Development of services is being held back by the self interest of the main vendors, says Monte Silver, who perceives a ‘Great Industry Conspiracy’.
Since GPRS, the hype surrounding non-SMS data applications has been constant. In Cannes/Barcelona, one can easily identify the annual killer-application. Mobile internet (WAP), MMS, streaming and mobile TV have all enjoyed the spotlight, but none have delivered the mass-market. While there can be no single explanation for this failure, this article argues that these services have failed because they have primarily been generated to serve the interests of industry – network and handset vendors, and to a lesser degree the operators.
Vendors promote services which favour perpetual network and handset upgrades. Operators promote portals to strengthen their brand. Unfortunately, the user is low on the totem pole. Although we all hold devices and are constantly pressing their buttons, we have not adopted data services. This article argues that if services continue to favor vendors and operators, rather than focus on client-oriented user and usability issues, DVBH and HSDPA networks will be launched, but not used either.
Seeking to exploit the internet revolution, the leading handset and network vendors saw a huge opportunity in GPRS if they could persuade cellular operators to upgrade networks and market internet-enabled handsets. There was only one way to off-set declining voice ARPU, it was argued, and that was by tapping into the ‘pent-up’ demand for mobile data, anytime, anywhere. Mobile internet was positioned as a natural extension of the PC experience and flattering analyst reports all agreed. Mobile internet hysteria was born.
Late 2000 marked the deployment of the first major WAP services in Europe. Ericsson and Nokia’s combined network sales that year was approximately $23.7 billion. In addition, Nokia sold 128,000,000 handsets, generating €21.8 billion. Operators searched for the killer GPRS application. Users found WAP, however, to be CRAP. Notwithstanding, with global networks emerging, operators needed to build strong global brand. The WAP portal was born. While teens continued to download ringtones, logos and games, and people SMSed themselves, the portals remained the domain of early-adopters only.
Despite the failure of WAP, vendors were now focused on the next really big thing – MMS which, of course, required new handsets. MMS was positioned as the natural evolution of SMS. But the user experience of the native client was horrible. Opening the MMS client required several clicks. Composing a combined visual and audio was close to impossible. Add operator interoperability issues and the need for a compatible handset on the receiving side. Operators could not commercialize MMS, so simply stuck poor quality photos on the portal to download. As an end result, MMS and picture messaging has disappointed everyone but the handset vendors. In 2000, about 400,000,000 handsets were sold. In 2004, 257 million camera phones were shipped worldwide, representing approximately 39 percent of total handset sales. In 2004, Nokia sold 207.7 million units, all with color screens.
The 3G opportunity
PC-based internet was evolving towards broadband and rich-media. Mobile broadband (3G) presented the vendors with an opportunity to completely replace existing networks and sell more phones. GPRS, they argued, could not support rich-media, and mass-adoption would congest the networks. There was only one solution: UMTS.
As only a few UMTS licenses were tendered per country, operators could not afford to be left behind, so they played along. In the UK, five operators paid 35.4B for UMTS licenses. In Germany, six licenses cost the operators $46 billion. In 2002, the first year of large-scale UMTS network sales, operators paid vendors $45 billion. Ericsson and Nokia received about 43% of this amount.
Vendors direct network and service evolution primarily for their benefit, not for that of operators or users. The focus on mobile TV is a perfect example. Watching countless live TV demonstrations at Symbian developer events and at recent 3GSMs, it was clear that streaming/mobile TV applications have had the vendor spotlight for years. The reason is simple. Of all the rich-media services, streaming generally, and mobile TV specifically, places the greatest demand on the radio network, the most precious network resource. So by promoting these services, vendors are forcing operators to upgrade networks and phones with each new radio technology.
Streaming is more demanding than other types of services for several reasons. Firstly, viewing content involves delivering and viewing content in real-time. Content is not stored so delivery and viewing occur simultaneously. Secondly, to meet this real-time requirement, the radio link between the user and network must be sufficiently broad and consistent to enable the user to watch the content smoothly, without delays. GPRS, the vendors argued, could not deliver the requisite bandwidth. Rather, UMTS was required.
Inherent problems
There are several inherent problems with streaming. Firstly, streaming over any cellular network (even UMTS) has proven problematic. One only has to commute on a trial or enter a metro station to know that network access is not constant. Secondly, and more importantly, cellular networks can not guarantee constant bandwidth throughout a viewing experience. The radio resource is limited. When few data subscribers are in the cell, bandwidth is available. When more enter, bandwidth is shared. Thus inconsistent streams and poor user experience are inherent. The situation is worse in crowded areas or peak times, which is usually when services are used. Thirdly, rich-media (even mobile TV) can be delivered in other ways which are more radio-efficient and user-friendly, as will be discussed below.
Clogged networks
Operators realized that even moderate adoption of streaming in peak hours would clog the networks. So they pixilated and shortened the content and, like MMS, stuck it on their portals. We, the users, continued to stand on the sidelines, while vendor and operator solutions served their needs, not ours.
Next came mobile TV. Mobile TV is like streaming, only it involves the real-time streaming of ‘always-on’ multiple-channel TV, delivered to many users simultaneously. It sounds complex. It is complex. However, as with Mobile Internet and MMS, the vendors took a popular experience – TV in this case – and just added ‘mobile’. A natural evolution.
A simple analysis of the network data capabilities readily display the many current problems, as well as the futuristic network/handset features that would have to be in place before mobile TV is possible. Firstly, current UMTS networks can not deliver a decent stream to a few users in a cell. Delivering multiple always-on channels to multiple users on current UMTS networks is not technically feasible. Aware of this, the vendors have introduced solutions that this time, they claim, will solve the radio problems and enable the delivery our much-loved TV to the mobile user. HSDPA today and MBMS tomorrow are the solutions, each requiring their own network upgrades and handset replacements. Alternatively, Nokia is championing DVBH, which requires building entirely new non-cellular networks, not to mention the need for new Nokia handsets.
One would expect that the industry will eventually ask fundamental questions before investing billions on new network components, frequencies and user services. However, the tendency to believe the latest hype is strong. No one wants to ignore industry leaders, analysts or the media, or allow the competitor to move forward alone. However, at some point, should one not suspect the message from an interested messenger?
The key message is this: it’s user and usability issues, stupid. We, the users, are constantly holding and pressing our phones, looking for things to do with it. However, an immediate user experience on the client is mandatory. Instant access to content and functionality is a must. The service is the client. It is what the users sees, and can press and can use immediately. Thus, to take user and usability factors into account means promoting a client experience that provides an immediate access to resident content and functionality.
Emerging alternatives
Despite the vendor and operator structural bias, numerous independent alternatives have emerged which address user and usability issues. For example, On-Device Portals (ODP) seek to overcome the WAP interface by providing client-based immediate access to locally-resident content that mirrors the operator portal. ODP deliver sample content to the handset in the background, usually ringtones, wallpapers and sample MMS clips. This allows the user to effectively and immediately discover the content on the handset, without a WAP connection. Accessing and purchasing the full-version still requires a WAP session, but it is greatly abbreviated. The win for the user is an immediate on-device experience with an attractive GUI. In addition, the access and purchase processes are expedited. The operator wins sees greater revenue from content consumption. Furthermore, the ODP generally mirrors, and thus leverages, the existing portal. ODPs currently operate on existing Symbian devices with Java support emerging.
A second area is that of background download. Client-based background download delivers high-quality rich-media files to the handset without any user-activity required. Whether overnight or while in the pocket, the device receives rich-media content transparently. That content is resident on the handset for immediate viewing when and where the user wants. No WAP sessions are needed.
Background download services are typically subscription oriented, with users subscribing to specific channels, such as video/music clips, TV mini-series or Podcasts. The win for users is that top quality content is immediately available for viewing when and where the user wants. Where once the consumption of media by train/underground commuters was difficult, with background download, captive commuters are the target audience. From a network perspective, congestion is eliminated. Background downloads can be scheduled off-peak. In addition, by avoiding real-time delivery, high quality lengthy content can be delivered. Background download services work on existing Symbian and Java MIDP2 handsets.
The client approach has one major drawback: the client must be customized per handset type, and installed on the handset. Deployment of non-native client solutions adds significant levels of complication. However, many examples of successful deployments of advanced non-vendor client-server services exist. Instant Messaging by Vodafone, Verizon Wireless’ BREW-based converged IM solution, and Push-to-Talk are but a few examples. The network and handset vendors will dislike the client view. Non-real-time services mean using the existing networks better. Furthermore, the clients are developed by others. Worse, the client works on existing, not futuristic handsets.
To summarize, cellular data networks and devices have significantly evolved since voice and SMS. Being vendor driven, data network and service evolution has been, and will remain, technology dependant, with each specific technology requiring expensive network and device enablers.
Will real-time services such as mobile TV remain industry favorites? Yes. Will the portal remain dominant? Yes. However, upcoming client solutions are placing emphasis on user and usability issues. At the end of the network is the user holding a phone. If the user is to use that phone for services, the on-device experience must be immediate and attractive. This requires a client approach.
Implementing a client solution is not trivial. However, considering the investments to date in networks, frequency, marketing and device subsidies, a slight shift in focus toward user and usability might actually show some return.