal users - they are a different community, with different ideals and aims and activities.
IMVU - have to be very clear on your policy. We have a very clear policy on what is allowable etc.
[Adam: there are already initiatives in progress on this, from Erik's Better EULA to Raph's Metaplace policy; it would be good to see more sharing between virtual worlds, more standardised policies, so users only have to read and understand ONE of them]
RM - what tools does IMVU offer to support UGC - issues of ownership (operator or player)?
IMVU - any products that someone creates derive from a core base product originallly created by IMVU. And then we track the chain of people creating from, creating from, creating from, ... etc. We then pay everyone along the chain (every person who's work got derived from gets paid).
[gaming this - what happens with two-person cycles?]
Shufflebrain - smilebox, which is online scrapbooking. Kind of like slide/rockyou for scrapbookers. There's things like this that are much more connected to / derived from the real world (as opposed to pure VW worlds) and that's part of this VG spectrum.
Meez - if you have a community of any size and you aren't in the VG market, then you're missing the boat in a big way.
It shows status, it allows gifting. It's the secondary model that combines the best bits of advertising. It's great way of monetizing the always-on hardcore, while you're simultaneously monetizing the masses via simple advertising.
IMVU - 90% of our revenues come from VG sales.
RM - how do you determine pricing?
Meez - pricing is easy, we look at Gaia, and copy it .
In South Korea, it's much more mature, so we look at that as a place to take input on pricing [???]
3%-8% of users will ever buy stuff.
We started at $6 / month for VIP package. We could have done $10/month and it would have been fine. [which begs the question: why are you voluntarily forgoing an extra 60% revenue? - one of the panellists came back and asked him this later:] we still haven't decided exactly how much to charge, we're still adding features to it, it's not feature complete yet.
We've noticed that if something doesn't have a status-element, it sells more poorly.
We have no secondary market, its all manually priced by us through trial and error.
The value of branded items within the system isn't always clear.
We had one body-type originally, only "thin". People complained that it was too thin, body dismorphia.
So, when we added more, we found that the other sizes were very popular:
17% of women choose "plus" (biggest size)
80% of men choose "buff"
Shufflebrain - people have been traditionally choosing just fantasy identities in the online games, but it's moving towards people choosing identities that represent themselves. So, now that's another big question you have to think about: which kind of identity will you service for your users with the kinds of goods you sell?
[Adam: IMHO this issue of "which identity are your users creating/using on your service?" is one of the few that advanced operators in all online games have been thinking about for years now, but many haven't - strangely it never comes up at conferences (I suspect that the people who know are keeping quiet about what's a clear competitive advantage to get right )]
Meez - why not photos? Too personal, too unchangeable. Avatars are much more flexible, changeable.
[Adam: interesting; the ease of changing/updating photos is changing already, good cameras are becoming ubiquitous (inside the DSi, for instance) and fast to upload to flickr etc ... how long before photo-avatars replace current ones?]
RM - as service provider, what effect is the financial crisis having on your clients?
V - in the VG space it's moved from "what are VG?" to "how do I do it?". It's been recognized as a much better way to monetize virtual communities than most other things.
So demand among publishers is currently very high for VG economies/systems.
We're also going after creative individuals, getting them to go direct to consumers, instead of being part of a bigger chain. Totally different dynamic to what they usually do.
I - we just added profit-based pricing: you say how much margin you want to get on the item you've created. This empowers the content creators [Adam: I'm not sure how that "empowers" them?]
RM - VGs are a $1.5billion global market. [Adam: this week at the Virtual Worlds Forum Europe one of the attendees mentioned a recent research report from Finland, IIRC, that pegged it at $2.1billion; we'll follow up on this later]. how do preferences for VG varying based on demographics?
SB - Only speaking from my direct expereinces...many social VG skew female, many games skew male. Ultima Online was male-skewed, There.com we had a mix - but females were leading the charge on content-creation, especially totally dominated clothing, especially especially the top creators.
Music downloads in rockband etc - this is part of VG too. That's a good that exists in the game that opens up a new activity. I believe this has cut across gender lines (both male and female)
[Adam: Well ... the game itself is doing the cutting across gender lines, it's arguable that whether or not it's an activity is irrelevant - it's a captive audience of people who like the game]
I think this is the future - VGs that open up new activities, as opposed to merely being about status etc [this is referring back to the slide that the RedPoint Ventures guy put up at start of conference]
I - I will get hold of the breakdown between M/F content creators, but I don't know that off the top of my head
Women are buying lots and lots of clothing items - "content consumption" skews as well as "creation".
M - need to consider that are limits to how many items you can make use of. So think about group-purchases as well - e.g. in Puzzle Pirates when you buy a ship, you also have to buy badges for each member of your crew (other players of the game who won't buy them themselves).
RM - how do international users influence your actions?
I - this is an English-language product, with very little internationalization, but 40% of our revenues come from international. When you build something good, people are going to want to be a part of it, wherever they are.
We're very clear that we're under US jurisdiction etc.
M - it's hard to monetize non-US traffic - but it's still EASIER with VGs than many other things.
e.g. banner ads for say Nigeria aren't going to get better than $0.02 a click - betting they will is a losing bet, but you CAN feasibly get Nigerians buying and selling VGs at much better monetization levels.
[Adam: interesting ... places like that leapfrogging real-world consumption, straight to VG consumption?]
RM - rockband / AG numbers?
M - Kerli, an Estonian goth-pop singer, who is very popular with part of our audience. We find in general that branded goods are a great promotional vehicle.
It is not clear yet that users are willing to PAY for branded goods at a price level that's big enough to make revenue the primary use of it. The premium branded good business in the US is still small, compared to other things.
2.5 % click-out ratio for people going out the chatroom to buy the songs on iTunes or etc.
[Adam: and some others that went past too fast for me to get down ]
Q: I thought most of the transaction model was that you sell virtual currency. Currency vs items, challenges there?
I - we do sell credits, that are used to buy VG. There's also promotional credits that we give every user when they get an avatar. And there's the VIP service, which is a monthly subscription that gives a batch of credits each month as an allowance.
M - less-active users you sell advertisers. Medium users, 1 item a month purchasers. High end, you have a subs offering for the hardcore so that they can go off and differentiate themselves.
We make VIP have value more than just the credits - being a VIP is a status item in and of itself - so people subscribe AND buy extra credits.
Q: how do you decide where to draw the line on how much credits to give away?
M - look at your sinks and sources of currency, your float. Just use standard economics measures. Keep tweaking to try and keep it roughly inline to prevent inflation and deflation.
Relatively straightforward to track in and outflows on a weekly basis and keep tweaking in response.
I - a lot of biggest CCs build up so much credits from their sales that they start to act as currency traders. Provides an RMT cash-out opportunity for users / CCs, which is good.
Q: what have you done about fraud?
M - Fraud goes up as soon as you either allow cash-out, or you reach a certain size of userbase.
We have fraud but we don't allow cash-out yet, so it's really all about cheating in-game, and we just have to use great reporting / metrics internally to detect and trace it.