A recent survey of developers of the Asterisk platform, an open source telephony technology, revealed a surprising number of companies are choosing to develop their own Private Branch eXchange (PBX) systems.
Those companies preferred to rely on in-house IT resources than work through integrators, or purchase off-the-shelf products from established manufacturers.
Conducted by media-processing hardware and software vendor PIKA Technologies Inc., the survey’s results were based on 322 Asterisk developers globally.
As many as 30 per cent of respondents were building in-house PBX systems. This surprises Terry Atwood, PIKA’s vice-president of sales, marketing and customer care, given the amount of IT and telephony knowledge typically required to deploy and support open source platforms.
“There’s a lot of work being done to make it easier, but it’s still not an easy thing to do,” Atwood said.
Despite the complexity of building and supporting an internal open source telephony system, he said, some users like open source platforms because they are free. “Companies like Nortel, Avaya and Panasonic have fairly substantial markups on their products.”
Besides being free, “for the technology geek, [Asterisk is] open and easy to modify”.
Another driver behind open source is the inability to find an existing system that offers specific functionality, such as an Asterisk-based paging system, said Irwin Lazar, principal analyst and program director for convergence and collaboration with Mokena, Ill.-based Nemertes Research Group Inc.
Actually, a similar percentage of companies that internally develop Asterisk systems emerged in a study by Nemertes, where the majority of companies were found to be small businesses.
Larger companies just can’t justify spending money developing what they can easily buy ready-made, said Lazar.
Besides, larger enterprises are generally more comfortable with systems by established vendors because they realize that while it may be more costly up front, the maintenance and management costs will be lower in the long term.
“When you’re using open source software that you downloaded from the Internet, you have to go to a message board to find the fix for something,” said Lazar.
The Nemertes study found those companies that initially considered Asterisk development were dissuaded by the lack of system support.
But while the idea of an open source platform is more palatable for small companies, Lazar isn’t so sure they can depend on it being scalable. He said he wouldn’t recommend using Asterisk for more than 5,000 users, but that would also depend on which features are needed.
Deploying and maintaining an open source PBX system internally is not difficult for most developers because they have the required technical expertise, said Atwood – difficulties arise when IT teams without the necessary skill are swayed by the zero cost and promise of system flexibility down the road.
Once they realize the project is overwhelming them, he said, the only options are to abandon development or inject money in order to continue building.
Most companies might decide to spend more to continue what they started, but really, the situation is not vastly different from the majority of IT implementations that overshoot schedules and budgets, said Atwood.
Although he thinks internal telephony expertise is growing, he doesn’t suggest companies build their own telephony systems. Instead, he recommends users contract with one of the many companies that make it their business to integrate Asterisk into their own applications.
Nonetheless, that surprising number of internal developers signals acceptance of the platform, said Atwood. “The success that the Asterisk community made from the hobby shop technical guru to business mainstream I think is going a long way towards proving there is credibility and reliability in open source in telephony and elsewhere.”
Asterisk is often used for research in government and education. But open source telephony will eventually make its way to the enterprise arena, believes Atwood, albeit in five to 10 years – a transition aided by increased reliability and adoption.
Some large multinational organizations are already considering deploying Asterisk systems, he said, “and when it’s getting that kind of corporate attention, it’s going to further the movement for sure.”
Enterprise acceptance will require companies, like Digium Inc., the Huntsville, Alab.-based company behind Asterisk, to provide a support arm to the open source platform, thinks Lazar.
“That’s what’s missing,” he said, recalling that Linux’s enterprise success didn’t happen until RedHat emerged to support it.
Ironically, Digium began offering a closed source version of the platform to rake in revenue from the technology, noted Atwood. Although he believes an open version will always exist, “the real money maker will be people taking it and closing it like Digium.”
But the Asterisk community can make money by providing system integrators with developer tools and other types of support, said Lazar.
Appliance companies, too, like Los Angeles,-based IP PBX system vendor Fonality will help leverage the Asterisk community, he said.
