Survey says: Voice over IP in five years

Voice over IP (VoIP) will achieve broad-scale implementation within a five-year time period, according to the majority of those surveyed for an independent global study done by Feldman Communications Inc.

The study, titled VoIP ’99, revealed that 96 per cent of the executives, analysts, IT managers and members of the media surveyed anticipate high-quality voice can and will be placed over data networks in the foreseeable future.

“People have been talking about this technology as being around the corner — we wanted to find out how big that corner was,” said Paul Feldman, president of Feldman Communications in Annapolis, Md.

“So we wanted to ask participants who should know more about it than anybody else what their expectations were in terms of delivering this technology.”

Nearly three-quarters (73.1 per cent) of the survey’s participants said they foresee that 15 per cent to 20 per cent of all voice traffic will run over data networks within the “near- to mid-term.” The study also found 83 per cent of respondents believe VoIP will be “broadly used” within the next five years. Of those, 28.6 per cent anticipated a considerable movement to embrace VoIP usage in the next two years or less.

In answer to a question regarding where the “greatest long-term applications” of VoIP would take place, large long-distance or telecommunications carriers led the pack with 41.1 per cent of respondents’ votes. Major enterprises with remote locations ranked second with 33.85 per cent, followed by ISPs offering voice as content with 17.7 per cent, and competitive local exchange carriers at 16.9 per cent.

As for barriers to VoIP being widely used, 61 per cent of respondents said one of the greatest obstacles would be that voice quality would be sacrificed. Service quality being inferior to the public switched telephone network was cited by 52 per cent as one of the biggest barriers, and 27 per cent said the problem was functionality will be diminished.

The same question showed that 19 per cent considered the hurdle of “initial investment” or the “challenge of getting voice and data personnel to collaborate” as a significant barrier to an organization adopting VoIP. (Respondents were allowed to choose more than one barrier to VoIP adoption.)

More than one-half of those participating in the survey had at least 10 or more years of experience in the industry, and two-thirds had six or more years’ experience.

“Most of the participants had tried voice over some type of data fabric, particularly IP,” Feldman said.

The survey was linked to the company’s Web site, but staff proactively approached 684 network managers, IT managers, CIOs, analysts and members of the media with the fairly detailed surveys. Feldman Communications received back 124 fully completed surveys. More of the respondents were from the telecom industry than any other field, followed by the Internet/e-commerce field, then the IT and software industries. As for positions, executive management sent in the most entries, with twice as many responses as any other group. Marketing and strategic planners followed, and everything else, including media and analysts, trailed after, according to Feldman.

When asked about the survey’s findings, John Armstrong, principal analyst with Dataquest Inc. in San Jose, Calif., said he thinks five years is a reasonable time frame for VoIP providers to be able to match the quality of what we see today on the public switched telephone network, but he doesn’t necessarily see it being implemented on a broad basis by then.

“It’s going to take some time because there’s a lot of infrastructure that needs to be upgraded on the network,” he said. “So in terms of general broad availability, I’d say the time frame is longer than five years.”

But another analyst, Thomas Nolle, president of CIMI Corp. in Voorhees, N.J., doesn’t expect VoIP to become viable at all. While he does agree data networks will have the quality-of-service capability to support high-quality voice, he said: “The real question is, do you believe data networks will absorb any consequential portion of high-quality voice traffic? And I would answer that question ‘absolutely not.’

“You cannot do voice compression without introducing enough delay to be perceptible to the listener, period. It’s not possible.”

Nolle therefore doesn’t think there will ever be a significant amount of what we would call voice over IP.

“I think eventually there are going to be alternative voice networks, but before voice is transformed into anything other than what it is today on a large scale, all of our current transport technologies — including IP — will also be transformed.”

He said he doesn’t put any stock in surveys such as the one conducted by Feldman Communications.

“Reality is not defined by a public opinion poll; only mass hysteria is defined by that,” Nolle said.

Carol Neshevich is a senior writer for Network World Canada