One of the most sophisticated business processes managed by the CIO today is invariably buying or developing something new. Every business has a well-defined process for investment management: the business plan, requirements specification, solution design, option evaluation, option selection, project plan and implementation.
But what about the next 25 to 30 years of living with the asset? Most hardware doesn’t last that long, but where the applications that support critical business are concerned, 25-year life spans are not uncommon. And yet Gartner research consistently finds that IT organisations have practically non-existent business processes for application life-cycle management.
Does this matter? After all, we have had 40 years of industrial IT without being very good at application life-cycle management, so why bother with it now? CIOs must address this issue for one simple reason: a great many applications in the operational portfolio of the organisation are likely to be approaching the end of their useful life at about the same time over the next four to six years.
On the way out
Hardware. There are still some obscure machines around for which support and spare parts are no longer available.
Software. The continued consolidation in the software industry means that many products are reaching the end of their support capability. While the CIO may need to run some lesser systems on unsupported software stacks, such a situation cannot continue indefinitely.
Skills. Many IT organisations are faced with an aging population of technical staff who will not be available to attend to the systems they have been developing for decades.
While most corporate IT teams could cope adequately with some of the application migrations, upgrades, replatforming or retirements that will be required, the cumulative effect over an extended period of time is likely to overwhelm most organisations.
A major proportion of the applications portfolio of large organisations will need substantial work to migrate, upgrade or replace over the next four to six years. Part of the reason for this lies in changing business requirements, but the biggest driver for change lies in the fact that different components of the stack on which the applications depend are becoming obsolete, see box.
The combination of these components creates a major new software challenge for IT management teams: how to plan and budget to refresh the majority of the application portfolio over a comparatively short space of time. There will be very few CIOs who escape this monumental task.
The challenge has many facets – budget, resources, prioritisation of activities and the vast complexity of the planning task. But at the heart, there is an unprecedented architectural and technical challenge to design the application portfolio that is going to be capable of taking the business to 2025 and beyond.
This task is made particularly complex since it happens at the same time that the software industry is going through a revolution in delivery models, with software as a service (SaaS) and business process outsourcing (BPO) undergoing substantial innovative investment and growing rapidly. Any applications architecture looking to create a stable, but agile, platform for a business in the future will be obliged to consider a wide range of application delivery options, even though some may not yet be ready for industrial implementation.
CIOs need to recognise that the focus on application acquisition needs to be matched or exceeded by attention to application life-cycle management. They should develop an inventory of applications, and generate predicted ten-year life-cycle plans for each one.
The whole IT management team should participate in the process, and be prepared to start sharing the results with selected business users.
CIOs should start to adopt application portfolio management as a critical business discipline for the IT management team, and use project portfolio management to guide available investment towards creating an applications portfolio that will support the business through 2025 and beyond. And they must be willing to think about adopting delivery models such as SaaS or BPO to focus available resources on truly differentiating business applications.
Andy Kyte is vice president and fellow at Gartner