Application Retirement: What about the Data?

by Jon Collins and Martin Atherton

Introducing application retirement: what’s really important to you?

Whatever can be said about IT in today’s organisations, our research shows us that we must start
from the premise that ‘it’s not wrong’. Plentiful reasons exist why infrastructure and applications are
as they are: some of them historical, because it made sense at the time, or organisational in terms
of who had the budget, or indeed because yesterday’s innovations can so quickly become today’s
legacy.

This is particularly relevant in how applications, and the repositories upon which they depend,
deliver services to business users. A senior business manager once said to us:

“What’s wrong with silos? At least they work. I would rather have off the shelf proprietary systems
and employ four or five extra people, than spend millions on a system, and the day after its
implemented, somebody comes along saying, “Can it do this?” and you end up with 3 or 4
programmers continually working on the system and modifying it.”

Against this background it is unlikely we will ever arrive at what could be considered as ‘application
nirvana’ – that is, the perfect software layer which supports the business’ every requirement without
ever needing to be changed. As organisations evolve and new capabilities emerge, from online
banking to mobile access, so do the software and hardware platforms upon which they depend. And
in turn, these impose new challenges on the data.

The downside is that IT has to deal with the consequences of constant change. We talk about such
challenges as fragmentation, inefficient operations, data security and so on, which often result from
where IT systems have not kept up with the changing requirements of the business. Change events
can be difficult to pre-empt or indeed predict – one strategic business acquisition or change in
market dynamics, such as a merger or indeed (as we have seen most recently) the credit crunch,
can result in even recently deployed applications becoming more of a burden than a help.

The insights upon which this paper is based are taken from a number of research projects
designed, executed and interpreted independently by Freeform Dynamics. This paper has
been authored independently by Freeform Dynamics and is sponsored by IBM.