It's a punchline right, "Good enough for government work?" Certainly not the attitude at your agency or mine, but "Good enough for government work" has something to teach us when evaluating IT alternatives.

Benefit/Cost Ratio is calculated by dividing the total present value benefits by the total present value costs. Where costs exceed benefits, the ratio will be less than 1 and break-even will not be reached. The same applies for marginal costs and marginal benefits. That is, when comparing alternatives it is sometimes easier to compare the marginal costs and marginal benefits of alternative choices. Where marginal costs exceed marginal benefits, the alternative is not optimal.

Applying this "marginal" cost/benefit analysis can also allow us to clearly see the law of diminishing returns in action. Beyond some point, each additional software feature or unit of hardware capacity (at almost always more cost) yields less and less output. Further, software features and hardware capacity do not necessarily equate to any organizational or societal benefit. In, the book, "Benefits Management: Delivering Value from IS & IT Investments" John L. Ward and Elizabeth Daniel teach us that, "IS and IT vendors are keen to promote the many features of their products and, all too often, organizations believe that the list of features equates to a list of benefits that the systems will provide to their organizations. However, it is seldom the case but can result in organizations buying and installing systems that either do not meet their needs or are overcomplex. As a result those systems tend to be underutilized and hence fail to deliver the expected benefits."

John and Elizabeth suggest that rather than starting with the features and functions of the technology or system, agencies should start with what it is that's causing the organization to consider the investment and what the project is expected to deliver. "It is only when this and the required change management actions have been identified that the IS and IT required should be assessed, leading to a technology specification that is 'sufficient to do the job'." Or, "Good enough for government work."


Post a Comment