Conventions and Starting Assumptions

I'll begin by making a number of starting assumptions and noting some conventions that will be used throughout the ROI model. The model is implemented as a spreadsheet. Formulas will not be shown: most are straightforward and it is hopefully obvious how the calculations are being done; showing them would add clutter and confusion.

There are a number of parts of the model which are highly subjective and/or will vary by company, industry, and so on. These parts of the model are marked with gray cells. The end of this chapter includes suggestions for dealing with this uncertainty.

Assumptions About Cost of a Fully Burdened Employee

We need to begin the ROI model by calculating the cost of a fully burdened employee per work day, per hour, and per minute as shown in Figure 7.1.^[2] By "fully burdened" we simply mean the total cost to the company: salary + total benefits. As a rule of thumb, the fully burdened cost of an employee (i.e., salary plus health insurance, vacation, holidays, sick leave, taxes paid by employer, and so on) is usually about 1.5 times salary.

^[2] All dollar amounts in this model, both for costs and savings, are shown in US currency.

Figure 7.1. Cost of fully burdened employee by day, hour and minute.

Initial Actual Data about Use Cases

Next we capture some actuals in terms of numbers of use cases that were entered into the requirements management tool as of the date of the ROI assessment; in this case, about 1.5 years after initial rollout (see Figure 7.2). Of the approximately 5,000 use cases entered, about 1,000 had already been implemented and shipped as part of some project and some 200 had been rejected, meaning that a decision had been made that these use cases would never be addressed in any future release.

Figure 7.2. Use cases entered, implemented, rejected, and remaining open.