Actual cost experience on prototype units, early engineering development hardware and early production hardware for the program under consideration should be used to the maximum extent possible. If development or production units (or components) have been produced, the actual cost information should be provided as part of the documentation. Estimates for Full Rate Production decision reviews are to be based at least in part on actual production cost data for the systems under review.

Estimating by actual costs is, essentially, an extrapolation of current program cost. In other words, current contract cost trends are used to estimate final system cost. The cost data is internal to the current system being constructed, which is not the same as “actual” historical data. There are several conditions that must be present to enable actual cost estimation. First and foremost, a program must be in a phase where the process of prototype development, low rate initial production, or full rate production has at least started – otherwise there is nothing “actual” from which to base actual costs. Second, there must be a reporting process in-place that enables the DoD agency to review accumulated actual costs as the prototype or system is being constructed. Although such a reporting process can vary significantly from program to program, the reporting process typically (a) occurs monthly or quarterly, (b) requires the contractor to provide percent-of-work completed to date and (c) requires the contractor to provide the cumulative cost it has expended for the completed work-to-date.

Note that the other cost estimating methods (analogy, parametric, and engineering) primarily rely on historical cost data. Such historical cost data is commonly described in terms such as “bid data,” “budget data” and “actual data,” to name a few. The “actual data” used for estimates by analogy, parametric, and engineering is NOT synonymous with estimating by actual costs because the “actual data” is NOT internal to the current system being constructed. That is, such “actual data” comes from a different contract or a predecessor of the current contract and does not truly reflect the “latest and greatest” costs of the system currently being built.

The technique of using actual cost data (or extrapolating future costs from actual costs) is based on data from earlier/previous units, prototypes, or production lots of the same system (not a similar system, as in the analogy method). The actual cost method is probably the most accurate cost estimating method when the data is available. The Office of Cost Assessment and Program Evaluation (CAPE) prefers this method since it uses actual or near actual data for the system of interest. The uncertainty associated with this method is based, as with the analogy method, on the technical assessment of the difference between an earlier version of the system, such as a prototype, and the current model under consideration. Obviously, the more the two versions are alike, and the further along the system is in the acquisition process, the more easily an accurate estimate can be made.

A drawback of estimating by actual costs is that it is usually too late to use actual costs to adjust or build a budget. In addition, an estimate of a prototype (actual) typically does not have a 1:1 correlation to the first production unit or lot. Therefore, one may need to repeat the actual cost method for a first production unit or lot.