Computer Futures

Stick or twist: when to modernise the legacy system

For decades, the modernisation of legacy systems has proved one of IT’s more intractable challenges – not least because it’s an area where even minor tech change can have a major impact on revenue streams.

With this previous generation tech (defined as still operationally critical), timing is always a factor. Most organisations want to migrate their legacy applications to newer platforms, but the potential disruption to core business operations means many stick with their ageing systems for a bit longer – and then a bit longer.

Modernisation is a decision that arguably pivots too often on how to rather than when to. Legacy’s longevity has shown a reluctance to replace it hasn’t brought about the demise of the banks or governments relying on it.  

But few would seriously propose unmodernised legacy systems can keep running indefinitely. They have endured due to business, rather than tech, imperatives. So the new market dynamics that are causing fundamental changes to the conduct of business will set the timeline for legacy modernisation.

Long-lasting legacy

Legacy systems persist in the big vertical sectors, such as banking, retail and government, that were early adopters of computer technology in the 1970s and 1980s. A 2021 review by the UK’s Financial Conduct Authority found that over 90% of financial sector firms are reliant on legacy technology to deliver their services. This proportion is highest in insurance, where almost three quarters of respondents classified most of their infrastructure and applications as ‘legacy’, with 100% of insurance sector respondents also reliant on legacy tech in some form. 

Financial sector incumbents must compete with agile, fully digitalised cloud based fintechs; their IT resources weighed down by technology that lacks flexibility and openness. In any vertical, this is compelling evidence it’s time for change.

The legacy overhead has also greatly expanded since the millennium. Twenty years ago, legacy systems would have been defined largely in terms of mainframe and mid range hardware from the 1970s and 1980s. Today, ‘legacy’ could apply to any still operational enterprise tech that has been deployed since 2000 – such as early Java and .NET applications – though with the commoditisation of server hardware, it mostly means software.

A well-known instance of ‘new’ legacy is Microsoft’s Windows XP operating system, first released in 2001. A 2019 survey by Spiceworks found that 32% of businesses polled still have Windows XP installed on at least one networked device, despite the operating system being unsupported by Microsoft since 2014. Spiceworks’ data also showed 79% of businesses were running Windows 7, which Microsoft ceased to support in January 2020.

If outdated versions of Windows continue to meet operational requirements, it’s understandable if organisations aren’t inclined to upgrade to the latest version. But without updates and support, the software isn’t being securely patched against emergent cyber threats. As the financial, regulatory and reputational repercussions of a successful hack-attack ramp up, arguments in favour of legacy modernisation grow apace.

Legacy tech skills shortages

Additional challenges related to unmodernised legacy tech are around running costs, skills gaps and operational agility. Legacy systems have a reputation for being high maintenance, and their OPEX (operating expenses) requirement is a regular argument for modernisation. For instance, Cloudsoft believes newer tech will consume less power (and generate less heat) than a legacy system to do the same amount of work.

Legacy systems’ running costs can also grow exponentially as they get older, and therefore start to claim a bigger proportion of already over-stretched IT budgets. Some sectors (retail is an often-cited example) are reported to spend more than half of their IT budgets on maintaining legacy systems. And while many new IT systems can be configured to run without human management, legacy systems are less open to automation.

Support contracts are another push point that can prompt companies to take the plunge and modernise their legacy platforms and applications. Allon Mureinik, Senior Software Engineering Manager at Synopsys Software Integrity Group, says: “Either some critical component in the legacy system reaches an end-of-life status, or the renewal of an existing support contract becomes abhorrently expensive.”

Another timing issue relates to the people who maintain and/or operate a legacy system, Mureinik explains: “Often, these systems are maintained by a small group of ‘veteran’ techies, and if one or more of them leaves or retires, the legacy support team becomes so small that it’s no longer an acceptable risk to rely on it.”

Younger replacements are difficult to recruit. Early career STEM candidates look for opportunities to gain expertise in the latest market leading solutions – acquiring skills and experience that enhances their value to employers. They’re less keen to work with tech that’s seen as a blast from the past.

Another prompt for legacy modernisation is incoming legislation. IT operations have, in recent years, become more subject to state and industry regulation. For example, enterprise governance chiefs have to ensure their IT provisions comply with data protection regulations such as GDPR (General Data Protection Regulation) in Europe, and CCPA (California Consumer Privacy Act) in the US. When Cap Gemini asked IT executives to rate the top challenges organisations face while preparing for the CCPA, for 42%, legacy IT emerged as critical. And 56% of respondents to a 2021 survey by Rackspace Technology reported that delaying application modernisations had resulted, at some point, in their failure to meet new regulations.

Digitalisation: a new change impetus?

These and other issues are brought into sharp focus by the global drive for enterprise digital transformation. Direct digital transformation investment is expected to reach $7.1 trillion by 2023, according to IDC research, as organisations strive to become ‘digital-at-scale’ enterprises.

And as they prepare to grasp the legacy modernisation nettle, this mega trend will cause many organisations to think about timing: when is the most favourable point to initiate the modernisation process?

“Digital transformation puts the dominant focus onto the new platform and exciting new tech,” says Lorna Rea, Central Government Consultant at BAE Systems Digital Intelligence. “Typically, this can result in a business leveraging guidance from design teams in isolation from legacy support teams.”

This is a common pitfall, Rea warns. Key factors – understanding how a legacy system is used, parallel running requirements, decommissioning considerations, data management and transfer concerns – are missed.

“Both future design and current/past legacy tech team perspectives must be incorporated for a successful, timely transition from legacy to a fully digitalised system,” Rea argues.

Digital transformation brings another new impetus as it requires those traditional hold-back risks and costs of legacy modernisation to be reassessed against a higher scale of ultimate benefits. And possibly, at long last, overcome.

 


For help recruiting tech professionals for legacy support teams or future design and operating needs, contact us today!