Over the years, Tech Trends has chronicled the progression of once-cutting-edge technologies as they become legacy systems in dire need of a modernization salvo. Just last year, in Tech Trends 2023, we discussed how mainframes can be viewed as “rusty but trusty,” ripe for connecting to emerging technologies through innovative middleware.1 In prior years, we focused on app modernization or cloud migration of aging databases2—in each case, championing a different aspect of the organization’s core tech stack.
This year, we take a step back to broaden the concept of core systems that need to be modernized. Businesses need to deal with aging networks that can’t keep up with 5G and Wifi 6. Their data centers are still being shifted to cloud, but data management needs to be cleaned up for generative AI’s primetime. Enterprise resource planning (ERP) vendors are pushing out new versions that require significant upgrades. Even relatively recent software-as-a-service (SaaS) implementations that were supposed to be a remedy for years of legacy core woes aren’t aging well. On top of all this, companies are dealing with a mix of contractors, captives, and traditional workers who aren’t ready to pivot to modern engineering.
Until now, companies have been monitoring their technical debt—the implied cost of not modernizing systems and working with suboptimal performance—through piecemeal assessments. But those that aim to lead in the future will need a more continuous and complete view of their core. Instead of watching their technical debt accumulate and overwhelm their systems, this view can provide guidance on which technology upgrades matter and when and why to make them. Going forward, the best proxy for understanding your legacy technology realities may not be debt at all, but health.
From a health and wellness mindset, organizations could treat disparate technology systems (cyber, data, infrastructure) like parts of a body, subject to thorough annual checkups, similar to those conducted by integrated medical care providers. Instead of fixing aging systems that break or bottleneck IT’s progress one at a time, teams can use preventative health assessments to identify and prioritize the areas of the tech stack that need treatment. These assessments can be rooted in real business problems: increasing costs and risks and stifled innovation and growth. For example, some aspects of core systems, such as the mainframes we discussed last year, may be in good health, only needing connectors to keep doing what they do best. Others may be due for a thorough upgrade or replacement.
Moving away from previously siloed modernization efforts, businesses are likely to have a highly customized and integrated wellness plan across their tech stack in the coming years. After all, today’s white-hot innovations will likely continue to become tomorrow’s legacy systems, in need of checkups, especially at the current pace of technology innovation.
Whether they are ERP systems or data centers, technologies that once revolutionized business are now prone to slowing it down. Up to 70% of technology leaders view technical debt as a hindrance to their organization’s ability to innovate and the No. 1 cause of productivity loss.3 Perhaps the population that suffers from this most directly is software developers, who spend an estimated 33% of their time dealing with technical debt maintenance.4 This time spent can also have an outsized impact on the productivity and satisfaction of developers, as discussed in this year’s trend on developer experience. As many as 78% of developers felt that spending too much time on legacy systems had a negative impact on morale; other impacts cited were employee and customer churn along with lost deals.5
As technology’s rapid advancement increases (as evidenced by generative AI), businesses and government agencies tend to struggle with two competing truths. On one hand, they deeply believe their future business models, existing products and services, and internal operations will be fundamentally transformed (or disrupted) by technology. On the other hand, they often struggle to make the necessary investments in their infrastructure, data, applications, cyber, and workforce capabilities in a way that could adapt to that future. This often leads to technical debt sprawl: In 2022, the estimated cost of technical debt in just the United States had grown to US$1.5 trillion, despite chief information officers spending 10% to 20% of their budgets on resolving issues related to outdated systems.6
Many companies have experienced subpar transformation programs that amounted to massive bets on a single dimension of their core systems, which ultimately failed to deliver the benefits promised. Instead of undertaking random acts of innovation or “low-hanging” transformation investments, technology leaders may need to face a hard truth: Their technology house is ailing. And they need new ideas for where to focus time and effort so they can begin to heal.
To move forward, leaders can look for a smarter way to invest in modernization: a systematic assessment of their company’s needs, strengths, and budget across the key areas of modernization spending. A holistic view of their organization’s systems, grounded in real business context, can help them eschew ever-growing technical debt for a more long-term view of technical health, one that improves over time and provides more confidence to business and technology teams alike.
Because organizations are unclear which sources of technical debt are causing the most drag, issues are often underprioritized or poorly managed.7 In reality, out of hundreds of applications or systems in a company’s core, just a dozen applications that have a handful of issues may be driving the bulk of the impact of outdated systems. Instead of applying funding every year that seems to go nowhere, companies could benefit from taking a step back to tackle key issues.
The investment drivers of a more holistic core modernization strategy span a variety of barriers, costs, and potential risks that companies may face by keeping legacy technology in place:
A core modernization framework built around technical wellness can be more comprehensive than the traditional framing of debt. In such a framework, the focus is placed on preventive care—in this case, using enhanced tracking, measurement, and predictions to address suboptimal legacy technologies before they become larger issues for the business. Rather than accumulating debt and paying it off periodically with large but perhaps ineffective investments, a wellness framing would encourage businesses to iteratively pinpoint their technical ailments and predict when investments would be most effective, based on cost, operational risk, and innovation readiness.
The focus of such a wellness diagnostic, or core checkup, would be on the five areas of statistically largest spend and biggest opportunity. Each has a current target for modernization, which is likely to change over time as emerging technologies either become more sophisticated or new ones become popular.
Infrastructure is the broadest category and often the most difficult and expensive area. Fortunately, as evidenced by the State of Utah, entire mainframe systems can be migrated to the cloud within 18 months if an organization is aligned on its transformation goals.8 Within this category, technical wellness translates to mainframes, servers, and end-user devices (such as virtual desktops) being migrated to the cloud across technical environments (sandbox, quality assurance, production). In addition, aging fiber, LAN, and WAN networks across facilities (such as data centers and corporate offices) are on a journey to modernize to 5G, Wi-Fi 6+, low-energy Bluetooth, and satellite communications. These upgrades enable companies to take advantage of private networks, advances in software-defined networking, and other advanced connectivity offerings.
Data life cycles (including cleansing, manipulation, and management) and data storage constitute this category. Businesses need to streamline data cleansing and manipulation through automation so they can spend less time on data management and more time on analyzing insights. Reporting on data usage and cleanliness, especially from a trust perspective, is also key, since AI models are only as good as the data they ingest.
Storage across data centers, offices, and remote assets can be modernized to cloud storage systems and data can even be streamed in real time. Amazon Web Services recently established streaming data pipelines that move data from connected devices across multiple sources to centralized repositories where it can be better leveraged, ensuring that users are working with the freshest information possible.9 This data can then be used in applications such as predictive maintenance, environmental monitoring, and smart city management. This application opportunity rests on a vast landscape of data stores, lakes, spindles, and drives—each carrying complexity and very real costs.
“Anything outside of using real-time data becomes very frustrating for the end consumer and feels unnatural now,” says Mindy Ferguson, vice president, messaging and streaming, at Amazon Web Services. “Having real-time data always available is becoming an expectation for customers. It’s the world we’re living in.”10
This broad category includes legacy custom applications that organizations have been modernizing over time through one or more of the “five Rs”: replatforming, revitalizing, remediating, replacing, and retrenching. It also includes package applications, such as ERP and SaaS applications, that require a clear upgrade strategy as vendors continue to improve their offerings, while dealing with the inevitable litany of customizations that complicate upgrade paths and integration.
Operations technology applications and product technology stacks, such as embedded products and digital offerings for customers, would also be considered for checkups in this category of modernization.
Many companies are struggling with a workforce that is a mix of internal and third-party contractors who may or may not be ready to bring about the modernization efforts described above. To improve talent acquisition and retention of tech teams, leaders need to prioritize the modern engineering experience, bolstered by investments in tooling (across the software development life cycle), processes, and culture, as discussed in our trend “From DevOps to DevEx.”
Finally, companies need to consider their relative health in cybersecurity across multiple areas: security and privacy, regulatory compliance, and ethics and morality. The first two areas can be tracked and improved through cybersecurity automation, especially to keep up with the growing amount of artificially generated content, as discussed further in our trend “Defending reality.” Ethics requires a more nuanced approach, and businesses should keep up to date on the latest thinking on technology’s potential harms to society.11
The benefits of a core wellness checkup across these five areas would be financial as well as intangible. For example, leaders who actively manage and reduce technical debt are expected to achieve at least 50% faster service delivery to the business.12 And the time returned to developers could result in many more features being developed to generate revenue from customers or efficiency from employees. Perhaps, most of all, an accurate tracking system for technical debt could allow organizations some peace of mind in knowing when and how to prioritize their investments instead of scrambling to keep up with the market.
As modernization needs progress over the next decade, what if technology could become adaptive and resilient, able to “heal” its own outdated code or system without limited intervention?
The idea of self-healing systems is not new. In the natural world, whether at the micro level (such as a broken bone healing itself) or a macro level (for example, an entire ecosystem rebuilding itself after a forest fire), nature has shown us the pinnacle of resilient design. It’s no surprise, then, that the field of biomimetics—design inspired by nature—has received more attention in recent years, and the applications within technology have already begun.13
For instance, self-healing raw materials such as ion gels have used clotting properties to heal damaged robot pieces, such as arms and hands, when the robot senses a cut in its material.14 This same process is being replicated with electrical circuits as well: When an electrical circuit is damaged, a capsule of liquid metal can be released automatically into the circuit to repair the electric connection.15
Crucially, self-healing systems are slowly graduating out of the world of atoms and into the world of bits. Consider the example of adaptive AI, which has progressed from human-initiated machine learning to unsupervised machine learning.16 This AI can not only solve challenges but also, in studying those challenges, learns to teach and reprogram itself by developing more advanced problems.
Along these lines, core modernization solutions are also poised to become adaptive. AI embedded into core systems can currently diagnose tech debt accumulation in tech stacks and support engineers as they write the necessary code to modernize (while also streamlining remedial tasks and compliance, which often pile up when tech debt increases).17 In fact, a recent global survey by Deloitte indicated that around 60% of organizations are already using AI to optimize code and identify bugs, while 50% are using it to manage code environments.18
Similar to a physician in training, these AI solutions for tech wellness are still prone to error and misdiagnoses; for example, they may be less effective at refactoring than debugging,19 but as they spend more time in their residency of core systems, they’re bound to improve. One day, AI could diagnose inefficiencies, develop a solution, and implement the solution without ever needing the support of human engineers.
As such innovations continue, longevity could be built into the five areas of core modernization from the onset. As more of the technology stack becomes software-defined, efforts to embed fault expectancy, monitoring, and self-healing could improve the “aging process” of our technology assets.20 As with human wellness, the goal of tech wellness would be for core systems to age gracefully, with built-in supports and checkups to allow them to fulfill their purpose.