Mainframe decision time
Mainframe decision time
Critical countdown for banks.
Shrinking pool of experts leads to tough decisions for banks
The need for modernization has been accelerated by the growing shortage of legacy talent. Legacy business and technology experts are retiring in droves, leaving a dangerous vacuum of technical and institutional knowledge. As digital innovators continue to disrupt the market leaving legacy technology behind.
Until recently the risk of moving from an inefficient mainframe to something else scared decision-makers in financial organizations. Migrating to another platform or the cloud risked the kinds of issues that generate headlines and firings. Banks peeled away some processes here and there, even as they kept relying on antiquated systems. The strategy has worked until now.
Banking leaders now see keeping mainframes as even riskier than doing nothing. Younger generations have zero interest in COBOL and the other programming languages used in mainframes. Actually, it’s not negative that they waited. That’s because modernization options are better (and lower risk) than ever. Upgrading infrastructure, while keeping costs under control, requires financial organizations to find the right path to modernization.
Time for a heart transplant?
The mainframe is the heart (and lungs) of many, if not most, financial services providers. The exception would be fintechs and non-financial brands dabbling in financial services. Born at the dawn of the computer age, mainframes enabled better, faster financial interactions. Checks (yes, those paper things) cleared in days, not weeks. When you called customer service, agents answered many questions in minutes.
Even with antiquated mainframes, banks settle many transactions in a single business day. That’s no longer good enough. Modern systems support real-time processing, AI-powered personalization, and next-generation digital experiences. New entrants to financial services aren’t constrained by mainframes. Just as crucially, the people who know how to coax the best performance from mainframes are retiring. Their GenX peers prefer to work with more modern systems. Millennials and GenZ? They view working with 30-year-old systems as a career dead end. Soon, organizations will lack enough skilled people to manage mainframes.
"Financial institutions have come to the realization that legacy mainframe databases are ill-equipped to meet the demands of digital banking."
Consider your modernization options
With delays no longer advisable, many banking leaders are taking a fresh look at their mainframe modernization options, which are:
- Rehosting: With rehosting, you migrate your current systems – with very few modifications – from mainframes to more modern infrastructure. This option reduces maintenance costs and, in a limited way, improves performance. Think of it as driving an old car on a newly paved express road. It’s a smoother ride, but you use the same antiquated technology. You need many of the same (retiring) skillsets to make rehosted systems “sing” that you do to run mainframes, so this option is a temporary fix. Many banks that have completed successful rehosting are now looking at the following two options.
- Replatforming: A popular option, replatforming means moving from mainframes to an entirely new platform, usually the public cloud. The critical difference from rehosting is that you modify key aspects of your legacy systems to take advantage of the elasticity and resilience of cloud infrastructure. Compared to rehosting, replatforming delivers more cost and performance benefits. Minimal code modification helps reduce perceived risk, and running in a modern environment creates an effective bridge to the future. Retiring skillset issues will resurface for some projects, but you’ll experience significant cost and agility benefits.
- Refactoring: With refactoring, you modernize your application code as you migrate from mainframes to the public cloud. Here, you gain the agility, speed, and cost-efficiency your fintech competitors enjoy. Finding people to maintain and manage your systems is much easier. You can start to look at improvements such as real-time processing. So why not take this path to true mainframe modernization? Refactoring at scale requires some degree of automation, and many refactoring applications produce “frankencode.” You move away from an antiquated COBOL codebase, but in some cases, the output falls short of the needs of financial service providers. Yet refactoring offers the most comprehensive solution to the retirement problem.
Choose the right path to mainframe modernization
Financial institutions have come to the realization that legacy mainframe databases are ill-equipped to meet the demands of digital banking. At Amdocs*, we see replatforming and refactoring as viable paths to modernization. However, in our opinion, more financial service providers should explore refactoring at scale. With effective refactoring, you liberate your legacy applications, making it easier to surpass the competition with customer experiences that include AI and personalization. Refactoring also transforms your ability to recruit and retain IT talent.
At Amdocs, we can typically provide a refactored sample of your code in about 24 hours. Our teams can discuss the pros and cons of replatforming and refactoring thanks to our experience with both paths.
What about the risks? Tackle those proactively by avoiding frankencode and approaches lacking large-scale financial service successes. Instead of settling for ready-made code examples or investing in a time-consuming proof of concept exercise, ask any potential partners to refactor some of your code as a precondition for more in-depth discussions. We can also share examples of large-scale modernization at leading financial service providers. Equally important, look at how companies such as Foyer Insurance , Société Générale or Deutsche Bank addressed all these challenging by successfully modernizing their infrastructure.
*Astadia joined Amdocs Cloud division in November 2023.
