The AT&T Labs researcher asserts that people are willing to pay far more for point-to-point communications than for famed content. He believes the main role of 3G will be to enhance voice services.
Content is attractive but it is not a moneymaker to the degree that operators need it to be. "What content has is glamour. What it does not have is money," says Odlyzko.
"Content, location-based services, and related novel features are probably best thought of as ways to induce more voice usage."
Odlyzko is not the only one claiming that turning mobile phones into expensive tamagochis or mini-TVs is not the best way to go to increase ARPU. In an article that criticizes NTT DoCoMo's plans to sell books, music, and magazines, even TV shows, columnist Tom Watson argues that the strategy is contrary to what we have learned from the recent history on the Internet.
"If the last six years of investment in the commercial Internet tells us anything, [it is that] faced with offerings from the best and the brightest media minds on earth, consumers still choose first and foremost to communicate among themselves," argues Watson
American broadband consumers do much the same thing that they did with skinny pipes. The number one activity is email. "They just do more of it, more quickly," remarks Watson.
There is some evidence from the industry that Odlyzko's assertions are valid. When it announced its recent financial results, industry giant Vodafone said that it expects voice to generate the bulk of its revenue, between 75 percent and 80 percent in 2004.
Voice services will drive market penetration to 1 billion in 2002.
However, Vodafone also states that subscribers have doubled their spending on messaging and data worldwide, each now spending, on average, seven Euros per month on non-voice services. (Except for the US as SMS had not yet been introduced there at the time).
Certainly, Odlyzko defines content very narrowly and it is hard to disagree that people spent more on telephone services than they did on books and movies when you see the statistics. But when someone like Leo Hindrey of Global Crossing claims the company's services will not become "anyone's dumb pipes", it becomes clear that the two are relying on different definitions.
"What Odlyzko does not know or refuses to acknowledge in the paper is that Hindrey includes application service providing, hosting, etc. in his definition of content," points out Ian Scales, editor of New Carrier magazine.
Hindrey's definition is typical of a telco as he defines content to include ASP and ecommerce-type services. While Odlyzko, as mentioned before is limited to music publishing, TV, and so on.
WAP has failed while SMS thrives because WAP is all about content, while SMS is point-to-point communications, argues Odlyzko.
But those who understand the mobile industry know that WAP has failed for a number of reasons, both technical and economic - such as slow connection times, incompatibility between vendor equipment, too much hype, and pricing errors.
Another disconnect with definitions is where the telecoms industry defines SMS as a data service, but Odlyzko defines it as a point-to-point service. Indeed, SMS is both content and point-to-point, as is email.
In Germany, for example, the majority of Vodafone's SMSs are person-to-person communications, but a fairly substantial 33 percent is generated by information services, such as daily financial news or alerts.
Clearly, Vodafone is keeping its emphasis on voice, supplemented by email, low bandwidth, and SMS. It looks like music and video service will be delayed until much later.
So is voice a 3G killer app?
"We'll each have our own killer app - 3G is a mobile access network. Whatever the application is on the core network, then the 3G services provide a mobile way of getting at it," says New Carrier's Scales.
Odlyzko's argument is one that is worth considering in principle. Voice looks to be important for the medium term. Odlyzko does suggest that data should be thought of as a way to drive or increase voice traffic. Vodafone's experience seems to bear that out.
The good news is that even if it turns out that gaming and video are not hot applications, the 3G infrastructures can support a shift in subscriber priorities. It's less of a problem for traditional operators to support voice or connectivity-related services, as opposed to those which are content-based.
Other network infrastructures, such as ADSL with its asymmetrical channels, don't have it so easy. If consumers decide that they would rather do personal communications, sending video files or streaming content to friends and family, the skinny return pipe may discourage them.
A quick look at a couple of younger alternative mobile operators in Europe, such as Orange (international), Diax (Switzerland) and E-Plus (Germany) suggests that connectivity and enabling of content, not necessarily becoming a TV or video distribution network, is the way forward.
For example, E-Plus just launched its Serviceworld platform, a portal that offers a combination of content, communication and commerce services. Subscribers can access the content and services offered there via voice recognition, WAP, SMS, or their PC. A single email address can be used for both WAP and fixed Internet communications.
Such features improve users' ability to communicate with each other via the network - the essence of what Odlyzko is saying.
Diax, owned by TeleDanmark, also sees voice and point to point communications as taking the lead. Diax spokesperson Monika Walser, told TheFeature.com, "We see voice dominating until 2005. Multimedia applications, such as games and video will only be successful if they are low-priced to consumers and that won't be possible initially. In the short to medium run we will focus on connectivity services, such as messaging, and email.
When asked about its priorities for UMTS, Orange said, "For us the number one thing that UMTS will offer is increased network capacity to handle our new subscribers [in other words more voice subscribers]. The second thing is access to our Orangeworld platform, which offers a lot of different services, video being just one."
E-Plus, will base its UMTS offering on the Serviceworld platform, offering access at speeds 10 times faster than its current GPRS services. The mobile operator plans to offer UMTS in 2002 and is working with KPN Mobile to create a portfolio of non-voice data services (KPN Mobile holds a controlling interest in E-Plus, and NTT Docomo).
E-Plus is arguably one of the most experienced mobile data operators in the region (it introduced HSCSD well over a year ago), so when it says data will enhance voice, it's speaking from authority.
Valerie Thompson is a freelance business and technology journalist, specialized in emerging networking and computing topics. She lives in Zurich, Switzerland.