The cellular industry has repeatedly attempted to port popular consumer services to the mobile environment. Internet became Mobile Internet. Television became Mobile TV. Despite the investment of billions of dollars in data networks, spectrum, devices, and marketing campaigns, very few services have ported successfully.

Yet digital music and podcasting prove that users will go to great lengths to mobilize entertainment, including actively connecting a media device to a PC and transferring to it content downloaded from the internet. But can podcasting become a cellular service enjoyed on handsets? Clearly, podcasting has certain attributes which make it suitable for the mobile environment. First, it is an "on-the-go" experience. Second, enjoying audio content is not effected by the handset's small display screen. In fact, given the prevalence of mobile phones, coupled with the ability to deliver content directly to the handset without any user action required, the mobile industry might be hard-pressed to explain a porting failure. Indeed, one may argue that such failure should challenge hyped concepts such as convergence. This article outlines a few of the critical issues that must be addressed if podcasting is to see even minimal mass-market penetration. First, what are some of the inherent "mobile-environment" constraints and how will they impact and define the service?. Second, is there a user willingness to pay for, and operator desire to launch, such mobile podcast services?

The manner in which mobile users discover and receive content will have a huge impact on the nature of the service. There are two alternative models: network-based solutions, and client-based solutions. Network-based solutions offer users access to podcast menus on the Operator's WAP Portal. Users, locate the appropriate podcast, then initiate a download or stream of the podcast in real-time.

Network-oriented delivery models have failed to appeal to the mass-market user. The click and wait, menu-intense experience of Mobile Internet has proven unappealing. It is doubtful whether posting podcast files on a Portal will be an effective way of increasing awareness and usage of the service. Furthermore, given the relatively large size of a podcast file, adding a lengthy download wait to a cumbersome Portal experience will kill the experience all together.

Podcasts can also be streamed off the Portal. Here, however, in addition to the cumbersome Portal-Pull issues, the user-experience becomes dependent on consistent and sufficient data transmission during the stream. For reasons beyond the scope of this article, providing bandwidth for short streams, not to mention lengthy podcasts is technically challenging. A user listening to a podcast while commuting by train will frequently lose coverage. Securing bandwidth in peak-hours or in congested areas is very difficult. It is thus doubtful whether streaming can deliver the mass-market with an acceptable level of service.

Whether downloaded or streamed, obtaining content via pull assumes that a user will regularly poll for content. Not only does the active user concept runs counter to the Podcast model of automatic content delivery, but a compelling mobile experience must be simple and automated. One must consider that the potential mass-market mobile user is not as "early-adopted" oriented as a current podcast user. Thus, the user-experience on mobile user must be as good, if not better than the iPod experience for the mass-market to accept it.

Client solutions reduce the amount of browsing associated with content discovery, delivery and consumption, and provide a more immediate, user-friendly experience. The first type of solution, offered by Pod2Mob, involves a client that displays a catalogue-list of available podcasts. The user scrolls down the list and selects one, which initiating a content delivery session (download or streaming). Content discovery is easier than in Network-based solutions, as WAP browsing to the portal is avoided. However, real-time delivery is required, resulting in either consumption delays, streaming-related problems, or coverage loss. With this solution, an active user is assumed, as a consumption decision must be made daily.

The second client solution, such as offered by MobiPod (a collaboration between www.Bamboomc.com and www.britcaster.com ) involves background download, where large files are delivered to the user transparently, without any user involvement required, for example overnight. Fresh content is available for immediate consumption for the morning commute with no network access required. Background download usually require subscription.

It must be noted that the transition from pull to push involves a conceptual shift among operators. Operators have invested heavily in WAP portals such as Vodafone Live. One key Operator goal is to drive users to the Portal, which strengthens operator brand. From a user-perspective, however, it is crucial to reiterate the assumption that the average Podcast user is more technically-orientated than the average mobile user. Ease of use is absolutely essential if mobile podcasting is to gain any degree of mass-market traction in the mobile world.

The second section of this article asks whether users will be willing to pay for mobile podcast services, and whether the operator will actually want to launch anything but a barebones service for PR purposes.

From a user-perspective, there is a significant rise in the number of people carrying MP3 players, media-enabled phones, and other media devices. People are clearly taking their entertainment with them. Also, working people have clearly definable windows of dead time while commuting to and from work. During these times, they are a captive audience. Will the mass-market, which holds mobile phones rather than other media-devices, be willing to adopt and pay for services which deliver personalized audio content to them?

One barrier might be the availability of free podcasts on the PC and the initial perception that internet data is and should be free. Whether users are willing to pay for personalized audio content on their mobile will depend of factors such as easy of use, content quality, and price. However, given the growing prevalence of people enjoying entertainment on the go, one does not have to invent a scenario of commuters enjoying a 15 minute targeted audio-program on the way to work. True, Podcasts are available free on-line, but it is quite likely that people will pay a small premium in order to receive interesting content on their mobile phones, rather than buy an iPod and then bother with transferring content from their computer to a device each morning.

One thing is certain: the operator has a keen interest to see the success of such operator-provided services. First, from a revenue perspective, operators often subsidize the handsets, yet sees no revenue when a user transfers music to it from the PC. Second, should the mass-market adopt iPod-like devices as their device of choice for media consumption, the mobile handset will be marginalized and viewed only as a tool for voice-calls. As these competing devices develop Skype-like internet telephone functionality over WIFI, operators will see their users gravitate to competing phone service as well. It is imperative for the operator that the mobile phone claim a firm stake as a media device, and that users load it with content that generates revenue for the operator.

Despite the operator interest to establish value and compete with encroaching devices, mobile podcasting poses a few challenges. While PC-based internet users enjoy inexpensive broadband, mobile networks are comparatively inefficient. Data transmission rates are slower and there is much less overall capacity. Thus, the internal cost to the operator of transmitting data is high. While the monthly charge for a high-speed residential internet connection might be 20 Euro, the average cellular user might be charged 1 Euro/MB for data usage. With the size of an average 30-40 minute PC-based podcast approximately 15 MB, the monthly amount of data traffic per user for a week-day service is 300MB!! The operator can not justify charging of a few Euros a month for a mobile podcast service, when a single Pull-downloaded video clip can generate a Euro or two.

Can mobile podcasting be made more efficient?

First, content files can easily be reduced in size by simple content transcoding. A 30 minute podcast can be reduced to 1.5MB, without impacting sound quality. Furthermore, the delivery frequency of a podcast service can be reduced. Finally, delivering shorter, 15 minutes podcasts, may be appropriate.

Second, and more important, the podcast files must be delivered during off-peak hours, ideally overnight. During peak hours and in congested areas, the cost of data delivery is at its highest. Delivery of large data files to a moderate number of users during peak hours will chill operator enthusiasm. Conversely, during off-peak hours, the network is empty, minimizing the cost of data transmission. This requirement would appear to point to a subscription push service model, with scheduled off-peak content deliveries pushed to the user.

Finally are billing and revenue. The operator is accustomed to pull-based charging models, where the user is charged per transaction. However, given the complexity of data billing, and the relatively large size of a podcast file, mobile users will only adopt podcasting if the pricing structure is clear and reasonable. A transparent subscription fee for the service, without any additional data charges, is mandatory. In terms of additional operator revenue potential, one point worth noting is advertising. Audio advertisements can easily be included at the beginning and during the podcast, creating further revenue sources.

By Monte Silver mobilemonte@hotmail.com