July 2007

Play No Evil offers up a good article breaking down why Free To Play is a superior engagement model to subscription or retail. Matt Mihaly from Iron Realms jumps in to offer a counterpoint on MapleStory’s user base (i.e. 57M is a gross number, not representative of active users) and offer up Habbo Hotel as another North American virtual item sales success story.

In regard to Maple Story’s user base, all 57M of them may not be active, but most of them are probably unique due to the fact that in China and South Korea, Maple Story requires users to enter their Social Security Number to sign up. So there likely aren’t very many duplicates, world-wide.

Regarding Habbo’s numbers, Paul Thind (Habbo’s GM) reported their active uniques at 7.5M world-wide (1.7M uniques/month) at last month’s Virtual Goods Summit. Habbo’s annual revenues sit north of $65M.

From the article on Play No Evil, I especially like this passage:

The choice of prices are important. Buying blocks of currency or subscriptions for less than $10 means that it is an impulse buy. $50, or even $20, to buy a game is above the “whine factor” for parents. Nexon’s ability to get payment cards into Target (see previous article) was critical – online game play becomes an impulse purchase at the checkout line. I think this is an area where many casual game companies “don’t get it”. Their prices are above the impulse purchase level.

In a related story, Fox News – that bastion of sound journalism – just did a piece on MapleStory and game addiction.

add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank


Last week I attended the Casual Games Conference in Seattle. One of the better talks was by Whon Namkoong, CEO of NHN USA. He discussed why he thinks the North American market is primed for virtual item sales games and the lessons his company has learned while trying to enter this market. I’ve summarized his points below.

In 2001, the US and Korean game markets were quite different:


  • Rampant in Korea
  • Not nearly as bad in the US

Broadband Penetration

  • 15% in Korea
  • 2% in US

Broadband Speed

  • 6.8mps in Korea
  • .2 mbps in US

Number of Online Payment Methods

  • 5+ in Korea
  • Only PayPal in US

Today, things have changed in the US:

  • 47% broadband penetration
  • 4.8mbps broadband speed
  • Lots of online payment methods

He listed the well-known advantages to a virtual item sales model:

  • Low barrier to entry for user
  • Unlimited ARPU
  • Dynamic item updates, support for season items

But more interestingly, Whon covered the lessons NHN learned about virtual item sales in the US market:

Lack of efficient payment methods

  • 90% of virtual item sales in Korea are paid for via SMS
  • Korean SMS transaction fees are less than 10% (vs 40-50% in US)
  • US requires bank account linked to mobile phone; Korea does not

US network cost is 5x more than Korea

  • In the US, NHN changed their Gunz product from P2P to client/server.
  • Resulting network expense made the product inviable – the more users, the more money Gunz lost.

Serious hacking attempts

  • Asian hackers do it for the money – i.e. dupe and sell
  • US hackers do it to “break stuff”
  • Korean hackers don’t reveal their methods, so they can continue to profit from them
  • But US hackers do it to share their knowledge, which causes a much larger problem
  • In 5 years of Korean hacking, NHN never experienced the US hacking techniques
  • Additionally, when ijji.com was launched in English, it essentially invited hackers from all over the world (rather than just those who could read Korean) – so hacking increased dramatically

Finally, Whon urged North American developers to work together to:

  • Support the use of prepaid cards and other payment methods
  • Make an alternative delivery system like P2P viable
  • Share info on hacking and hackers

add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank

Worlds In Motion – Blogs to Watch

Thanks to Leigh Alexander at Worlds in Motion for naming FreeToPlay.biz one of three “Blogs to Watch,” alongside two of my favourite blogs: Raph Koster‘s and Nabeel Hyatt‘s. I had the pleasure of meeting both of them at the recent Virtual Goods Summit.

add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank

THQ and Shanda Announce Company of Heroes Online(TM)

So finally I can blog about this, as it’s something Relic‘s been brewing for a while now. Shanda, the company that coined the term “Come-Stay-Pay” to describe their free to play model, will be delivering Company of Heroes Online.

Obligatory press release quote:

THQ Inc. (NASDAQ:THQI) and Shanda Interactive Entertainment Limited (NASDAQ:SNDA) today announced a co-development and publishing agreement for an online-only version of THQ’s critically acclaimed 3D Real-Time Strategy (RTS) PC game Company of Heroes. Company of Heroes Online (working title) is currently being co-developed by THQ studio Relic Entertainment, the creators of Company of Heroes, and Shanda. As part of the agreement, Shanda has obtained the exclusive license to operate Company of Heroes Online in Mainland China, and THQ and Shanda will also share the benefits of the games potential release outside that region. Company of Heroes Online is scheduled for release across Mainland China in 2008.

The original Company of Heroes was released to broad critical acclaim in 2006. The game is set in World War II and brings to life the journey of the men of Able Company as they fight across Europe. The RTS hit has won more than 37 awards to date, including six PC Game of the Year and 12 Strategy Game of the Year awards. In addition, the game is listed in more than 10 Top 10 Games of 2006 lists. Company of Heroes Online is set to deliver a persistent Company of Heroes experience for online gamers where players will be able to build their character up from private to general through new multiplayer cooperative missions, gameplay modes, and player versus player combat. The teams at Relic Entertainment and Shanda are joining together to create a persistent RTS experience across the expansive battlefields of World War II Europe.

CoH is a fantastic game (93% on Metacritic) and THQ & Relic’s partnership with Shanda is a great way to expose it to a lot more people. In mainland China, Shanda receives the exclusive license to operate the game. But in the rest of the world, Relic & THQ are free to operate CoH Online under a revenue sharing agreement with Shanda.

Shanda also recently announced a very similar deal with Tecmo for Dead or Alive Online, a game they aim to release prior to the 2008 Beijing Olympic games. I’m interested to see how they overcome the controller peripheral issue without limiting their market.

In any case, from a Relic perspective the Shanda deal is a great way to get some cutting edge persistent world development experience in the short term.

add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank

I don’t know if Maid Marian‘s Sherwood Dungeon RPG is indeed the most successful Shockwave free to play MMO, but it’s done pretty well for Gene Endrody and his wife – the only two employees of the developer, Maid Marian.

More interesting though is that the lone source of revenue for all of Maid Marian’s games is Google Adwords PPC (Pay Per Click) ads.

Recently, I had the chance to sit down with Gene and discuss Maid Marian, Sherwood and his company’s other projects. He was kind enough to offer up some very interesting info for those of us who aspire to create a free to play game as a “lifestyle business,” as Gene calls it.

Here are the Top 10 things I learned over coffee with Gene at Caffe Artigiano.

1. Revenue
Sherwood Dungeon sees 1M unique users each month, generating 10M ad impressions via Google Adwords. Gene claims a 4-5% CTR (click through rate), where traditional banner ads see a paltry .3% CTR.

Out of respect for Gene’s privacy I won’t reveal everything, but based on what he told me I estimate Maid Marian is grossing approximately $700-$800K/year, with 50% of that number being reinvested into the business.

Pretty solid income for a husband-wife team with a homegrown, free to play product supported solely by ads.

2. Technology
Maid Marian utilizes Shockwave for all their games. And Gene says that while Shockwave has just 55% penetration among the general populace (and Shockwave installations are declining), his market – teens – sees about 80-85% Shockwave penetration.

So the higher Shockwave penetration among Gene’s target demographic makes it an excellent technology platform.

Note: As a point of comparison, Flash has 98% penetration and is coming on strong as a robust game development platform. Interestingly, Gamasutra and Next-Gen.Biz both ran articles this week on Flash game development.

3. Servers
Sherwood Dungeon needs just 6 servers: 2 web servers, 3 game servers and 1 test server. Crude load balancing is accomplished via a round-robin DNS splitting scheme that directs incoming users evenly among the servers.

Maid Marian’s servers are provided by Peer 1 – a hosting company with a Vancouver data center – at a cost of $200 each per month and provide 2000gb of monthly bandwidth. But Gene is quick to point out that equivalent servers can be had for as low as $80/month these days.

Because all the user data in Sherwood Dungeon is stored client-side, there are no database security concerns or associated db speed drag. Data encryption and some client-side checks attempt to curb cheating or too rapid character progression, but of course player cheating is possible with this scheme.

4. Demographics
Because Gene’s game is web-based, he uses Google Analytics for most of the coarse metrics on his user base (country of origin, platform, unique users, etc). Some interesting stats:

  • 15% of Sherwood’s traffic is from US (largest single territory)
  • 5% of Sherwood’s traffic is from Canada
  • Maidmarian.com is the 500th most visited site in Jamaica

Hungary and Poland also have disproportionately large representation in Sherwood Dungeon. The game outperforms in areas such as Eastern Europe that aren’t well served by traditional game distribution methods.

5. Usability
In the interest of making Sherwood Dungeon more accessible, Gene has chosen not to expose player levels to each other. This means Player A can’t see what level Player B is at. Gene believes that this encourages more experienced players to help out new players, supporting the casual nature of the game.

Additionally, player levels are not even considered in PvP (Player versus Player) mode. So even if an experienced player chose to pick on a new player, ganking is much more difficult.

Players can also change their character type or look at any time. Players are not locked into a character after making their initial selection.

6. Content
Most of Sherwood Dungeon is procedurally generated, which saves a ton on level data. Among the items procedurally generated:

  • Dungeons
  • Islands
  • Height maps
  • Trees
  • Ramps

Gene uses seeds to ensure the same result every time with his procedurally generated worlds.

7. Genres
Maid Marian features three types of games, theorizing that users who are temporarily tired of one type can jump to another for a while. These game types are:

8. Distribution
Rather than block third party sites from linking directly to his games like many casual game providers do (at least the ones that rely on revenue from ads on their sites), Gene chose to embrace them. As a result, web game aggregators such as Addicting Games have become great secondary sources of new players.

Maid Marian’s highest referrer is Arcade Town, accounting for 5% of Gene’s total traffic. But Gene ensures his Google Adwords are preserved even if the third party iframes his game. Using his logs, he’s able to identify sites that violate his affiliate policy by stripping out his ads.

However, the sites are free to bracket his game with their own ads, forming an ad-hoc revenue sharing program.

9. Client Size
I’m a big believer in tiny client sizes and Sherwood Dungeon is a clear winner here. Clocking in at just 2mb (compared to 4mb for the already tiny Runescape and Habbo), Sherwood’s puny downloadable client makes it the fastest loading free to play MMO I’ve ever experienced.

Couple that with deferred sign up (as I mentioned in my previous article), and you have a deadly fast way to get players into your game.

10. AI
In Sherwood Dungeon, the AI monsters utilize distribute processing to reduce server load. This means that the brain for the monster you and your friends are fighting resides on your PC, not the server.

When you log off mid-fight, the control of the monster’s brain is seamlessly handed off to another user’s PC. A monster might “lose his mind” momentarily if his AI can’t successfully land on another user’s PC, but it’s often tough to tell the difference between a monster that’s flailing with purpose and one that’s not.

A big thanks to Gene for his time. I had for more interesting points, but had to cut them down for this article.

Also thanks to Raph Koster, for directing a small subset of his traffic my way twice in the last couple weeks.

add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank

Whether you’re making a casual MMO like Maple Story or a virtual world like Habbo Hotel, here are 10 ways to remove game-killing barriers to entry and create the largest possible addressable market.

1. Free to Play
The Free to Play business model is here to stay – and growing every day. In an entertainment world filled with endless choices, asking someone to pony up $50 before they can play a game is rapidly becoming a non-starter. The focus now is on getting players through the front door, keeping them happy, then monetizing 5-15% of them. Non-paying customers become “content” for the paying minority, so don’t think you can ignore them.

2. Integrated graphics support

“If our games required a video card, we’d lose 80% of our audience”
– Min Kim, Director of Game Operations, Nexon North America

“Graphics are not important – the mind models the situation”
– Daniel James, CEO, Three Rings

Enthusiasts who purchase the latest, greatest video card make up just 4% of the market. Integrated graphics (i.e. no dedicated video card and therefore lower graphics performance) accounts for over 60% of all new computer sales. It would be foolish to develop a Free to Play product requires a video card when success in the F2P sector is partially reliant on addressing a large market and monetizing just a small fraction of your player base.

3. Multiple, regionalized payment systems
Finding the right payment method is a key success factor for Free to Play products. When a user finds a payment method they’re comfortable with, they are fiercely loyal to it. But there are nearly as many payment methods as their are markets. Erik Bethke of GoPets says his company utilizes 90 different payment systems worldwide in order to address the local preferences of each region and make it as easy as possible for users to pay.

Many factors influence payment method selection. Credit card penetration in China is low, so billing customers via their land-line telephone provider has become a widely used payment system that provides excellent security in exchange for high surcharges. In Europe, SMS payments are hugely successful and carriers take anywhere from 10-30% surcharges versus the 40-50% fees of North American carriers. PayPal, checks, points cards and more are also used.

We have three people on staff whose full-time job is to open envelopes with single dollar bills and quarters in them. The users can’t figure out how to get the cash to us. One user sent in a $5 bill in a $14.95 FedEx package so it would get to us on time.
– Craig Sherman, CEO, Gaia Online

4. Little or no download
Get users into a game as fast as possible. If your game requires the user to download client software, make it as small as possible and give the user something to do while they wait for the game to download and install (i.e. setting up their character).

But better yet, make your game in Java, Flash, Shockwave or Silverlight so it’s playable within a browser. A game delivered via Java applet (i.e. Puzzle Pirates, Bang! Howdy, Runescape) can be downloaded and installed in under a minute. A signed Java applet will even avoid tripping a user’s installed spyware detectors.

Only ~30% of players actually tolerate downloads at all, the other 70% preferring to play online. I believe this percentage of download-intolerant players is increasing.
– Daniel James, CEO, Three Rings

5. Deferred sign up
How many times have you been faced with filling out a mandatory sign up form before you can starting playing a new game? The barrier of filling out one more form and becoming a member of yet another online site/network/game/etc that might eventually spam you – before you even try the product – is a huge barrier to entry.

Why not let a new player name and create their character, enter and start experiencing the product, then ask for sign up information along the way? A game that gets this right is Maid Marian‘s Shockwave MMO Sherwood Dungeon, which allows you to start playing immediately after you enter your desired character’s name. Despite its simplistic graphics and lack of server-side character saves, Sherwood has attracted over 1M users to its Free to Play ad-supported game.

6. Easy to understand world/lore
Pets, penguins, pirates, party goers – these are some of the most successful Free to Play virtual worlds and games. If you want to keep your game’s potential market big, utilize commonly understood worlds, characters and rules as often as possible. There are exceptions of course, but generally the more jargon and fiction you graft onto your property, the greater the barrier to entry for new players.

7. Quick to play core
Build your game or virtual world around a quick-to-play core mechanic that loops into a larger meta-game. A game that can be played in small 5 minute chunks that feed into a higher purpose.

The casual MMO Puzzle Pirates was designed with short play sessions and a solid meta-game in mind. However, the average Puzzle Pirates user spends 2.5 hours per day in the game – 30 days a month. And while some players do drop in and leave, others spend up to 9 hours a day in-game. Ultimately, the game’s short compulsion loops keep players online longer than traditional, longer compulsion loops that take 30-60 minutes to complete.

8. Warp, don’t walk
Spending precious minutes walking to destinations is, for many, a significant barrier to entry and a big waste of time. Many games and virtual worlds allow “warping” between areas to avoid long marches or simply a point-and-click interface with the world.

9. Spending limits
It seems counterintuitive, but enforcing spending caps on some or all of your player base (depending on your product’s demographics) may actually increase your user base. Habbo Hotel puts spending caps on all payment methods to control the influx of cash into their economy but also to allay parents’ fears. Users can spend money only on 2-3 predefined days of the week.

Limiting how much a player can spend spend in a short period of time benefits the game by reducing parental concern and decreasing incidents of buyer’s remorse in new players.

10. Secondary markets
The presence of a secondary market can drive the primary market. Wizards of the Coast had this observation, as told by Daniel James at this year’s Virtual Goods Summit:

Wizards of the Coast had some interesting things to say, that secondary markets, for example of Magic Online, have been incredibly valuable in driving the primary market. People will buy way more cards in the primary market because they know they can flip them. Mostly they don’t, though, they just hold onto them. Which is a great tip for people thinking about this.

So embrace secondary markets as more users will choose to participate in your primary market if they believe they can sell their goods to others when they’re done the game.



add to del.icio.us :: Add to Blinkslist :: add to furl :: Digg it :: add to ma.gnolia :: Stumble It! :: add to simpy :: seed the vine :: :: :: TailRank