At least £480bn of the government’s operating revenues and £210bn of non-staff expenditure, such as pensions and other entitlements, were reliant on legacy ICT systems in 2011-12, according to a recent National Audit Office (NAO) report.

Entitled, ‘Managing the risks of legacy ICT to public service delivery’, the NAO found that the strategies government bodies had been applying to legacy ICT are unlikely to deliver the wide-reaching levels of transformation that have been laid out in the government’s digital strategy.

The report highlighted that with legacy ICT systems paying out significant sums – the Department of Work and Pensions’ (DWP) system paid out £84.3bn of state pension and associated benefits, and HM Revenue & Customs’ (HMRC) systems administered £99.6bn of VAT receipts (net of repayments) – the potential for failure would endanger the payment of pensions and benefits and the collection of revenues.

The Cabinet Office and the Government Digital Service (GDS) mandated that central departments should work to overhaul legacy systems and implement innovative digital products for citizen use to save taxpayers £1.7bn a year by 2015 – but it’s a challenge replacing these reliable workhorses.

Replacing these core systems would mean recapturing company knowledge and any changes to a single application could jeopardise the data in another application. This carries a lot of operational risk. The overhaul of the NHS systems, where there has been a fall in the accuracy levels achieved when comparing the new system to the old system, is a shining example of the pitfalls involved, though it is by no means the only case where such replacement projects have gone awry.

A lower risk alternative should be considered. One which concentrates on how these mature systems can be updated and evolved to meet new needs, instead of being ripped out. When core system business logic is captured in millions of lines of application code (usually COBOL code) such systems are highly valued business assets and contain valuable business processes. Organisations should look to evolving and modernising these COBOL applications as a safer, more stable option.

Such systems may be seen, rightly or wrongly, as sluggish and costly to maintain, due to inadequate investment, tooling or platforms. However, modernising the underlying technology and processes enables development teams to deliver innovations to these business-critical applications, quickly, in response to market needs. Building on what already provides core business value, while revamping elements such as development frameworks and delivery processes, can yield the required benefits without incurring any of the risks associated with system replacement.

Ripping out older systems because they are considered ‘legacy’ carries a lot of risk. Retaining core business value in existing applications and modernising them to support future needs removes those risks and assures a better return for future innovation projects.