We have also included reasonable estimates for staffing costs

I am sure Brent, Dave, and Marty followed a meticulous process, and their conclusions are no doubt accurate based on the question they attempted to answer. My position is that they answered the wrong question. TCO comparisons are inherently difficult because of the required underlying assumptions.

Normalizing assumptions between two competitive solutions is always difficult, but even more so among different deployment models. Not long ago, No Jitter found itself in a TCO debate over premises-based solutions from Lync and Cisco. The core problem was that the direct and indirect costs of Lync vary wildly based on a site's overall Microsoft commitment and investment, resulting in multiple (correct) TCO conclusions.

Consider the TCO question when it comes to something as simple as buying or leasing a car. The key assumption that drives TCO is the term of the lease. Other key assumptions that impact the outcome include estimated resale value, estimated mileage per year, depreciation rate, and so on. The buyer tailors these assumptions to determine the correct path for their particular situation. Two buyers looking at the same vehicles may come to perfectly sound different conclusions.

The authors of the TCO study made a number of assumptions about enterprise UC requirements, and then attempted to normalize the bids for fair comparison. They state, "The three RFP authors worked together to create a normalized and consistent TCO comparison for all 23 of these solutions." Direct comparisons are difficult without normalizing the level of functionality and services quoted.

Unified Communications deployments are typically highly customized, and organizations select exactly what they need, integrated into their specific framework, and aligned with their organizational objectives. For example, the authors stated that "if a solution did not include audio or web conferencing, we have added in reasonable costs for procuring these capabilities separately." Inherent in that statement is the assumption that the customer wants audio and web conferencing as part of their core UC solution--two services that many organizations use a la carte today.

It also assumes "reasonable costs for procuring these capabilities separately." How organizations conference varies significantly. Some are mostly internally focused, putting more emphasis on the equipment; others rely on more distributed uses, which puts additional emphasis on access; and some are video-centric. An organization that regularly uses video with external organizations may find itself a prime candidate for a cloud service like Vidtel or Blue Jeans--regardless of the UC model deployed.

"We have also included reasonable estimates for staffing costs." That's noble, but staffing costs are only a part of the human equation. There's a shortage of both Cisco and Microsoft UC engineers, so additional costs include position search, time to fill, turnover and retention costs, and ongoing certifications.

One of the big drivers of cloud adoption is to reduce the burden of UC on management. In addition to the obvious issues of delivering core services, linux dedicated server management attention also goes toward staff management, resource management.

With premises-based implementations also come issues like border security. To mobile-enable staff often requires specialized equipment, licenses, and additional bandwidth. Two colleagues communicating from remote locations over mobile clients requires two connection paths to the server--a duplication that is eliminated with UCaaS.
Click on their website www.mileweb.com/public-cloud for more information.