At the Gamasutra-attended Virtual Goods Summit 2009 in San Francisco, reps from leaders in the microtransaction-based market, such as Nexon and IMVU, discussed the nature of a successful virtual goods market based on first-hand knowledge.
Attending the panel, which was moderated by Lightspeed Venture Partners' Jeremy Liew, were Nexon America (Maple Story) vice president Min Kim, Dai Watanabe from DeNA, which runs Japanese mobile games portal Mobage Town, Cary Rosenzweig of U.S.-based virtual world IMVU, and Akio Tanaka, of Japan's Infinity Ventures, which exclusively and globally funds microtransaction-based products.
The Product You're Selling
The first question posed to the panel was the nature of the items themselves. What makes a good virtual product? Says Kim, "Everything we sell is functional. Whether it's a piece of clothing for your avatar or it's an item that helps you earn experience faster, there's some sort of function there."
Rosenzweig notes that though certain items are popular, what IMVU's users crave is choice, and that drives long tail business. "The top 10 items for people dressing up their avatars... represent 0.2 percent of sales of items. Our users demand to go deep, deep, deep into the catalog."
Before you talk items, it's about getting users into your network, and getting them to pay. Watanabe maintains that user acquisition is key to sales. Mobage (short for Mobile Game) Town pulls users from its simple games into a broader community. "The key is how we can manage people to move from the game to the community. The game is what drives conversation among users."
"Once we successfully bring that traffic from the game to the community, we have avatars," says Watanabe. There are three key points to generating sales of items, he says -- what's in fashion (which requires constant updating), delivering rare items (which can be sponsored items, such as a Coca-Cola branded T-shirt), and giving users the chance to get new items from old items.
Rare items in Mobage Town are often delivered via a vending machine -- put in credits, and you have a small chance of getting a rare item. More recently, the service has introduced a concept called "avatar mixer", which allows users to drop old items and some money into a machine and get rewarded with newer, rarer items.
But the majority of IMVU's items, unlike in most MMOs or even many other non-game virtual worlds, are user-created. Says Rosenzweig, "Once users buy credits from IMVU and they're inside IMVU, we leave them to buy [items] from other users." IMVU considers itself "the eBay of virtual items," in that sense. He notes that the staff does pick items to promote to its user base -- but "when we try to pick the winners, we don't always succeed." IMVU attracts multiple subcultures, such as anime fans and goths, so items may not be broadly popular across the entire user base.
Tanaka notes that tracking player motivation on item purchases is important. "People spend a certain amount of money to express themselves, but they spend even more to impress others," he says.
"War has the highest ARPU," says Tanaka. People will spend even more to win in a conflict -- a Chinese colleague once said to him, "the ultimate motivation is to terminate everybody on your screen."
Watanabe notes that the more you obfuscate the exchange rate of real money into virtual currency, "people stop thinking about how much the currency is worth."
As an experiment, says Rosenzweig, "last month we did a sale for credits for only our VIP subscribers. We did a 50 percent off sale for one day... and revenue for the whole business tripled for that one day." Most importantly, this didn't seem to be people banking cheap credits ahead for a long period of time, as you might expect. "There was a slight slowdown on the next day or so but then it went back to average." This, coupled with an item sale, created an "upsurge in buying in the catalog."
Tanaka notes that what's most important is making sure that you "don't underprice your items, because you may find much more upward mobility in your pricing range." He cites an example of the most popular farming game in Japan -- the operator wanted to start credits at approximately $1, but he counseled them to start their range at $5 -- "and it worked."
Flavors of Payers
Liew asked the panel about their top buyers -- a question that seemed to make the panelists uncomfortable. While nobody wanted to state it directly, there seemed to be a consensus that unhealthy or socially unacceptable motivations, like hoarding or seduction through gifting, are behind the behavior of some of these users. "The top buyers, if you look at the number you can just worry about if this guy's okay," says Watanabe.
Kim calls them "outliers." He doesn't target these individuals, Instead, Kim says, "What we're not trying to do is target a high ARPU. What I'm looking to do is have a lot of users paying something a lot more healthy a month." He suggests that "20% of your paying users will probably make up 80% of your revenue."
Tanaka said that gifting is a "huge part" of virtual item purchases. Stats revealed that 29 percent of men and 21 percent of women are buying items only to gift them to other users. "Virtual goods are going to cost more than real goods in some of the markets. And that's a real trend," says Tanaka.
The subject of offers is currently a hot topic in virtual goods circles -- rather than having players pay cash for items, you allow them to earn credit through offers such as free trials of Netflix -- which will see the video rental service paying the game provider directly for the acquired users.
Says Rosenzweig, IMVU experimented with showing its users a five minute ad -- and pegged the one credit payout, worth half a cent, at different lengths of time. At first it was the first five seconds; that paid out almost instantly. Thirty and 60 seconds also paid out quickly. Even two and a half minutes generated payouts -- at which point, "you are valuing your labor at 12 cents an hour," he says.
Making it Work
IMVU has jumped up in revenues, says Rosenzweig, because "in 2009 our focus was lasered in on increasing the number of people who paid us money. The first is the fundamentals -- we were improving the product. We had a big project to fundamentally change the user experience. The day we launched that, our core metrics jumped." The other factor? "The payment products -- once we launched them. People how they want to pay how they want to pay."
Though Mobage Town is mobile-only in Japan, the dynamics that drive it could map to systems like Facebook, it seems. Says Watanabe, "Eighty to 90 percent of our users are coming from word-of-mouth. We are giving some amount of virtual currency for a friend invite, but it works well because it's easy to pull up the contact list. And it's very easy to track these links."
When it comes to targeting specific markets, it's more than just the volume of users that's crucial to track, says Tanaka. "Japan and China have large internet populations but the value of users are very different." China has huge numbers, he says, but the average rate of activity is very low. Japan has fewer, but the users are very active.
"I would say that Japan's active rate is higher than Facebook," he says, and advertising to acquire users is worth it in that market because the users are so valuable. DeNA, operators of Mobage Town, paid "$5 per user" it acquired via a marketing campaign, says Tanaka.