Complexity in IT is more complicated than a variable analysis for an experiment. Yet reducing complexity is recommended in industrial engineering to reduce the number of things that can go wrong. Reducing complexity is also part and parcel of lean sigma and the ethos of lean IT. However, before embarking on a project to reduce complexity, knowing how to estimate complexity helps demonstrate the concept and the benefit of lean IT projects.
Proxy Measures for IT Complexity
- Number of software applications used in the organization
A numeric count of how many applications are used by the organization is an indication of the diversity of IT tools. This measure also has the benefit of being easy to measure. The degree of improvement is also simply to calculate.
- Diversity in the software portfolio
A list of software applications provides the numeric count of how many applications are used by the organization. This is called the software portfolio. The diversity of the software portfolio is a comparison of the software portfolios between business divisions, work sites or departments. The greater the difference between each software portfolio, the greater the diversity and the higher the complexity the IT department faces. While engineering or finance may have a few tools that other departments do not, differences in core applications like Operating Systems, email software, financial software and web browsers indicate a dire need for a coordinated consolidation of IT tools.
- Number of Operating Systems used in the organization
The number of Operating Systems used in the organization is a weaker measure of IT diversity than the sheer number of software applications. However, determining the versions of operating systems and the distance between them (how far behind some versions are from others) can reveal a need for tighter configuration management of IT assets. This measure of diversity has the potential to reveal possible problems like security holes from un-patched server software that an application count will not find.