The Mermaid of Convergence

 Last update: 2005/03/08

Starting with the 1980s, the increasing number of cases where data processing and telecommunication systems needed interaction or integration led people to believe that the corresponding businesses would converge. A superficial reading of the business implications of this technological convergence - note that even this type of convergence is questionable in many instances - made the top management of IBM and ATT think that they were bound to become competitors because they shared much of the basic technology in their products and services. 

So ATT, whose Bell Labs could brag to be second to no one in information technology - witness the successful UNIX story - tried to create a computer company inside the group and, when it failed, invested billions of dollars to acquire the highly successful NCR, and bundled it with its other computer-related properties. The resulting entity was transformed into a money loser in no time. The end of the story was written a few years later when ATT decided to spin off its newly acquired computer company and, since it was already at it, it also spun off its old manufacturing arm bringing to an end the glorious Bell System and, with it, the Bell Labs that people used to know. 

In parallel, IBM had developed a global network to connect its dispersed business units. Having done the job for itself, it started selling communication services to other companies using that network. Eventually IBM realised that telecommunication services were not really its mission and decided to shed the business because it was "non-core". Whom do you think they sold it to? ATT! 

With the application of digital technologies to audio and video - the realm of entertainment - made possible by MPEG, the idea of a further convergence between the businesses of entertainment, telecommunication and computers took further root. Today the "convergence frenzy", as many other frenzies, has fortunately taken a more sober attitude, but it is still worth revisiting the saga so as to avoid any future confusion about technology-induced dreams and business reality. 

The first claimed aspect of convergence concerns hardware. Its thesis is the following: there is manufacturing convergence because different industries use the same types of components. This is a very naïve reading. For several decades integrated circuits were a common technology for a multitude of businesses: telecom switches use computers, computers use serial data transmission chips, CE devices use microprocessors, and so on. ATT and IBM thought that technology convergence would put them on a collision course. The result? They are as far away as ever.

The second claimed aspect of convergence concerns software. Its thesis is the following. broadcast devices, such as radio and TV, and telecommunication devices such as telephone and fax have always been rather "dull". But now users expect more from their devices, e.g. more advanced application programs made possible by more advanced operating systems. So expect to have convergence in this area as well. This is again a naïve reading, but with an element of truth, only because the more traditional industries are slower to adapt to the new world, not because software technologies provide a tougher barrier than hardware technologies. 

The third claimed aspect of convergence concerns the business domain. Its thesis is the following: telcos buy CATV operators, CATV operators buy Hollywood studios, CATV and satellite operators buy terrestrial broadcasters, CE companies buy Hollywood studios, Internet Service Providers (ISP) buy content companies, and so on. The response to this claim is very simple. Business convergence happens because of abundance of capital and lower legislative constraints, not because there is any more internal compelling reason for this to happen than there used to be. The story of the acquisition of Time Warner by AOL teaches a lot.

The idea of companies trying to extend their reach over the entire value chain is by no means new, what changes is the driver. In the early decades of the last century, RCA saw radio broadcasting as a means to entice consumers to buy radio sets. It set up the National Broadcasting Company (NBC) to feed the market with content so that people would buy its radios. Fortunately those were years where a glimmer of rationality still existed and PAs forced the use of a broadcasting standard, so that people could still tune in NBC even if they had not bought the radio from RCA and in turn people who had bought an RCA radio could tune in other stations than NBC. The same happened with music records which were seen by RCA as a means to entice consumers to buy record players and so Columbia Records was established. Legislative limitations that are now being weakened were established decades ago, exactly because capital-rich companies have always aimed at other companies next to their businesses, not necessarily because there was "convergence". 

A second example is the fact that MPEG-1 and MPEG-2 were conceived as universal languages to represent digital audio and video, so that each piece of digital audio and video can move from one industry domain to another without consideration of the lower layers: the same "video file or stream" could move from/to satellite, CATV, VoD system, DVD etc. The reality, however, is that Video CD is a replacement of VHS, DVB is little more than a better exploitation of satellite and cable bandwidth and DVD is just a replacement of laser disc. Universality of MPEG digital language notwithstanding, we see that, so far, content companies tend to restrict the flow of their content within the traditional industry bounds.

The concept of convergence starts making sense if we see it as "the transformation of a set of so-far vertical businesses into a logical set of horizontal businesses". This can be achieved when there is a reduction of dependence of a layer on the technology used at a layer below. This is what the computer industry has achieved by moving from a situation in which each computer manufacturer had proprietary technology elements, starting from components and platforms up to OSs and applications, to the situation of today where applications are almost independent of the CPU, the computer platform and the OS. 

The following table provides a comparative analysis between digital television and the PC. Looking simply from the technology viewpoint one could say that there is no difference because they both have a CPU with RAM, an OS, applications, peripherals, network interfaces, user interaction devices and audio and video outputs. However, apart from these technological similarities, all the rest is different. Even though the situation is evolving, but at a slow pace. The conclusion is that we are still a long way from convergence between computers, entertainment and telecommunication.

Copyright © 2003 chiariglione.org