Analysis After many years of anticipation, it is fair to conclude that the mobile games market finally arrived in 2003. Japan and Korea aside, the main territories of the world had merely flirted with wireless gaming since 1999, when the first WAP and SMS projects went live. Throughout this period, developers and aggregators bemoaned the general lack of interest from network operators in the sector.

Whilst Japanese networks created tightly controlled data services infrastructures and encouraged content partners with generous revenues shares, the prevailing attitude among their European and North American counterparts was "you build it, we’ll host it and we’ll take 80 per cent too".

Two factors re-drew the landscape:

• Mobile phone penetration in developed countries started to reach saturation, meaning that networks had to look towards non-voice services to increase revenue and act as a differentiating point to reduce churn.
• Downloadable technologies (and enabled handsets) emerged to offer a compellingly simple (and billable) platform for the delivery of games.

In 2003 the first downloadable services were launched en masse. Vodafone’s Vodafone Live and T-Mobile’s T-Zones introduced pay-per-download games in just three or four ‘clicks’ to subscribers. The services were heavily advertised. Throughout Europe most operators launched similar initiatives. Meanwhile in the US Verizon and Nextel led the way for a similar transformation.

The approach taken by operators today is drastically different from that taken during the era of WAP. Now, most networks have a dedicated content unit that liaises with aggregators and developers. The average revenue share is 50 per cent. It is still high compared with i-mode service in Japan (where the share is 91 per cent), but in Europe and the US airtime is cheaper so it is harder than in Japan for networks to profit from the airtime revenue accrued during the download.

The ‘walled garden’ built by operators like Vodafone also contrasts with the open environment of Japanese operator DoCoMo’s i-mode (which has tens of thousands of unofficial sites). Some European networks contend that the system of offering a set number of games on a one-off payment model (rather than subscription) suits the Western mindset best. In order to nurture such a system, the operators need to control it tightly. This control extends to dictating APIs (Application Programming Interfaces) to handset vendors in order to make the consumer experience fast and intuitive. This was vindicated by the overwhelming popularity of Vodafone Live’s Sharp GX10 phone, the handset most strongly configured for the service.

Of course, not all networks are as hands-on as Vodafone. Some appointed third parties to build, manage and provide content for their games platforms. This was the strategy of Norway’s Telenor, which uses Macrospace to deliver Java games. Meanwhile the French enabling firm in-Fusio still works closely with the remaining European networks that support its ExEn games engine. In the US and parts of the Far East, Qualcomm does the same with its downloadable engine BREW.

By the end of 2003 operators were able be cautiously optimistic about the future of mobile gaming. The services were working well technically and delivering hard cash at last. The predominant pay-per-play payment model has also shown itself to be robust. But there is plenty to be done.

A 'billion-dollar' industry

Although downloadable applications have proved popular with consumers, the numbers are still tiny relative to the traditional console and handheld gaming business. The few publicly available numbers reveal that the market has so far barely touched the mobile owning mainstream, except of course in Korea and Japan. According to Screen Digest estimates, in 2003 these two countries accounted for 80 per cent of worldwide games download revenues of €380m.

But the vast majority of major operators in North America and Europe have now established games download services on their networks. The penetration of Java handsets is also rising fast in these markets, such that we expect games-enabled handsets to have risen three-fold in Western Europe to 110m by the end of 2004, and more than double in North America to 70m. As a result, we are expecting download spending outside the Far East to really take off this year.

Whilst games download spending for Japan and Korea combined is expected to grow by around 50 per cent in 2004 (to €446m), Western Europe and North America combined is forecast to grow at 10 times that rate. A 500 per cent jump will result in games download spending in these two regions of €332m in 2004. Although seriously trimmed, Japan and Korea’s majority share of worldwide mobile games market will remain in place for another year. We reckon that Japan and Korea will have accounted for 51 per cent of global spending in 2004.

In total, we anticipate the wireless games download market will prove to be worth around €880m ($1,070m) in 2004, breaking the billion-dollar barrier for the first time.

Despite explosive growth in Western Europe and North America this year in particular, we do not expect combined spending in these two markets to exceed spending in Japan/Korea until 2006 at the earliest. One of the contributing factors is the level of advancement of the mobile universe. In terms of games-enabled handset penetration, Western Europe and North America are around 2-3 years behind these Far East markets. Yet a much more significant factor is the games ARPUs deriving from games-enabled handsets in the market.

It is clear that wireless gaming ARPUs (per game-enabled subscriber) are significantly higher in the Far East than those being achieved in the US and Europe. We estimate that in 2003 games download ARPUs were four times higher in Japan/Korea than in the newer markets, although the gap was expected to have narrowed to 2-3 times higher by the end of 2004. But we think that if the wireless gaming market develops in line with the business models that are currently being deployed in Europe and the US, then there is little chance that mobile gaming ARPUs will match those of Japan and Korea right through to the end of the forecast period in 2010.

The business models common in Japan and Korea are different from those gaining traction in the US and Europe, although in different ways.

Flat-rate or pay-per-download?

The Japanese (i-mode) model is a flat-rate subscription to an unlimited supply of content, which has been highly successful at encouraging both wireless gaming usage and stimulating a thriving and innovative wireless games development industry in the country.

Replicating this model in Europe would undoubtedly generate similar benefits for the European industry, and consequently for wireless games spending in the region. Some operators question whether this model would be financially viable, given the low price of airtime in Europe compared with Japan (where NTT DoCoMo makes its money out of the airtime component of this model).

The i-mode model is being deployed in several European markets, and although it is early days, it’s fair to say that its success rate has so far been ‘limited’. However, there are major mobile operators in Europe who are convinced of the need to move to a flat-rate pricing structure for data services, and we believe are prepared to go it alone if necessary.

In our view, the introduction of flat-price pricing would result in a major fillip to the European mobile gaming market. If i-mode does take off in Europe, or some other form of flat-rate price model becomes widespread, then our forecasts for wireless gaming spending in Europe and worldwide will be revised upwards significantly.

In Korea, the predominant business model for wireless gaming is the pay-per-download model found in Europe and the US. But games download ARPUs are comparable with Japan, and (paid for) download rates are considerably higher than that achieved in the US and Europe.

The major difference between Korea and Western Europe/US is in pricing, with mobile games downloads in Korea averaging around 1,500 won ($1.30) in 2004. This is between a third and a quarter of the average price charged by operators in Western Europe and the US. Yet pricing differentials are dwarfed by the difference in download rates, with about 10 times the number of games purchases per enabled handset in Korea than outside the Far East. As a result, Korea has much higher mobile gaming ARPUs (for enabled handsets).

Certainly not all of this massive difference in download rates can be explained by the cheaper Korean download price. But it is a major contributing factor, with the whole spectrum of potential users highly incentivised to download new games, or even just try out mobile gaming for the first time.

The Korean model shows that there is an alternative to the flat-rate model to generate higher mobile gaming ARPUs, and that is more competitive pay-unit pricing. This is certainly a serious consideration for European and American operators wishing to increase their overall download revenues. Both of these strategies have a corollary in the fixed-line data market. Several European broadband ISPs are now finding that subscription-based models for fixed packages of content are a far more attractive consumer proposition than pay-per-download. Meanwhile, after years of prevarication, the market for (legal) music downloading has taken off with the introduction of the 99 cent download (and cheaper) by companies like Apple, Real and Wal-Mart.

Operators' key gaming issues

Because usage and awareness are still relatively low, networks must now address key issues if they are to maintain their position as the drivers of the market.

These include:

Marketing

Often there is still a vacuum where promotion should be. Thus far, virtually all marketing has been generic rather than specific. Of course, revenues may not justify the advertising of individual titles. But more must be done to show subscribers how easy and enjoyable mobile gaming can be.

Meanwhile, specialist portals such as Germany’s Jamba and Zed do advertise heavily. Operators must surely be aware of how the ringtone business was poached - they will be wary of this happening again in gaming.

Access

Operators should consider broadening the scope of the games decks. They are right to maintain quality control. But if the sector does surge, again customers will defect to portals and magazine ads for their games as they did in the ringtone market.

New technologies

Making available demos, pay-per-level, access to leagues and tournaments, multi-play and so on - if they are implemented well - should improve the user experience and grow the market. Also significant could be the downloading of mini games engines that allow for further play. For example, a user could purchase a chess engine and then play a nominal amount to play games with other owners of the same engine. Many of these technologies will be made easier to implement by the arrival of MIDP 2 for Java and by standards agreed by trade bodies (such as the Open Mobile Alliance’s work on digital rights management).

Alternatives to the operators

There are plenty of alternatives available to the keen player of mobile games, should operators fail to entertain subscribers or keep their partners happy. In Europe, the web now provides an open-ended channel for masses of Java games vendors to offer games using premium SMS as a payment mechanism. Virtually all ISPs (Internet Service Providers) and some major consumer brands offer downloadable games from their sites. But there are hundreds more specialist sites dedicated to selling Java games.

These websites are generally more localised than operators, working with local studios to make available games that cater for regional tastes. They can handle any number of titles and generally shuffle them far more reactively than the networks. Some offer a high revenue share for content suppliers - up to 80 per cent. However, the norm is rather less attractive with portals subtracting up to 40 per cent in billing fees, and another 20 per cent for marketing/advertising.

The remainder is split in half leaving the content partner with 20 per cent of the download fee. Players can also acquire games from interactive digital TV services and from adverts in the press. In the case of the latter, there is often diffidence from developers to be involved thanks to the sometimes transient nature of the organisations placing the ad. There have been reports of illegally posted games and non-payment of content providers in this sector.

At present, web and TV portals could control as much as 50 per cent of the business in some European territories, but overall we believe that operators command 80-90 per cent of the European market. In the US, their share is negligible because the non-adoption of SMS there makes billing a problem for any would-be games supplier apart from the operators themselves.

In the future, operators may face competition from retailers (although it won’t be competition if the retailer is the part of the operator’s own high street network). Some stores have already begun selling games. During the WAP era these came in the form of scratch cards that revealed ID numbers to unlock games WAP sites. In the download era, they use the same basic model although others have been tried such as the CD-ROM that allows the transfer of games on to a phone via a cable or bluetooth link.

Self-evidently, this model is only likely to attract hardcore mobile gamers rather than the more casual mainstream market. The sector may improve when retailers formalise a technique to ‘flash’ games into handsets on the premises.

Mobile games platforms

There are many ways to get a game on to a phone. The qualities of these platforms determine the gameplaying experience - which is why some work and some do not. Text games work because they are easy to understand. But their technical limitations reduce their potential. Browser-based (WAP) games don’t work because they are slow and static. Downloadable games have taken off because they play like conventional handheld games.

SMS

SMS games have been eclipsed by downloadable games - for obvious reasons. Downloads look like most people’s idea of an electronic game - colour, graphical, interactive. SMS games are really text-based quizzes. But they still fulfil a role.

As the name implies, SMS is ideal for sending individual short messages, but pretty hopeless when it comes to creating compelling multi-turn experiences. The user interface is clumsy. However, when WAP was failing during 2001/02, text games were the industry’s only viable platform.

SMS games are generally a ‘dialogue’ between the subscriber and his/her own network. For this reason, developers and publishers have sought close partnerships with individual operators. In general, operators share between 20 per cent and 40 per cent of premium SMS revenue with games providers. In Europe operators bill with a so-called ‘0 and 10’ model, where all ‘Mobile Terminated’ or incoming traffic is free to the user, while ‘Mobile Originated’ traffic costs $0.10/£0.10 to send. Elsewhere a ‘2 and 10’ model remains. Here, the user pays around $.02/£0.02 to receive an SMS message and $0.10/£0.10 to send one.

Generally, SMS games pay the content developer based on Mobile Originated revenue share deals. Networks pay between 10 per cent and 30 per cent. Premium priced SMS

Premium pricing simply means that the cost of an SMS transaction is variable based upon the message’s contents and/or where it is sent. In Europe, premium SMS is commonly used as a download billing system and it’s become a fairly lucrative way for game developers to earn revenue.

Premium SMS is also used as a payment method when charging directly for downloads is not possible. For example, off-the-page retailers and web portals will use it to charge for Java games that are free to download (excluding airtime costs).

Developing for SMS

For obvious reasons SMS games have evolved along certain common lines. Player messages have been made quick and simple (more complex responses reduce the number of messages they will send) - one word or even a single letter. Good SMS games are as much about narrative as they are about technology. The best games of this type can generate around 200,000 texts a day.

WAP

During the heady, early days of WAP some aggregators enjoyed isolated success. Perhaps the biggest of all was that of US developer Jamdat, which created a variation on ‘scissors paper stone’ called Gladiator for subscribers to the US carrier Sprint PCS in October 2000. By December 2001 over 2.5m games had been played by nearly 1.1m users and generated over 14m minutes of usage. A sequel to this game was then developed for wireless networks operated by Alltel Wireless - including AT&T Wireless, Qwest Wireless, Sprint PCS, Telus Mobility and Verizon Wireless.

But it was to be a brief burst of successful activity. WAP was flawed as a protocol for gaming. Essentially, WAP-enabled phones are small browsers that connect to the server just like an Internet browser connects to web servers on the Internet. The processing power is held completely on the server side so users can really only make choices and be progressed to a choice of further WAP pages. The games are implicitly turn-based and often very slow. So WAP never caught on as a games platform.

There is no long-term future in the ‘format’ as a stand-alone gaming platform, although many operators still offer WAP games for those who know how to find them on their browsers.

Mobile games platforms: Java, BREW, ExEn, Mophun

For years, downloadable gaming was the nirvana of the mobile games industry. Downloadable games permit animation, can be played offline and can be monetised so that consumers play the game along the lines of a traditional retail model - ie, they pay once and play offline all they like.

Of course, downloadable applications are only possible when a technology is developed that enables a program such as a game to run, to be transmitted over the air and to be understood by the user device. And these are just the basics. Other requirements include the ability to handle graphics, sound, different screen sizes and user interfaces, security, digital rights management and so on.

The most widely adopted downloadable technology is Sun Microsystems’ Java variant for small devices called J2ME. However, the perceived shortcomings of J2ME - as well as the precise demands made on mobile devices by games (as opposed to other downloadable applications)- has led to rival technologies being developed. They include BREW, ExEn and Mophun.

These platforms had achieved varying levels of market penetration. Java has been adopted all over the world, BREW in the US, China, Taiwan and Korea; ExEn in France and Germany; Mophun in parts of the Nordic region.

Copyright © 2004, Screen Digest

Screen Digest the Newsletter is the international media business's leading news & market research journal. It has been published for more than 30 years and is read in over 40 countries. Subscription details here. It is published by Screen Digest, a research company which produces a rapidly growing number of major business reports on media markets.

Related stories

Archos ships video, audio, gaming handheld
Skulls Trojan keelhauls Symbian phones
Spurned woman deletes ex's gaming data