Lessons from leaders in virtual goods October 30, 2009Posted by jeremyliew in digital goods, virtual goods.
1 comment so far
Worlds In Motion has a good summary of my panel today from the Virtual Goods Summit, where leaders in Asian and Western virtual goods social networks shared what lessons they had learned.
Why do people buy virtual goods? February 8, 2009Posted by jeremyliew in digital goods, virtual goods.
I have an opinion piece in the WSJ.com on why people buy virtual goods today.
What could Facebook do to increase its digital goods revenue – increase Motive. 3/3 February 5, 2009Posted by jeremyliew in digital goods, facebook, gifts, virtual goods.
I’ve been posting on what Facebook could do to increase its sales of digital gifts, breaking down tactics into three categories, Means, Motives and Opportunity. We’ve covered Means and Opportunity, and today we’ll address Motive.
Let us start by understanding the motives of gift givers. Gifting serves the same purposes on Facebook as it does in the real world. Firstly, it serves to strengthen social ties. Secondly, it serves to draw the receivers attention to the gift giver. Facebook can increase motivation for gift giving by playing into these two familiar behaviors, looking to the real world for conventions that can easily be borrowed.
Strengthening Social Ties
One of the strongest conventions of gift giving is reciprocity. It is awkward to receive a holiday card from someone that you did not send a card to. So too with virtual gifts. But right now it is difficult to know who has given you a Facebook gift. Since virtual gifts are given with the context of the wall, the wall is the best place to highlight gift giving. If each time I visited a friends wall I could prominently see what gifts that friend had given me, that would increase the pressure for me to buy a virtual gift for my friend to accompany my wall posting, especially so if it was a gift giving occassion (such as a birthday, holiday etc). Of course the opposite is also true – if a friend had not given me any virtual gifts you would not want to highlight that at the point at which I was considering whether to give a gift myself.
Gift giving is strongly influenced by immediate social norms. If I were to show up a dinner party empty handed when all the other guest had brought a bottle of wine, I would also feel awkward. People look to the behavior of others to see what is appropriate for their own behavior. Once again, the wall is the right place to highlight this. Right now the wall displays all postings in reverse chronological order. Since a new wall posting appears on the top of the wall, you will only see the most recent postings on your friends wall, many of which may not have digital gifts attached. What if the top postings on the wall were those with virtual gifts attached, and then reverse chronological order after that? (Perhaps with some time limitation, so that top posts would be virtual gifts received say in the last week). This would create a sense of social pressure to a visitor to the wall who would see virtual gift giving as a social norm. This will be especially effective around traditional gifting occasions as before. By highlighting desired behavior, you can influence social norms in the direction that you want.
Drawing Attention to the Gift Giver.
Facebook can be a noisy environment. On your birthday you can receive 10s and even 100s of birthday well wishes. That is a lot of messages to sort through, and often these wishes are not responded to individually due to the volume. How can I make my well wishes stand out from the rest? How can I show how good a friend I am or how much I care? One way is to attach a virtual gift. Because the gift is not free, the very act of attaching a gift serves to differentiate my message from the rest. This is visible not just to the recipient, but also to all other visitors to the profile. Facebook could makes product changes to make this differentiation more prominent. One way would be to “pin” gifts to the top of the wall for some period, as noted in the preceding paragraph. Another would be to similarly “pin” gifts received to the top of the News Feed page for some period, ensuring that the gift is noticed and emphasized to the recipient. Finally, having a small profile picture accompany the gift, instead of just the name of the giver on the News Feed would serve to further draw attention to the gift giver.
Building from this approach, gift giving draws attention but it currently cannot draw gradations of attention. The absence or presence of a gift is the only distinction because all gifts currently cost the same (with the exception of free sponsored gifts). If Facebook were to provide gifts of different prices and levels, this would enable a gift giver to express their interest in a more nuanced way. One problem with implementing this approach on Facebook is the sheer volume of available gifts. There are over 300 Facebook gifts available today. It will be hard, if not impossible, for a gift recipient to tell what is a more valuable gift versus a less valuable gift just by looking at the gift. HotOrNot’s Meet Me solved this problem by starting with a small range of gifts with value tied to a conventional scale; flowers, ranging from the least valuable daisies to the most valuable red roses. Given the profusion of gifts available on Facebook today, Facebook would need to find some other way of demonstrating value to a recipient than relying on the image itself. Perhaps it could show the point value of each gift when the gift displays. But that is a bit crass – it is like leaving the price tag on a gift. It may need more creativity to make this obvious. Facebook could change the background color of the gifts according to a scale of value that is well enough understood: perhaps white – bronze – silver – gold? This would allow for the current large range of gifts but make it easy to tell at a glance the gradations of value.
Obviously, allowing gifts to have a range of prices will increase the average sales price of gifts, hence increasing revenue from digital gifts sales.
Motivation for Gift Receivers
Looking at gift givers motivations is only half the story. The other half of the story is the gift recipient. What are their motivations?
One simple dynamic to increase gift recipients’ motivation to receive gifts is to make gift getting competitive. Keep track of how many gifts have been received and display this prominently. This could be done on the profile page; in the same way that number of friends is tracked and thumbnails of friends shown, number of gifts receieved and thumbnails of gifts received could also be shown. Or it could be made even more explicit with leaderboards for the people who have received the most gifts. The power of displaying metrics to drive behavior is well documented by game designers. (If you haven’t read Amy Jo Kim’s work on game design for social environments, you should). Once people want to receive more gifts, they will start acting in ways that encourage gift giving, whatever that might be.
Game design provides a second possible mechanic to induce gift recipients to want more gifts; collecting. Gifts are all treated the same right now. If Facebook were to offer awards and achievements for getting “sets” of gifts, you would most likely see some users work very hard to collect gifts to complete those sets. PackRat has shown just how powerful and addictive collecting behavior can be on Facebook. Facebook could for example offer a free [birthday cupcake] to give to someone else if you were given five [birthday cupcakes], or put a custom Christmas skin on your wall if you received 10 Christmas themed gifts.
I believe that through increasing the motivations of gift givers and gift receivers, Facebook could see a more than doubling of their virtual gifts revenue.
We are looking at how Facebook could increase its digital goods revenue by improving the Means, Motive and Opportunity for users to buy digital gifts. Yesterday we looked at Opportunity. Today we’ll look at Means.
Last year Facebook switched from denominating gifts in dollars to gift credits. This was a good first step as users tend to be more willing to spend virtual currencies than real money, even when they are readily interchangeable.
Currently there is only one way to buy Facebook gift credits, and that is via a credit card. But a lot of Facebook users don’t have or don’t use credit cards. They may want to be able to buy and give gifts, but they can’t do so. This is a common problem for a lot of game developers, including many game developers on Facebook. The techniques that worked for them can work for Facebook too.
As a start, Facebook could enable additional payment mechanisms, including Paypal, cell phone billing (including premium SMS) and direct debit from checking accounts. With such an international audience, additional payments mechanisms would allow many of Facebook’s international users to more easily buy gift credits.
Some Facebook users, especially those younger than 18, may not have access to any payment mechanisms other than cash. Accepting cash in envelopes for Facebook points would not scale very well. However, many game companies have been successful in getting their branded prepaid cards distributed at retail. This is one way of turning user’s cash into a payment mechanism that can be used online. Facebook has the brand awareness to do the same thing by striking deals directly with the two biggest distributors of prepaid cards, Incomm and Blackhawk. Alternatively, if they did not want to deal with retailers directly, a company like GMG Entertainment could handle it for them.
However, some users don’t have any money at all to spend on gift credits. $uperRewards and MyOfferPal have found one way to reach this market, through incentive offers. These companies allow users to trade their attention (through filling out market research surveys, applying for credit cards, getting a free trial of a service, signing up for email newsletters or other activities) for virtual currency. They take the bounty paid by the company acquiring the user, and use some of that to buy the user their virtual currency. Facebook could enable users to buy gift credits with incentive offers.
The combination of these tactics to increase Means to buy digital goods could provide an additional lift of 50-100% in digital gifts revenue.
What other ideas do readers have?
Tomorrow we’ll discuss the last, and arguably most important factor, Motive.
What could Facebook do to increase its digital goods revenue – increase Opportunity. 1/3 February 3, 2009Posted by jeremyliew in digital goods, facebook, gifts, virtual goods.
Last September I estimated that Facebook is doing around $35m in digital goods sales. In November anonymous insiders suggested that it was closer to $50-60m in digital goods sales. That’s a healthy run rate, but I think Facebook could make a few product changes to do even better.
Facebook has made some smart changes to its digital goods/gifts product recently, highlighting upcoming friends birthdays on the home page, prompting gifts on birthdays and allowing users to buy birthday gifts in advance. As birthday gifting is the most common usecase for Facebook digital goods, these are changes that increase a users opportunity to buy virtual goods.
Facebook could increase gifting opportunities by prompting gift giving on other calendar events as well. Some are natural gift giving holidays, such as Christmas, Chinese New Year, Valentines day etc. It already provides a selection of holiday specific gifts for each of these occasions. Facebook could generalize the “birthday” section on the homepage into a “calendar” and note upcoming events that are gift worthy, and perhaps even suggest recipients. For example, if you are “in a relationship” with someone on Facebook, then they would be a natural person to prompt for a Valentines day gift.
Facebook could also prompt gift giving on certain notifications in the feed that might be “gift worthy”. Especially notable are changes in relationship status (e.g. moving from “in a relationship” to “married” or to “single” for example), but others might include changing address (housewarming gift?), changing educational or job status (graduation gift, or new job gift etc).
Finally, Facebook could prompt for gifting more generally through more, and more direct, calls to action. “Give a gift” is the fourth option for writing on someones wall, and moving it to first would likely increase gifting immediately. So would prompting for gift giving (not just comments) on Status Updates on a profile and in the news feed. Simply putting a link to “gifts” as the default first application in the application box in the right rail would help.
I suspect that increasing opportunity could increase gift giving by 50-100%.
Tomorrow, we’ll cover means.
Why do people buy virtual goods? January 13, 2009Posted by jeremyliew in digital goods, game design, game mechanics, virtual goods, virtual worlds.
1. Attention in a noisy environment (usually digital gifts)
2. Self Expression
3. Increased Functionality
and later proposed a fourth use case, convenience.
Vili proposes a different taxonomy:
Purely “utilitarian” or use-value0based attributes can be divided into two categories: performance (simple numerical advantage) and functionality (new abilities and options). Virtual goods also have attributes capable of generating emotional or hedonic responses, particularly their visual appearance and sound, but also any background fiction or narrative associateion with them. Hedonic attributes are difficult to distinguish emperically from the conceptually different social attributes, which refer to attributes that make virtual items suitable for creating and communicating social distinctions and bonds. Such attributes are provenance, customisability, cultural references and the “branding” of an item with a known commercial brand. Rarity is perhaps the most socially oriented attribute of virtual goods, because its value is strongly associated with its ability to distinguish a (small) group of owner from non-owners
In the paper he gives examples of each of his classes of virtual goods. He also summarizes some previous research on digital goods. In particular, he notes advice from Oh and Ryu to game designers based on research on Kart Rider and Special Force:
– Balance between items that can be purchased with real money and items that must be earned through gameplay, and build synergies between the two categories
– Allow players to keep “ornamental” items permanently, but make “functional” items consumable
– In the case of items that ive the player a performance advantage, do not disclose the exact numbers;provide approximate descriptive texts instead
– Introduce items linked to specific events and communities (e.g. Christmas decorations and guild emblems).
It’s useful to read the whole paper (around 15 pages)
Facebook’s digital goods revenue $50-60m sources say November 12, 2008Posted by jeremyliew in digital goods, facebook, gifts, virtual goods.
In September, I estimated that Facebook’s digital goods sales were on a $35m revenue run rate. Silicon Alley Insider quotes an anonymous insider to say:
Facebook’s revenue this year will be about $265 million, the source says, which is less than the $300 million expected. The source estimates that this is composed of about $180 million of ad revenue, $50-$60 million of virtual gifts, and some smaller revenue items.
I was in the ballpark!
IMVU founder’s framework for digital goods: three key questions October 20, 2008Posted by jeremyliew in digital goods, game design, game mechanics, games, games 2.0, mmorpg, virtual goods, virtual worlds.
Eric Reis, one of the co-founders of IMVU, posted last week on the three key decisions you have to make when thinking about virtual goods business models:
UGC or First Party content?
First Party – more control, but higher costs and harder to anticipate what users will want
UGC – Massive breadth of content, but have to put systems in place to deal with adult content and copyrighted content
Subscription or a la Carte payments?
Subscriptions – Greater game balance between rich and less rich players, lower fraud rates
A La Carte – Easier to monetize players without credit cards (e.g. teens)
Merchandising or Gameplay?
Gameplay – Virtual goods are functional, part of the core game mechanics, and confer benefit in the game. Demand is driven by game mechanics alone, and requires a delicate balance to ensure that players with money do not always beat players with time, skill and passion.
Merchandising – Virtual goods are not just functional, but also associated with self expression or attention in a noisy environment (see my previous post on the three use cases for virtual goods). This creates potential for greater demand for virtual goods, but requires the creation of a marketing and merchandising capability in the company.
Reis believes this framework can be used to describe any virtual goods business:
You can use these three questions to analyze existing businesses. For example, IMVU is a user-generated, a la carte, merchandising product. Habbo is first-party, a la carte, merchandising. Mob Wars is first-party, a la carte, gameplay. WoW is first-party, subscription, gameplay.
1 comment so far
It has often been said that good games are easy to learn and hard to master. At this years Penny Arcade Expo, Jamie Cheng (CEO of Klei Entertainment, who are the developers for Sugar Rush, Nexon’s first “built for built for North America” free to play title) notes that these principles are even more important for free to play games:
With a free-to-play game, a positive initial experience is crucial. Cheng recently played Jonathan Blow’s Braid, and very early in the game came upon a puzzle that was so frustrating he put the game down. His main motivation for returning and giving it another shot, after which he very much enjoyed it, was that he had already paid for the title.
Free-to-play games don’t have the luxury of player obligation, he says, because they haven’t made a concrete monetary investment. Games that don’t immediately grab players simply won’t be played.
For a similar reason, games need to have depth and complexity beyond their approachable entry point — after all, free-to-play games are only monetized if players keep playing them, as opposed to retail games, which are monetized as soon as the player buys them. “If nobody spends their time, we’re not going to make any money,” Cheng says.
By extension, a social experience is crucial in creating the word-of-mouth marketing that will allow the game to reach a long-term dedicated audience.
“For example, I’ve bought probably 40 additional tracks for Rock Band, as well as the N+ cooperative levels. I bought those not just so I can play by myself, but so I can play with my friends,” he says.
Longevity is also key: “We’re not trying to build a grindfest, we’re trying to build a game that grows and evolves with you,” Cheng explains.
Some social games such as Friends for Sale (a Lightspeed portfolio company) and (Lil) Green Patch take this to such an extent that many new players don’t even realize that they are playing a game, but think of the app as a lightweight communication tool. But over time thanks to smart application of game mechanics, a hardcore player base develops. Traditional and casual game designers should take note.
Facebook selling digital gifts at a $35m run rate September 2, 2008Posted by jeremyliew in business models, digital goods, facebook, gifts, virtual goods.
In January of this year, we estimated that Facebook was selling digital gifts worth $15m per year. We based this estimate on an analysis of the number of each gift available each week over a 7 week period.
Facebook creates a certain fixed number of each type of gift. When the number remaining for any particular gift drops below 100,000, Facebook displays the number left. (The most common size runs are 100,000 and 1,000,000 but they range as high as 10,000,000 and as low as 15,000.). For those items where less than 100,000 remain, we can track how many gifts had been sold in the preceding week by subtracting the number remaining from the number remaining the previous week.
We updated the analysis this month and found that Facebook has dramatically increased its sales rate of digital gifts. As before, we tracked the number of digital gifts available of each type (where data was available). We ordered the items by bestselling (as defined by Facebook) and, because data is sparse, we divided the list into groups of 20, took the average of each group of 20 items and applied that sales rate to all items in each group.
Once we excluded the free gifts, the averages looked like this:
By multiplying each average by 20 and adding the totals we came up with virtual gift sales of between 390,000 and 600,000 per week, with an average of around 470,000 across the three weeks.
The vast majority of facebook gifts are bought from the first screen of gifts in the directory – almost 80% of the total sales come from the group of the first 20 gifts. This points to the self reinforcing nature of popularity (the crowdiness of crowds rather than the wisdom of crowds) when popularity data is made public.
We need to take into account seasonality. In most retail environments, something like 40% of sales occur in the last 8 weeks of the year. Judging from the high number of holiday themed gifts over the holiday period last year, the same seems to be true of Facebook:
Holiday themed gifts (e.g. Santa hat, eggnog, Happy New Year!) dominated the list of top selling paid gifts, averaging 4,755 sales per week.
If we apply this assumption to our weekly average sales numbers, we multiply by 73.3 (instead of 52) to get to an estimate of annual sales. Using this estimate we get a range of between 28,500,000 and 43,500,000 in annual Facebook gift sales, with an average around 34,500,000.
Unfortunately however, as only one item in the first 20 had “number left” data available each week (Bear Hug), this also makes our estimate prone to significant error. Bear Hug was consistently around 6 or 7 among the most popular gifts, behind most of the birthday gifts, so hopefully it approximates the average popularity of the top 20. Applying 25% uncertainty to the average of the top 20 bestselling gifts creates a similar range of between 28,000,000 and 42,000,000.
In both cases, these estimates are double the number that we estimated at the beginning of the year which was around 15,000,000 digital gifts. That estimate was based on data drawn during the holidays and may have been high on a run rate basis.
Given that Zuckerberg has estimated that Facebook will do between $300-$350 million in revenue this year, digital goods constitutes a meaningful secondary revenue stream for the company.