Is there early take-off phenomenon in diffusion of IP-based telecommunications services?

Abstract

Traditionally, the diffusion of telecommunications services has been considerably affected by the presence of critical mass and network externalities, and thus has shown the so-called “late take-off” phenomenon. However, as telecommunications networks evolve from circuit switching to packet switching, especially IP networks, and thus enabling diverse new services, it seems these traditional telecommunications diffusion patterns are changing. By comparing the diffusion of IP-based services with those of circuit-based services and durable goods, we have found the late take-off phenomenon is not involved in the diffusion of newly introduced IP-based services. Moreover, we concluded that the diffusion of IP-based services is much faster than that of durable goods, thus showing an “early take-off” phenomenon. Based on this empirical result, we suggest that telcos (telecommunications companies) and (ISPs) Internet service providers should pay as much attention to the growth stage as the introduction stage in their development, in order for their IP-based services to be successful in the market.

Introduction

One of the typical characteristics of networked industry, network externality, is defined as a quality of certain goods and services such that they become more valuable to a user as the number of users increases [1]. In the case of telecommunications services such as telephone, FAX, and the Internet, the utility of the existing subscribers increases as new subscribers are added to the network [2], [3], [4]. This is called positive network externalities or the “network effect,” and it is believed positive network externalities affect the rate of adoption [1], [3], [5].

Alongside network externalities, another distinctive feature shown in the diffusion of telecommunications services is the existence of a critical mass. Critical mass is defined as the minimal number of adopters of an interactive innovation for the further rate of adoption to be self-sustaining [6]. Considering the purpose of telecommunications services is to connect people so that they can communicate with each other, it is crucial for telecommunications service provider to reach critical mass [7]. Thus, the existence of critical mass, combined with network externalities, affects the diffusion pattern of telecommunications services. That is, a telecommunications service diffuses slowly in the early stage of diffusion, and the diffusion speeds up once critical mass is reached [8]. This is called late take-off phenomenon since the take-off time of telecommunications services occurs later than that of durable goods. The presence of critical mass and the resulting late take-off phenomenon give telecommunications service providers an important strategic implication. For example, through empirical research on the French PTT's Minitel, a videotext service that was successfully introduced, Allen [2] suggested that promotion strategies such as subsidized pricing and free trials of services in the introduction stage were critical for Minitel's success.

However, as telecommunications networks evolve from circuit switching to packet switching, enabling diverse new services such as Internet provision services like ADSL, cable modem Internet access service and Internet lease line service, and Internet application services like Internet banking and web hosting and so on, these traditional diffusion patterns are changing. In this paper we investigate the diffusion patterns of new IP (Internet Protocol)-based telecommunications services as a pilot study, and conclude that diffusion patterns of new services are quite different from those of old services, at least in the Korean Internet industry.

The paper is arranged as follows. In Section 2, we review previous research related to network effects and critical mass in the diffusion of interactive innovations which cause the “late take-off” phenomenon. In Section 3, we identify the “late take-off” phenomenon in the diffusion process of traditional telecommunications services in comparison with those of durable goods with the Bass diffusion model. In Section 4, we analyze new diffusion patterns of IP-based services and compare them with diffusion patterns of old circuit-based services and examine some factors triggering the “early take-off” phenomenon of IP-based services. Finally, in Section 5, under the “early take-off” phenomenon of IP-based services, we discuss the implications for telcos and ISPs and suggest further research, considering the limitations of our study.