More Case Studies / White Papers
Contributed by
David Mullens
ADAPT
In the earliest days of computing the possibilities for using computers to automate processes, render them speedier and hopefully more accurate (see GIGO) led to the first development of commercial computing, with the construction of LEO 1 – Lyons Electronic Office, to process the payroll of the Lyon's organisation.
The LEO 1 was the non plus ultra example of the ADAPT model – whereby the whole solution was produced from the bottom up, including even the manufacturing of the hardware for the one specific application.
No one seems to have performed an analysis of the cost effectiveness of the solution, but it would surprise me very greatly if the benefits of the computerised processing of the Lyons payroll were greater than the very considerable cost of purchase and subsequent maintenance of the hardware and software.
ADOPT
Though we should be grateful to those who created LEO 1, it perhaps provides an extreme example of some of the issues that may occur with bespoke development versus the adoption of a standard software solution.
Although the LEO analogy is certainly imperfect, the costs of bespoke software development, testing and implementation are huge, with the consequence, it is generally reckoned, that on average about 80% of IT Development costs go into maintenance whereas just 20% goes into new developments – i.e. developments to respond to new market / business needs.
Thus is born ADOPT – an approach that says: take all you can from what already exists on the market and change it as little as possible – as every change you make will cost money – not just short term but long term too. In doing this the compromise is that you have to change your way of working – your business processes instead, but the advantages are many. In particular, by being as standard as possible you are easily able to take the software provider's updates quickly and easily, not only gaining business advantage by early adoption, but also minimising costs as the supplier who can spread his costs over many customers will almost certainly be more economical.
Of course this may not always be the best way, if your IT provides you with a concrete market / business advantage you may well prefer the higher costs of in house development, but it is well worth assessing the direct benefit of any development - whether in house or developed by a third party, looking at the projected lifetime costs and understanding the concrete financial benefits this development will achieve. It is likely that in 95% of cases there will be a packaged solution providing 80% + of your business functionality needs, to which you can adapt your processes or maybe (if really justified) add bespoke development to satisfy the most pressing.
BEWARE
Even in 2011 there are still large companies that live on producing bespoke systems that lock you in to large scale systems development at very high cost and doubtful value, this is often simply because they have large numbers of programmers needing work to keep them off the street. An old manager of mine – running a $2bn budget – used to say that if the project was due to take more than one year to implement, it should be killed immediately, by the time it starts to deliver results your business needs will certainly have changed. Broadly speaking complex systems very nearly always fail in one way or another – usually cost and time.
Evolution beats Revolution – a lesson Central Government nearly always fails to learn.