Managers
choosing the build option often take a very short term view of software
costs. It is not unusual for most to believe that initial design
and programming make up the majority of those costs.
In
Perspectives on Software Engineering, Marvin Zelkowitz from the
Institute for Computer Sciences and Technology, National Bureau of
Standards, Washington, D.C reports on extensive research findings on the
costs associated with custom development. He identifies several
categories of cost. Only 5 % of the costs are in the initial design,
and only 7 % in the initial program coding. On the other hand,
15 % of the
overall cost is spent on testing, and a significant 67 % of the costs
occur after the testing cycle is completed.
Mark Lutchen,
former CIO of PricewaterhouseCoopers, makes this point as well. He
indicates that when evaluating whether to buy or build, it’s critical to
thoroughly understand total costs during the software lifecycle. This
step is important, says Lutchen, because 70 percent of software costs
occur after implementation. (InfoWorld, February, 2007).
Most managers are
inclined to believe that once a software application has been built and
delivered they can begin seeing a return on their investment and costs should
go down. In fact, less than one third of the time spent and costs
incurred have been completed on the day the application has finished
the testing cycle.
Unfortunately,
this means that outside consultants or internal IT departments can
sometimes provide software delivery at what
appears to be a low cost, while it is virtually certain the majority of costs
in time and money will just be starting.
|
NOTE: We are not
without a bias in the Buy vs. Build decision. However, this
article includes important, objective
references on the decision making process.
For anyone considering building a case management system, or
modifying an existing system that was built from scratch, we strongly
recommend you read the article in detail and review the references
provided.
|