The idea of packaging up voice fragments into digital Internet packets – a technique known as voice-over-IP (VoIP) – makes great sense. It means separate and expensive voice networks can be scrapped, reaping an immediate return on investment. Existing applications such as call handling, voice mail and email can at last be merged. The number of new applications is limited only by the imagination.

The trouble is, the first iterations of the technology have proved neither reliable enough, nor cheap enough. Voice applications are highly sensitive to latency problems – even a small delay in delivering a packet leads to distortion and loss of sound. That renders the public Internet nearly unusable.

Even where a virtual private network is used, problems occur. Suppliers claim up of 99.9% uptime, which sounds good; but that still means nearly nine hours of downtime a year. Both statistics and Sod’s Law say the telephones will fall silent during the busiest of trading hours.

Inevitably, the scare stories – especially where emergency calls or big customers are involved – have frightened off many customers. For these reasons, many critics say VoIP is a technology that has more intuitive than practical appeal.

To date, customers seem to agree. Suppliers such as market leader Cisco might be selling thousands of IP phones a year, but the facts show that VoIP equipment amounts to little

more than a rounding error in the company’s annual revenue figures. Giga Research points out that less than 2% of all lines installed in 2001 are VoIP compatible.

What will make business buyers overcome their apprehension? And why now? There are a host of reasons.

First, the mobile network operators are already beginning to win over consumer acceptance for advanced digital technologies such as advanced call handling, multi-media messaging, instant messaging, text messaging, and video distribution. But there is a disconnect emerging between what youngsters routinely do on their mobile phones and what big businesses can do for their top people. When the IT directors try to replicate or integrate these facilities for competitive advantage, they will find integrated IP networks, handling voice and data, far cheaper and far more flexible.

Second, the PBX (private branch exchange) equipment suppliers are already losing the battle: investors in non-IP solutions are no longer convinced, customers are buying very little, and managers are trying to plot an IP based escape route. Ask Alcatel, or Siemens, or Ericsson, let alone a pure IP company such as Cisco, where their research and development spending is going.

Third, the investment cycle is right. In the three years leading up to the so-called ‘millennium bug’ in 2000, businesses throughout the world invested in new telephone switching equipment. By 2007, these will mostly have been written down. In order to use the latest facilities, organisations will have a choice of new VoIP equipment, or investing in convoluted bridging technology. Not all will go for IP, but many will.

Fourth, the reliability issue will gradually be resolved. By 2005, over half of the population in the US and in Europe will have broadband connections. It will be resilient and fast enough for VoIP and, in case of emergency, the POTS (plain old telephone service) will still be there. Meanwhile, enterprise suppliers, in conjunction with telecommunications carriers, will offer virtual private IP networks that work ever closer towards 99.999% uptime.

Ultimately, those companies that use VoIP will be able to demonstrate better customer service and better services and applications for employees. They will be able to more effectively leverage their existing IT investments. And, ultimately, they will be more profitable. Who will bet against any technology that offers that?