Before embarking on any strategic initiative, having a clear understanding of the likely costs and benefits seems obvious. But all too often, this is not the case. The result can be outright failure or limited success at the best. IT seems to be particularly prone to this sort of problem. Why should this be the case?
To help answer this question, I’d like to look at one particular kind of strategic activity: changing the environment in which major applications run. Many large organisations have critical applications running in a well-established environment, often in mainframe-class systems. If the applications do not work properly or are not available, the organisation suffers, as may some of its customers, especially in the public sector. The impact of any change, therefore, has to be considered very carefully, as it may destabilise the operation.
An organisation considering such a change has two possible strategic options: keep the current environment, making incremental improvements, or switch to a different one. The first option includes moving to new hardware and software releases, enhancing automation, and making progressive application extensions such as new interfaces and improved internal structures – using service architecture, for instance. It is a low-risk approach, with minimum impact on the ability to continue delivering the IT services.
Moving to a new environment is more radical. The motivation for change may be outside the control of those involved with the current systems. External directives about approved technologies, and acquisition by or merger with another organisation can be factors driving change. The motivation may be internal, based on a belief, not always rational, that another environment is cheaper or better value, a view often imported with the arrival of new senior management.
There are three choices available for a change to a new operating environment: move the existing applications, write new ones, or move to something else such as packages. Whichever is chosen, the financial implications must be carefully assessed, as wishful thinking about what is possible is dangerous.
Three factors must be included in cost/benefit calculations. First, the total costs of ownership (TCO) of the current and planned environments have to be calculated, usually with the expectation of a reduction. Secondly, the cost and duration of the implementation have to be estimated. Finally, the impact on the business during the implementation of the move must be determined. Calculating the TCO of the current environment is easy because the environment exists and its costs are generally (though not always) well known. Working out the other costs is difficult and frequently underestimated, often grossly so.
The cost of implementing the move and especially the business impact may dwarf any TCO savings, leading to a long payback period. Coping with new requirements during the move is particularly difficult. Do they go in the new environment, or both the new and the current? Failure to make the changes can significantly affect the business, financially and/or legally. Failure to implement new requirements for compliance, for instance, can lead to serious consequences for the organisation concerned.
However, people do move, although with mixed results. I can point to cases where wishful thinking prevailed: a great deal of money was spent to get to something that does not work as well and has a much higher TCO.
A look at moves that are successful reveals a number of common factors. The move should be to working applications meeting all the existing requirements, with little or no change needed. The target applications should scale to the required performance levels. Finally, they should be in production in the same legal and regulatory régime as the current applications so that processes such as auditing and compliance are satisfied. And of course the new environment should show real value.
Whatever strategy is being considered, a detailed analysis of the financial implications of the alternatives is essential. Some of the data may be hard to estimate but getting them is important. And it is essential to take all possible steps to avoid wishful thinking. The last thing anyone wants is to finish up saying ’It seemed like a good idea at the time…’
I’ve discussed IT costs in more detail in a White Paper Understanding IT Costs.