Telepresence: It's Not That ExpensiveTelepresence: It's Not That Expensive

Part of the ongoing conversation about telepresence is about its high cost, which includes both the initial cost of the telepresence suites or endpoints, and the ongoing cost of the network and services to support it. Telepresence suites can cost $300K plus, and the cost of the service and network can be as much as $18K/month per location. Ouch. But let's take a look at what this really means to the enterprise.

Part of the ongoing conversation about telepresence is about its high cost, which includes both the initial cost of the telepresence suites or endpoints, and the ongoing cost of the network and services to support it. Telepresence suites can cost $300K plus, and the cost of the service and network can be as much as $18K/month per location. Ouch. But let's take a look at what this really means to the enterprise.In my conversations with enterprises I am finding out that a quality telepresence installation is in high demand. Hewlett Packard told me their internal rooms that have been in place for six months or so (so that their users have discovered them) are being used 9 and 10 hours a day. Employees come in early to talk to colleagues in Europe, and stay late to have meetings with colleagues in Asia. These are big numbers. In traditional video conferencing installations we most frequently see rooms that are used both for regular meetings as well as for video, and the video is only in use a few times a week.

Andrew Davis of Wainhouse Research did some direct analysis on this topic, using primary research data he collected from enterprises. Here is a graph he presented at VoiceCon in Orlando last month:

So here the conclusion is that telepresence is cheaper by the hour than traditional video conferencing. A quick backwards calculation shows that Andrew's assumption for telepresence room use is only on the order of 6 hours a day, not the 9 or 10 I talked about above. Still this shows that telepresence rooms are much more highly utilized than traditional video setups.

Now the numbers above are also calculated using $8,000/month as the operating cost instead of the $18,000 I quoted above. If the costs were really $18,000 it would add $90/hour to the telepresence usage cost. One of the readers of my last post said they eliminated their telepresence rooms because of the high cost of the network. So let's take a look at this component and see how it can be addressed.

The key requirement is to put in place sufficient bandwidth for the telepresence traffic, above and beyond what was required by the business before telepresence was installed. The WAN service with the best offerings in this space today is MPLS based with clear SLAs that address the need for video to have very low loss and jitter, and latency that is reasonable given the geographic distances involved. Unfortunately, the cost of this bandwidth can vary widely across geographic regions around the world.

The first logical approach is to find a WAN service provider with a sufficiently large footprint to be able to supply service to all the enterprise locations of interest. If a company is primarily based in North America, Europe and key developed locations in Asia, this may be a good and cost effective solution. If the company goes further afield and includes locations where telecom is still highly regulated, the costs may jump up. Additionally the enterprise may find that certain carriers have good coverage for portions of their needs but not such good coverage in other regions. These carriers may offer partnering agreements with additional service providers, but now you are crossing from network to network, and getting a real guarantee on the traffic quality and QoS parameters is hard and potentially expensive.

Where a problem like this exists, there is usually an entrepreneur providing an answer, and in this case one of the solutions comes from a Toronto-based company called IPV Gateways. These folks provide a switching service that will connect two or more WAN service provider backbones and maintain the QoS across that connection. Think of them as a high quality peering point between dedicated enterprise VPNs on multiple carriers. This kind of service, and I am sure others will develop, will help make high quality high bandwidth pricing more transparent by allowing enterprises to pick the best carrier in each geography, and bridge between them to support video across the enterprise.