One of the key issues in implementing VoIP is cost. Until around 2005, organizations that implemented VoIP did so because of its real or perceived cost savings over traditional telephony.

In many cases, they did find substantial savings by eliminating costly third-party contracts for moves, adds and changes (MAC), reducing the amount of cabling required in new buildings, or leveraging idle capacity in their data networks. Indeed, those savings existed, but so did additional costs -- effectively negating any net savings for the first 12 to 18 months.

For the past two years, cost savings has been a secondary driver to "futureproofing" the network. IT executives see IP as the platform for the future, and they want to be prepared for new applications -- starting with voice -- within a converged infrastructure. As a result, they are more accepting of VoIP's hidden costs, such as consulting, training and ongoing maintenance.

The focus has shifted from proving the business case of VoIP vs. TDM to the business-case vendor comparisons. In other words, it's no longer a matter of whether a company switches to VoIP; it's a matter of when.

In Nemertes' benchmark research, 55% of the 120 IT executives who participated said they started their VoIP implementations within the past two years. Still, only about 18% of companies surveyed have completed their VoIP rollouts, and those are primarily small and midsize businesses (SMB).

As these IT staffs assess VoIP, they want to evaluate the cost components -- and how key vendors compare. During that time, Nemertes has been tracking VoIP costs for four years and has interviewed nearly 400 organizations.

What companies spend on VoIP depends largely on a few factors: how large their deployments are, which vendors they use and how they design networks. The trouble is that most IT executives cannot calculate accurate costs until they are well involved with their rollouts, particularly with respect to start-up and ongoing operational costs.

It's vital to understand the cost components of VoIP. For example, it's great to find very low-cost VoIP deployments, for instance, with an open source solution. But if that solution requires significant training and programming and support is mediocre, it may be worth spending the extra money to go with a solution that is not open source.