You are here

Useful life

7 September, 2015 - 14:22

The final issue we will consider is how to plan for a system’s useful life. This is important for two reasons. First, as with any investment, this information is used to determine whether a new system is worthwhile or not. If you have a system that is expected cost USD 5 million and bring in USD 1 million per annum, you know that the system must have a useful life of at least five years.

Second, planning the useful life of a system gives you at least a chance to decide how and when to withdraw or replace the system. Most development projects do not address the issue of decommissioning at all. As a result, systems live for much longer than anyone would have imagined. This is how so many systems were in danger of crashing on the first day of the year 2000 – none of the developers had imagined that their systems would last so long. A perhaps more extreme example is that of the United States Department of Defense, which is reputed to have had more than 2,200 overlapping financial systems at one time 1. Efforts to reduce this number have not been very successful, proving that it is much harder to kill a system than to create one.