As I gathered my thoughts in preparation for my Virtual VoiceCon (registration is free) keynote presentation this Wednesday I reminded myself that significant changes in the enterprise telephony market take time to gain a foothold and spread to the masses.Perhaps no major change has taken more time to evolve than the paradigm shift from a hardware- to software-centric system design. Microsoft may currently be promoting its software-powered vision of an enterprise communications ecosystem, but as I have pointed out in earlier No Jitter blog posts and feature articles the software age began in the early 1970s with the introduction of the first Stored Program Control (SPC) PBX by Northern Telecom. It took more than 30 years before the development and availability of open source PBX solutions using third party hardware gained hold in the market. Most of the old line PBX suppliers now market solutions using OEM hardware elements ranging from servers for call control and features to media gateways for port interface requirements. I have heard complaints about closed, proprietary telephony systems from the time I entered the market three decades ago, but hope the chatter will soon fade.
Looking back at several other major technology advances that revolutionized enterprise telephony system design, such as digital and packet switching and signaling protocol format, one must walk away with the conclusion that market acceptance and customer implementation is measured in years, not mere months. Though digital PBXs arrived in the 1975 time period, AT&T continued to retain more than half the annual domestic market share until the early 1980s with an analog switching system platform, and it took more than a dozen years for half of the domestic PBX installed base to migrate from an analog (much of it AT&T) to digital switch design. Similarly, the first digital telephone instruments were introduced in 1980, but analog station shipments continued to outpace digital until 1990. It wasn't until the mid-1990s that the installed base of digital PBX telephones outnumbered analog.
The evolution and market acceptance of IP telephony systems followed a similar pattern, though it occurred at a slightly accelerated pace. The first IP telephony systems were shipped before the end of the last century, but circuit switched design platforms continued to dominate until the middle of this decade. Until recently, there remained strong system design elements of traditional digital PBXs in many products shipped by market leaders such as Avaya and Nortel. It took less than ten years after the first IP telephone instrument was commercially shipped for annual IP shipments to outpace TDM (analog, digital) shipments, and the installed base still remains more of the latter than former (though the tipping point is very close).
Between the market growth of digital PBXs and the introduction of IP telephony systems came another technology advance that was supposed to change the look and function of enterprise telephony, but utterly failed to live up to even the most conservative expectations: Computer Telephony Integration (CTI). The recent attention given to Unified Communications (UC) eerily parallels that given to CTI during much of the early-mid 1990s. Articles about CTI proliferated in the pages of Business Communications Review and CTI presentations at "PBX in the 90s" (the original name of VoiceCon) changed the nature of the conference for a few years. Magazines and conferences dedicated to CTI soon appeared on the scene, until all things CTI faded by the end of the decade. CTI did not replace the need for advanced PBX telephone instruments, nor did it make the core PBX system obsolete. It's interesting to remember that a major player pushing CTI was Microsoft, the designer and developer of Telephony Applications Programming Interface (TAPI). As regards Microsoft's initial push into the enterprise telephony market, it seems that some things never change.
As we are about to begin the second decade of the century, it is beneficial to remember that SIP has been among us for many years, as have other now hot industry topics such as Cloud Computing (dating to the early 1990s concept of Grid Computing), and virtualization (going way, way back to the 1960s partitioning of mainframe computers). For those that don't remember the past, or are too young, it seems that everything old is new again. I don't worry that my field of expertise will be extinct anytime soon, because I will evolve with the market at a nice slow pace befitting my advancing years.
