It doesn’t take much to convince IT professionals that it is a good idea to have an applications and system portfolio that is optimised: containing the most suitable technology to support the business and little in the way of duplication and technology legacy. A typical corporate IT systems portfolio will have a wide range of applications ranging from enterprise level behemoths to local ones serving small groups of users; so the question arises: what can be rationalised and optimised, and how do we start the process of pruning and reshaping?
The top concern is, of course, systems that are simply too old, that are well past their ‘sell-by’ dates, with either the hardware or software (or both) being obsolete. In itself age is not the issue, but the fact is that such systems often fail more frequently, take longer to bring back to normal service and rely on a dwindling number of people around with the right skills to feed and water them. Surprisingly often we find systems like these still supporting vital business processes, and with every passing month the likelihood of a serious impact on the business as a result of a substantial failure increases. It is unlikely that such systems can be upgraded easily or cost effectively, so the system’s retirement is usually the best option, with a replacement system procured to perform its functions and services.
The flip side of that same coin is, of course, when a business needs to re-engineer its processes to improve productivity, profitability or customer services, and finds its current technology cannot be reconfigured easily to support the new ways of working. Again, ‘retire and replace’ is likely to be the most suitable strategy, with utmost care taken in developing the business case and implementation plans to achieve the necessary results.
Like tree rings or geological rock strata, applications can build up in layers within a business. At one end of the spectrum organisations that have grown through acquisition and consolidation inevitably build up portfolios of applications that duplicate workflow and services provision. Here, a forensic assessment of the landscape needs to be done, with the support of the business’s leadership and other internal stakeholders, to identify those applications that should be retained (and possibly enhanced) and those that should be divested or retired. The costs associated with duplication can be significant, and such a patchwork of systems is often vulnerable to cyber-attacks, because plugging every gap is such a vast and complex undertaking.
At the other end of the spectrum is the applications portfolio that has grown like mushrooms in a dark cupboard. Where there is a lack of ‘governance light’ all manner of side projects, sub spend-limit procurements, prototypes, experiments, quick fixes and so on may proliferate. The issues facing the organisation will be many: excessive operational costs, security breaches, unmanaged vendors, poor support contracts etc., alongside the lost opportunity costs due to lack of technology standardisation and common working processes. This is always a difficult scenario to face, but as with all good journeys it starts one step at a time: identify some ‘quick wins’, systems that are dormant or with few active users, and begin to decommission them. In parallel start building a roadmap for the key applications and plan out how they need to evolve over time, and what other services may be subsumed into them. Weed out duplicated business processes, shut down non value-adding vendor relationships and put the change management engine into overdrive with internal stakeholders.
The trigger for application and system rationalisation is often not that it is a ‘good thing to do’ financially and operationally, but because other changes are afoot like reaching the end of a hosting contract, implementation of a cloud migration strategy, business divestment etc. But we suggest that, ideally, it should be a core activity of ‘good housekeeping’ that is far easier done on a regular basis, rather than through a big, disruptive effort every few years.