ISG Provider Lens™ Mainframes - Services and Solutions - Mainframe Optimization Services - U.S. Public Sector 2025
Public sector investments in mainframes is growing as AI helps drive efficiencies.
The legacy mainframe is experiencing a renaissance that began in 2024 and is set to continue into 2025. Mainframe systems, traditionally regarded as liabilities due to maintenance challenges, are being reevaluated and recognized as strategic assets.
Investments in training and AI are starting to pay off, leading to reduced challenges in mainframe maintenance. The unique capabilities of mainframes to support AI models while reliably managing and protecting missioncritical data allow new value to be added to the IT ecosystem.
These new value propositions, when weighed against the high risk of migrating data and applications off mainframes, are prompting agencies to reassess the risk-reward relationship. Consequently, the choice to shut down mainframes is becoming less attractive.
Technology trends
Hybrid Architectures
As organizations come to grips with the reality of the eternal mainframe, they are refocusing on how these machines can function more efficiently and add value to a more connected IT ecosystem. In smaller implementations, agencies can completely exit their mainframes and shut them down. However, shutting down mainframes that support more complex and extensive mission-critical functions has proven to be much more challenging. Agencies and organizations are asking.
• What applications should we move off the mainframe, and which should stay?
• How can we efficiently access the data on our mainframe to support AI models and enhance decision-making?
• How can we optimize applications remaining on the mainframe?
New tools and services are emerging to help agencies address these questions and provide solutions. The focus on risk reduction and regulation compliance continues to drive much of the decision-making in state and local
government modernization strategies.
Importance of Access to Data
The requirement for access to data to support and train AI has created a demand for new strategies to leverage hard-to-access mainframes. Data stored on mainframes is typically the most important to the business operation, requiring robust protection and security layers. DevSecOps strategies, APIs and microservices that make data more accessible and secure will become “must-have” capabilities in the coming years.
AI
Applying AI to reduce the costs and risks associated with maintaining and modernizing mainframe applications is emerging as a leading use case. The use of AI for automatic translation of COBOL to more modern languages, which are easier to maintain, has not met expectations. However, using AI to map and document mainframe applications,
making them easier for humans to work on, has been significantly more productive.
AI also facilitates the creation of knowledge graphs and documentation customized to specific users, further improving the maintainability of mainframes.
AI is also helping to reduce the challenges posed by the limited number of skilled mainframe developers. AI code assistants and chatbots streamline the learning process and enable newer developers to write better code.
AI also holds the promise of repositioning the mainframe as a strategic asset. The announced IBM z17, equipped with Telum 2 chips, is set for release in 2025. This new product is poised to innovate mainframe capabilities with built-in AI accelerators designed to optimize running ML and LLMs. The mainframes’ ability to handle large volumes of data, coupled with its high processing power and robust security features, makes it well-suited for processing AI models. With many agencies’ most valuable data residing on mainframes, running models on these machines reduces the risks of moving business-critical data. While the z17 is not yet available, the perspective around the value that mainframes can offer is beginning to shift.
Rising Software Licensing Costs
The fees required to run mainframes have become an increasingly large share of overall costs. Many long-term licensing agreements are up for renewal across the industry in 2025, allowing ISVs to take advantage of this scenario
by demanding higher fees. Some licenses are reportedly increasing between 125 and 175 percent. Service providers work with clients to devise alternative strategies or leverage bargaining power to reduce software costs.
Rising software costs are also squeezing margins for MFaaS providers, making it increasingly difficult to sustain this business model. This challenge reduces the number of players in this segment, as economies of scale and a laser focus on costs are essential.
Investments in Training Are Paying off
The aging workforce of skilled mainframe developers has posed challenges for many years. Service providers that established training programs to attract younger developers to learn mainframes and build their skill sets are realizing the return on their investments.
These firms report an increase in younger professionals on their payrolls who possess the skills to maintain legacy code.
Changing Political Environment
The changing administration at the federal level in the U.S. is expected to have farreaching implications. Based on articulated policy changes, we anticipate changes in IT investments at the state and local levels.
One of the new administration’s primary goals is to reduce the size and influence of the federal government. This strategy will likely place greater responsibility on states to deliver services and make decisions. With added stress on legacy systems, state and local governments must be able to support these changes. Resources should also be allocated to states to support increased investments in modernizing mainframes.
This shift in responsibility may foster increased collaboration among state, local and education (SLED) agencies. States are already banding together to procure software, hardware and services and are beginning to exchange technology and services. This trend is expected to accelerate as they adapt to the shifting political landscape.
Offshoring Becoming Less Taboo
Restrictions on where sensitive government data can be stored have made it challenging for state and local governments to leverage low-cost offshore resources. However, the growth of hybrid architectures and mainframe
optimization strategies, such as DevOps, AIPs and microservices, are making it easier to offshore less sensitive code and data to lowercost regions. Governments are becoming more comfortable working with offshore teams if developers do not access sensitive data.
Uneven Growth
Innovation in the mainframe services market is being rewarded with growing revenues. Providers that invest in tools, processes and technology see significantly better growth rates than those that do not. Leaders are experiencing growth rates of approximately 20 percent, while laggards are generating little to no revenue growth from their mainframe business. This disparity is likely to grow in the coming year as new technologies and approaches are embraced and improved to deliver differentiated services and products.
While the mainframe’s role within IT stacks is evolving, the competitive positioning of service providers has not considerably changed. We have seen leaders gaining momentum, while firms not adapting to the mainframes’ changing role will likely become increasingly marginalized.
M&A Activity
M&As in the mainframe space have slowed since a particularly active year in 2003. In early 2024, IBM acquired Advanced and Rocket Software closed the acquisition of OpenText.
Access to the full report requires a subscription to ISG Research. Please contact us for subscription inquiries.