This is probably one of the more controversial debates between the open source community and proprietary software vendors. Everyone has something to say about which side has the cheaper total cost of ownership (TCO), with dozens of tools available to test your TCO, all of them producing slightly different results.
Gartner introduced the TCO concept as a metric to gather data about any hidden cost that may impact on migration and new system implementation. Factors like administration, downtime, retraining were, therefore, included.
Depending on which side of the fence you sit on, open source can be much cheaper or more expensive than commercial and proprietary software (CPS). For example, training and supporting a thousand users, who are used to the ins and outs of Microsoft Office, on a rival software package can prove to be a very daunting and frustrating task.
But the very concept and viability of TCO has often been questioned, especially when compared to the Total Benefits of Ownership (TBO), which takes a bigger, holistic point of view.
Ultimately TCO is about money saved or spent, calculating how much it might cost a particular company to run a particular system. Rather than being a forecasting tool, it has been argued that it is more reliable as an accounting one.
As a word of caution, do not use TCO on its own for best results. Use it in conjunction with TBO and return on investment (ROI) assessments.
You can read about the ten other myths of commercial and proprietary software here.