In most companies, the computer services department is strictly responsible for its own budget, which comprises a small percentage of total company expenses. In these instances, it is usually sufficient to periodically track the department’s actual costs against its budget and halt any further financial analysis. However, there are cases in which the computer services department is responsible for very large projects that either have a major impact on available company cash or that must be completed on time in order to give the company a competitive advantage through some technological innovation. The analysis noted in this post can be used in this situation.


The computer services department is, from time to time, responsible for a few very large system implementation projects. It is critical to track the success of these projects in order to ensure that they are completed on time and within the prescribed budget. If not, a company can suffer a severe shortage of cash due to the allocation of extra funds to such a project, or will not have the benefit of the new system until much later than expected.

The measure to use when tracking such a situation is a continuing comparison of budgeted to actual costs for each major computing project. It is not sufficient to review the current actual cost of an ongoing project to its eventual total cost, because this approach gives the analyst no information about whether the project will be completed on budget.

A better method is shown on below exhibit, in which the budget to actual comparison is conducted for each project on a milestone basis. By breaking down the analysis into smaller pieces, it is easier to determine in advance if a project will probably run over budget.

Comparison of Budget to Actual Costs by Milestone

Milestone         Budget       Actual      Total Variance
Storyboards     $175,000  $135,000  -$40,000
Initial coding     572,000    600,000  -12,000
Testing              305,000   300,000   -17,000
Procedures         58,000     65,000     -10,000
Installation       129,000   150,000       11,000


In the example, the project is cumulatively under budget at all milestones until it reaches the installation phase, where it finally incurs a loss relative to the budget. Also, losses are incurred at the initial coding, procedure writing, and installation phases, so all of these work areas should be subject to a review to see if the problems can be avoided for the next computer project.

For those companies in which it is critical that projects be completed on time, a similar analysis based on time, instead of money, can be easily developed. Like the analysis in the above exhibit, it splits a project into milestones and assigns a completion day to each one. By tracking the cumulative variance at each milestone, one can easily determine where variances are arising and roughly when the project should be completed. This analysis is shown on the next exhibit. For both types of analysis, the accuracy of the initial budgeted figures will increase as the computer services staff gains experience in compiling initial budgets, so it is important to go over the final results with them at the conclusion of each project, so that they will receive feedback on their previous estimates.\


Comparison of Budget to Actual Times by Milestone

Milestone         Budget (days)       Actual (days)      Total Variance (days)
Storyboards     32                        29                       -3
Initial coding   110                      115                    +2
Testing            45                        40                       -3
Procedures      15                        20                      +2
Installation      30                        35                      +7


Besides the tracking of project costs and completion dates, some controllers may feel that the overall costs of the computer services department as a proportion of total company expenses are too high. One way to determine the reasonableness of this cost is to calculate the total computer services cost as a percentage of total revenues, especially on a trend line that reveals any significant changes over time. However, this calculation may yield inaccurate results if a company uses sophisticated computer systems as a linchpin of its overall competitive strategy, which would justify a higher expense level. If computer expenses are treated as a commodity, however, a reasonable way to determine if the cost is excessive is to calculate the com puter cost per user and compare this figure to that of other organizations.


The calculation is:

(Total cost of all computer hardware) + (Computer depreciation expense) + (Computer outsourcing costs) + (Allocated personnel and occupancy costs) / Number of employees using a computer

Note: It is very important not to use the total number of company employees in this calculation, because there may be a significant number who do not use computers, such as those on a production assembly line. Including these people in the measurement would significantly skew the cost per user downward. Also, the computer depreciation expense should accurately reflect the actual number of years over which this equipment is used—internal accounting rules may result in depreciation periods that are too long and therefore spread costs over an excessively long period.

This post has focused on only a few key aspects of the computer services department, with a particular emphasis on its ability to efficiently and effectively complete hardware and software creation and installation projects in a timely manner. These are crucial factors in those companies that rely on advanced computer technologies to ensure their competitive survival. The controllers of these organizations should strongly consider the use of the measurements noted in this section to keep close track of computer services activities. Some of these measures can also be duplicated for the engineering department, the subject of the next section, which operates in a similar manner.