Two weeks ago, while I was preparing the piece that ultimately was headlined Unifying Unified Communications for ITBE's site Unified Communications Edge, Radvision vice president Robert Romano said something very interesting to me.
The goal of the piece was to define three levels of standards and/or proprietary protocols that will facilitate unified communications. My idea-which I tested out by speaking to experts such as Bob-was that connections were needed at three levels.
The first is at the granular level. VoIP, IM, videoconferencing and other communications channels must be supported.
The second stop is at the intra-vendor level. Thus, VoIP, IM, video conferencing and other communication channels offered by a vendor must be seamlessly integrated for the platform to operate most efficiently. More superficial linkages will be inefficient and overly complex.
And, at the top, unified communications platforms from Avaya must be able to speak to platforms from Cisco, etc. Companies merged through acquisitions usually don't come to the first joint meeting with the same UC vendor under contract. Nor do companies entering a long-term work arrangement have the same UC vendor, except by chance. Good standards and protocols can enable the disparate systems to work together.
I was discussing these issues with Romano when he made an intriguing point about video:
When I take a look at value, video probably last from a cost/value perspective. It takes the most bandwidth, is the least tolerant of packet loss and latency. But you can live without video at the end of day -- but you can't live without voice.
The comment wasn't directly related to the story I was writing, so it didn't make it into the feature. But it raises a very interesting issue, especially considering that video -- both in its standard and high definition flavors -- clearly is becoming a big part of the future of UC.
The bottom line is that building a UC infrastructure capable of delivering video may not pay off in the short haul if the organization's UC goals are strictly data-centric. However, putting an infrastructure in place that is capable of handling the most demanding application from the ground up-even if it is not deployed immediately-seems to be a smart approach. (Mike Vizard at CTO blogged today about a trend of using application accelerators to bolster video traffic.)
The bottom line is that organizations must decide whether it is wiser to create suitably robust networks from the ground up or to wait until a decision is made to deploy demanding application.