Outdated Software in Focus

We are facing a situation in which the shortage of skilled workers hits old technologies particularly hard. COBOL, VB6, and Delphi are just a few examples of programming languages where expert knowledge is fading away while maintenance and support efforts are rising. Many of our mid-market clients report how difficult it has become to find even individual developers for such systems. After all, a large share of these experts will retire in the coming years, and young talent can hardly be motivated to work in walled-off legacy stacks.

This trend is also reflected in manufacturing: in mid-2023, more than 700,000 industrial positions remained unfilled in the United States, which according to figures from Seraph is partly due to the aging workforce and the growing skills gap in both classic and new technologies [1]. The situation is worsening in Germany too, especially in areas where older hardware and software are still in use. This accelerates the departure of experienced professionals who do know these old systems but are approaching retirement.

We see a serious problem here: companies that rely on COBOL, VB6, or Delphi cannot simply shut down their legacy systems overnight. In many cases, business-critical workflows, order-handling logic, or booking routines that have proven themselves over years are buried in these systems. The same goes for other outdated software landscapes that are highly coupled and difficult to integrate into modern IT architectures.

At the same time, we are convinced that companies should actively engage with these legacy solutions before the last specialists retire. Beyond the dependency on rare expertise, these systems lead to high maintenance costs and growing technical debt. The question of modernization strategies therefore moves more strongly into focus: should we modernize legacy systems in the long term, or is it more sensible to replace the legacy system entirely?

Not every business has found seamless ways to bring new talent into the company or to train existing employees in older technologies. We therefore advocate a proactive course. At its core, the point is to mitigate the risks before the developer shortage threatens operational reliability. Especially in industries where legacy systems still play a major role, we have to preserve the remaining know-how and modernize at the same time.

Why the Skills Shortage Hits Old Technologies

The shortage of skilled workers in old technologies is not an isolated phenomenon. Rather, it is the interplay of demographic developments, technological dynamics, and changing career aspirations. According to Fraunhofer IGD, the number of working people in Germany will fall by around 3.5 million by 2035, which exacerbates an overall deficit of qualified workers [2]. Young developers often want to work in forward-looking fields such as AI, cloud computing, or container orchestration. COBOL-based insurance portals or Delphi front ends for warehouse management appear rather unattractive to them.

On top of this, many vocational and degree programs focus on modern technologies. Classic languages like COBOL appear less and less often on the curriculum. Specialized knowledge that is indispensable for the continued operation of existing solutions is therefore being lost. Many software landscapes from the 1970s through the 1990s are still running today in banks, insurance companies, or industrial environments — all with their own proprietary dialects and quirks that are barely documented.

Our experience suggests that the gap between old and new technologies smolders because even the few experts who still master COBOL & co. can hardly be motivated for repetitive maintenance tasks. In addition, companies are often unwilling to free up large budgets purely for maintaining aged systems. When key people then retire, valuable code knowledge remains unsecured.

A further obstacle is the stigma that clings to old technologies. According to Seraph, younger IT professionals avoid jobs associated with physically demanding or "outdated, boring" activities [1]. Programming itself is not physically demanding, of course, but the association with dusty systems puts off many talents. Companies thus lack suitable candidates to staff project teams for legacy applications.

Challenges for COBOL, VB6, and Delphi

COBOL, VB6 (Visual Basic 6), and Delphi are prime examples of legacy languages still in use in many German companies. These technologies provide an essential basis for back-office processes, ERP solutions, or industry-specific applications. However, several factors fuel the skills shortage and deter potential developers:

1. Lack of training: Universities and continuing education institutions now focus heavily on modern frameworks, cloud technologies, and agile methods. Delphi or VB6 often no longer appear in the curriculum at all.

2. Limited community support: Compared with current technologies such as .NET Core or JavaScript frameworks, the community around VB6 and Delphi has shrunk considerably. The same applies to COBOL — help forums and documentation are outdated or hard to find.

3. Limited career outlook: Anyone specializing in COBOL today is betting on skills that will not expand going forward. Salaries and advancement opportunities often appear less attractive.

4. High technological complexity: Grown-up Delphi applications or monolithic COBOL programs are only fully understood by a few senior developers who have spent years getting to know all the system's interconnections.

We frequently observe that companies with these legacy languages are stuck in a dilemma: they urgently need maintenance and further development, but at the same time many IT professionals see these technologies as a dead end. As a result, the issue of skills shortage in old technologies is escalating: the existing systems do still run, but they are barely covered in terms of personnel.

Risks for Maintenance and Security

Old software brings a range of risks when expertise about it dwindles. On the one hand, the likelihood of failures increases as soon as no one knows where in COBOL routines a critical booking logic is implemented. On the other hand, security gaps grow because outdated development environments are no longer regularly patched, or because manufacturers have ended support. We observe that companies then run the risk of breaching compliance requirements — especially in banking or the public sector.

The shortage of skilled workers for older technologies also means rising costs for emergency solutions. If a senior developer is suddenly unavailable, no one can responsibly take over the problem. Companies often have to bring in expensive freelancers who can still work with Delphi control desks or VB6 front ends. Such specialists are rare and command correspondingly high day rates. At the same time, maintenance costs rise because all the workarounds that have piled up over decades are difficult to disentangle. A hidden technical debt arises here that limits productivity and scalability in the long term — companies effectively pay twice, because there is no quick fix.

In addition, outdated technologies impair innovation capability. When processes are constantly slowed down by manual interventions or error-prone parts of the system, less room remains for new projects and digital business models. While competitors develop their apps in modern environments and adapt them flexibly, COBOL, VB6, or Delphi often lack simple interfaces or automation options.

A recent example we encountered was a mechanical engineering company in Baden-Württemberg that had to postpone important orders because the programming of an old Delphi application could not be updated in time [3]. Only a handful of employees could still adjust the code, but they were leaving the company or retiring. Without prepared succession plans, significant delays arose.

Strategies for Modernization

In view of these risks, we recommend developing viable modernization strategies in good time. There are different approaches — for example, gradually rewriting individual modules of a COBOL application or providing a modern replacement for an old VB6 front-end system. In principle, it is worth setting the course early toward a future-proof architecture.

Depending on the status quo, different strategies come into play:

• Incremental migration: Here, individual parts of the legacy systems are renewed bit by bit and transferred to more current technology stacks. We have had good experiences with the strangler fig pattern migration, in which old modules are gradually replaced without an abrupt complete switch.
• Big-bang redevelopment: For heavily fragmented systems, a freshly built framework can be sensible, although this carries considerable risks. We recommend weighing up the big bang vs. incremental migration question in advance.
• Modular refactoring: If COBOL logic is still partly fresh, refactoring can help to distribute central functions across new services. This reduces dependencies and makes later updates easier.

The pace at which a company can change plays a major role in the approach. Accompanying audits are often necessary to scrutinize the current system landscape and realistically estimate the modernization need. This brings the topic of reducing technical debt to the fore: if we wait too long, maintenance costs add up and even higher levels of specialist know-how become necessary.

We also rely on agile project methods to realize iterative partial successes. This gives our clients time to build new competencies internally. It is important that decision-makers at the highest level signal their support and free up resources for modernization initiatives — only then can the full potential of a switch be unfolded in the long term.

Securing and Expanding Know-how

Even when we replace or modernize legacy systems, we have to secure the existing knowledge about these technologies. Without clear documentation, subsequent development teams are confronted with functions whose origin they hardly understand. Often, COBOL, VB6, or Delphi contain routines that have grown over years — for example calculating customer-specific discount rules or intercompany settlements.

We advise capturing senior experts' knowledge early, before they retire. Structured knowledge management with code documentation tools, diagrams, and interview guides can help here. Enterprise wikis or collaboration platforms in particular prove useful for capturing recurring sources of error, special cases, and workarounds. It is also important that younger employees acquire a basic understanding of these old languages and "bring documentation to life."

A further key lies in targeted continuing education. Numerous companies are already starting to qualify their specialist staff for both modern and older technologies. We recommend setting up training programs for COBOL or Delphi so that a new cohort of developers can confidently move through legacy code. External training, bootcamps, or mentoring models are also conceivable, in order to pass on knowledge in stages.

In manufacturing, according to Seraph, training and reskilling programs are seen as a sensible investment for closing the skills gap [1]. Especially where traditional craftsmanship meets high-tech, a comprehensive training concept can help to make existing workforces future-proof. The same goes for software companies and IT departments: anyone who cannot replace COBOL, VB6, or Delphi immediately should offer structured mentoring that gradually introduces younger generations to the old systems.

The Role of Automation and Process Adjustment

For many functions previously handled manually in VB6 user interfaces, automated alternatives now exist. We argue for automating repetitive routines to reduce operational effort and direct the remaining developer resources to strategically important tasks. In some cases, it pays off to spin out individual sub-processes into microservices and thus build a modern, service-oriented architecture. Anyone wanting to break up large monoliths can take their cue from our experience report on the monolith to microservices migration.

Automation solutions also help to relieve the specialists who still hold the system knowledge. When everyday tasks such as data entry or simple code builds disappear, more capacity remains to think through critical legacy sections and design a roadmap for renewal. By digitizing classic integrations, interfaces become more uniform and the risk of errors within legacy systems decreases.

We know from projects in banking and insurance that well-planned automation can significantly extend the lifespan of old technologies. While work on replacement or modernization is already happening behind the scenes, automated processes create time savings and stability. The principle becomes clear in data migration in legacy systems: anyone who establishes clean, automated workflows here early on reduces error rates during the move to a new system and can keep live operations stable.

Continuing Education as an Antidote to the Skills Shortage

Given the persistent trend of skills shortage in old technologies, continuing education is a key solution. According to IBM, many technical skills — for example in AI or cloud computing — become outdated every 2.5 years, which is why constant adaptation is required [4]. The same applies to legacy technologies such as COBOL or Delphi, except that the learning curve here is often steeper, because the systems must be more comprehensively documented and standard training is rarely available.

We have learned that companies offering continuous learning opportunities and internal training formats benefit from lower turnover rates. In production too, institutes such as Fraunhofer IGD recommend creating training scenarios using virtual technologies and digital twins, so that a new generation of workers can learn routine tasks more quickly [2]. Applied to legacy software, a kind of virtual "sandbox" could be developed in which juniors can safely practice in old programming languages.

Moreover, a comprehensive continuing education program helps us increase the attractiveness of IT jobs, even when older technologies are involved. Anyone combining a secure workplace with exciting modernization projects signals to candidates: "Here I can shape things, learn, and take on responsibility." That is a motivator that can counter the supposedly old-fashioned image of COBOL or VB6.

Measures to Minimize Risk

To arm themselves against outages and knowledge loss, companies should take various measures early. This includes a structured onboarding program for new employees who come into contact with the old code. Equally essential is documenting all important application areas. We generally advise:

1. Documentation concierge: The team designates an experienced person who systematically searches and documents all code modules.
2. Version control: Even with older systems, it pays to use modern tools to track all changes.
3. Test automation: Even if COBOL or Delphi seems unfamiliar, basic tests can be automated to safeguard system stability.
4. Handover plan: Anyone retiring hands over knowledge in defined months to a successor or to a documentation team.

With these to-dos, we can already reduce risk significantly and increase maintainability. Sudden knowledge gaps can thus be prevented. Last but not least, a contingency scenario should be defined — that is, who can step in if a critical module suddenly fails and the few experts are not reachable. Such precautions protect us from complete standstills and prevent worse damage.

The Path to Replacing Legacy Systems

Ultimately, no company can avoid the question of whether and how to replace or modernize old technologies in the long run. From our point of view, it is worth conducting a comprehensive inventory and developing a roadmap plan from it. This plan defines which modules, databases, or front ends are renewed first and how we secure ongoing operations in parallel.

In practice, this often takes the form of carefully transferring individual system parts into new architectural components — for example, from a monolithic Delphi program to a microservices structure. The goal is to gradually port the functional scope to a modern platform while keeping an overview via modernization strategies for legacy software. It must not be forgotten that the old logic is still business-critical. A clean migration path and good tests are therefore decisive to prevent downtime or data inconsistencies.

The discussion of whether to choose big bang vs. incremental migration plays a central role here. With a Big Bang, the entire system would be replaced in one go — a highly risky undertaking, especially when little specialist personnel is available. We therefore more often recommend a gradual approach, combined with projects for the parallel training of new staff. The company can thus "rehearse" multiple times and learn how old and new can coexist.

Costs, Benefits, and Future Outlook

A core concern is the cost question: is it worth investing several years in modernization? We believe it is, because the long-term savings are considerable. Companies updating their old systems can reduce maintenance costs of legacy systems and at the same time lower the risk of expensive production outages. In addition, recruiting IT professionals becomes easier if you offer a change of perspective: "You can contribute to our modernization while at the same time building valuable knowledge of legacy software."

Looking at the IT sector as a whole, the world of work is changing rapidly. According to Bitkom, more than 600,000 IT professionals will be missing in Germany by 2040 if no countermeasures are taken [5]. This shortage applies not only to new technologies but also to traditional languages that remain part of our IT landscape. Anyone who invests early in training and modernization will be less vulnerable to acute personnel gaps — and at the same time ready for the requirements of future markets.

Conclusion

In the German mid-market, the shortage of skilled workers for old technologies is an increasingly pressing topic. COBOL, VB6, and Delphi continue to keep critical business processes running in many industries. But when the last developers retire, it is high time to act. We recommend a holistic approach that pursues three goals in parallel: securing knowledge of old systems, training employees, and initiating sustainable modernization.

It is essential not only to look for external specialist personnel in the short term, but also to inspire young talent for the modernization journey. Through targeted continuing education, mentoring programs, and automated processes, legacy systems can be transformed step by step. Clearly defined role allocations, good risk management, and open project teams that are willing to dive into old code bases all help here.

In the long term, there is no way around a systematic replacement or thorough renewal. Whether we replace individual services using the strangler fig pattern migration or instead launch a comprehensive modernization project: what matters is that a concrete plan exists and that those responsible tackle the issue consistently. In this way, we reduce dependencies on dying-out know-how and create the basis for greater innovation and competitive advantage.

Anyone acting today not only protects themselves against costly system failures. They also lay the foundation for a future-oriented IT landscape in which modern technologies and proven structures can calmly coexist. We see a major opportunity in this approach for sustainable growth, stability, and a smooth generational change — even when it comes to COBOL, VB6, or Delphi.