Banks have spent years investing in technology to become faster, more efficient and more responsive to changing customer expectations, yet the accumulation of those investments is creating a problem that receives far less attention than the latest digital banking or artificial intelligence initiative. Across many financial institutions, successive waves of modernisation have introduced new applications for customer engagement, payments, lending, compliance, fraud prevention, data management and operational automation, often without fully replacing the systems that came before them. Each platform may have been acquired to solve a legitimate business problem, but the combined environment can become increasingly expensive and difficult to manage as applications overlap, integrations multiply and critical information is distributed across different technologies. The result is a growing disconnect between the capabilities banks continue purchasing and the operational complexity they must absorb to make those capabilities work together. As financial institutions prepare for another wave of investment in AI agents, cloud infrastructure and real-time financial services, the competitive advantage may increasingly belong to banks that can simplify their technology estates rather than those that continue expanding them.
Technology Investment Has Created an Accumulation Problem
The modern banking technology environment is rarely the result of a single architectural decision. It is more often the product of years of individual investments made in response to changing business priorities, regulatory requirements, customer expectations and competitive pressures. A bank might introduce a new digital onboarding platform to accelerate account opening, purchase a separate fraud detection solution to improve transaction monitoring, implement additional customer analytics capabilities and later deploy an automation platform to reduce manual processing. Each investment can deliver measurable improvements within its intended function, but the cumulative effect may be a technology estate containing numerous applications that perform related activities while relying on different data structures, integration methods and operating arrangements.
The problem becomes more pronounced when institutions expand through acquisitions, enter new markets or operate multiple business lines with independently developed technology strategies. Retail banking, corporate banking, wealth management and payments divisions may maintain separate systems for customer information, workflow management, reporting and risk assessment, even where substantial opportunities exist to share infrastructure. Over time, applications introduced as temporary solutions can become permanent components of the operating environment, particularly when replacing them would require changes to established processes or create disruption for employees and customers.
This accumulation is not necessarily evidence of poor technology management. Banks operate within demanding regulatory and operational environments where replacing a functioning system can involve considerable risk, making incremental investment more attractive than wholesale transformation. However, the consequence is that the cost of technology increasingly reflects not only what a bank is purchasing today, but also the complexity of maintaining everything it has purchased in the past.
The Real Expense Is Often Between the Applications
Software licences, cloud subscriptions and vendor support contracts are visible components of technology expenditure, but they represent only part of the financial burden associated with a fragmented banking environment. A substantial amount of operational effort can be consumed by the work required to connect applications, reconcile information, maintain interfaces and ensure that changes in one system do not disrupt another. These activities rarely attract the same attention as a major platform implementation, even though they can become recurring obligations that continue long after the original project has been completed.
Consider a bank operating separate platforms for customer onboarding, customer relationship management, fraud monitoring, payments processing and regulatory reporting. Information collected during onboarding may need to be transferred into several downstream applications, with each platform maintaining its own version of customer records, identifiers or business rules. When customer details change, the institution must ensure that those changes are reflected consistently across the relevant systems, while failures in synchronisation can create additional work for operations teams and potentially affect customer service or compliance processes.
The integration burden becomes more difficult to control as the number of dependencies grows. Adding one application does not always create just one new connection, because the new platform may need to exchange information with multiple existing systems and participate in several operational workflows. Even where modern APIs simplify connectivity, banks must still manage authentication, data transformation, access controls, monitoring, version changes and the handling of failed transactions. As the environment becomes more interconnected, the cost of maintaining reliable operations can increase faster than the number of applications alone might suggest.
For technology leaders, this creates a challenge in evaluating the true cost of a new solution. A product that appears commercially attractive when assessed through its subscription fee and implementation budget may become considerably more expensive once integration, testing, ongoing maintenance, cybersecurity and operational support are included.
Banks May Be Paying Several Times for Similar Capabilities
Technology duplication is another consequence of fragmented purchasing decisions, particularly when different departments select applications to meet their immediate requirements without sufficient visibility into capabilities already available elsewhere in the organisation. Separate platforms may provide overlapping functionality for document management, workflow automation, customer communications, analytics, identity verification or case management, with each system carrying its own commercial agreement, administration requirements and technical dependencies.
In some circumstances, maintaining different solutions is justified because business units operate under distinct regulatory, operational or customer requirements. However, duplication becomes costly when the same underlying function is implemented repeatedly without delivering a corresponding improvement in business outcomes. A bank may have multiple tools capable of automating workflows while continuing to rely on manual processes because those tools are not integrated into a consistent operating model.
The financial impact extends beyond licence expenditure. Employees may require training on several systems, technology teams must maintain different skill sets, and security specialists must evaluate the controls and vulnerabilities associated with each platform. Vendor management teams also face a larger portfolio of contracts, renewals and service arrangements, increasing the administrative burden of managing the technology estate.
This is particularly relevant as banks invest in AI capabilities. Many existing software vendors are introducing AI assistants, automated decisioning and intelligent workflow features into their products, creating the possibility that banks will acquire similar capabilities through multiple platforms. Without a clear understanding of what is already available and where particular technologies should be used, institutions risk paying for overlapping AI functionality while making their architecture more complicated.
Fragmented Data Can Undermine the Value of AI Investment
The growing importance of artificial intelligence makes technology fragmentation a more strategic concern because AI systems depend on timely, reliable and appropriately governed information. Banks may possess extensive customer, transaction and operational data, but that information is often distributed across applications designed for different purposes, with inconsistent definitions, duplicated records and varying levels of accessibility.
A bank attempting to develop a comprehensive understanding of a customer’s financial relationship may need to combine information from core banking, lending, cards, wealth management, customer service and digital engagement systems. If those applications maintain different customer identifiers or update information at different intervals, the institution must invest in additional data engineering and reconciliation before it can generate a sufficiently reliable view. The challenge becomes greater when AI models are expected to support operational decisions rather than simply produce reports or answer general questions.
AI agents introduce another dimension because they may need to retrieve information from several applications and execute actions across multiple systems. An agent supporting a customer service process could require access to account information, transaction histories, case management tools and communication records, while an agent assisting with financial crime investigations may need to combine alerts, customer profiles, payment information and previous investigation outcomes. Each additional system creates requirements for integration, permissions, monitoring and control over the actions the agent is authorised to perform.
This means that the value of AI investment can be constrained by the complexity of the underlying technology estate. A sophisticated model cannot compensate for inconsistent customer records, unreliable interfaces or poorly defined ownership of operational data, and introducing more AI tools without addressing these issues may simply add another layer of technology to an already fragmented environment.
For banks, technology simplification is therefore becoming closely connected to AI readiness. The institutions best positioned to deploy AI effectively may not necessarily be those purchasing the largest number of AI products, but those that have established clearer data ownership, reusable services and consistent integration standards across their existing platforms.
Every Additional Platform Expands the Operational Risk Landscape
The cost of technology fragmentation is not confined to financial expenditure or operational inefficiency. Each application introduces dependencies that must be understood, secured, monitored and maintained, while the connections between applications create additional points at which technical failures or security weaknesses can affect banking operations.
A fragmented environment can make incident management particularly challenging because the source of a problem may be difficult to identify when a customer-facing service depends on several underlying systems. A delay in payment processing, for example, might originate from an external provider, an internal integration service, a data validation issue or an application that is functioning correctly in isolation but receiving incorrect information from another platform. Resolving such incidents requires visibility across the entire transaction journey rather than individual application performance alone.
Cybersecurity teams face similar complexity. Every additional platform may introduce privileged accounts, third-party access arrangements, software vulnerabilities and configuration requirements that must be managed consistently. Security controls can become uneven when applications are deployed at different times under different technical standards, particularly where legacy systems coexist with cloud services and newer API-based infrastructure.
The growing regulatory emphasis on operational resilience reinforces the importance of understanding these dependencies. Frameworks such as the European Union’s Digital Operational Resilience Act (DORA) require financial entities within scope to strengthen their management of information and communication technology risk, including relevant third-party dependencies. Although the requirements vary across jurisdictions, the underlying challenge is increasingly common: financial institutions need to understand not only which technologies they operate, but also how failures within those technologies could affect critical business services.
This creates a strong argument for treating architecture simplification as part of operational risk management rather than merely a cost-reduction initiative. Reducing unnecessary applications and dependencies can improve the ability to identify incidents, manage changes and maintain consistent security controls, provided consolidation does not create excessive reliance on a single provider or system.
Simplification Is Becoming a Strategic Technology Decision
The difficulty for banks is that simplifying a technology estate can be considerably more complicated than purchasing another application. Systems that appear redundant may support important business processes, contain historical data or depend on interfaces that are poorly documented. Removing one platform can therefore require changes across several departments, along with extensive testing to ensure that regulatory reporting, customer services and operational controls continue functioning as expected.
Deloitte’s research on banking software engineering highlights the scale of this challenge. Its analysis, based partly on interviews with banking technology leaders, identifies inefficient development practices, costly maintenance, integration difficulties and legacy applications as persistent obstacles to engineering productivity. Deloitte also reports Gartner estimates that US banking and investment services enterprises spent approximately $107.8 billion on software in 2024, illustrating the scale of investment involved, although that figure should not be interpreted as a measure of unnecessary spending.
The solution is unlikely to be a wholesale reduction in the number of vendors or an attempt to consolidate every function onto a single platform. Specialist banking applications often provide capabilities that general-purpose systems cannot easily replicate, while excessive consolidation can introduce concentration risk and reduce flexibility. The more useful approach is to identify where duplication creates little additional value, where integrations impose disproportionate costs and where common capabilities could be delivered through shared infrastructure.
This requires banks to evaluate technology at the level of business capabilities rather than individual products. Instead of asking whether a particular application is performing adequately, institutions need to understand how many systems contribute to the same function, how much those systems cost collectively and whether the resulting architecture supports the bank’s longer-term operating model. Such an assessment can reveal opportunities to retire applications, consolidate overlapping functionality, standardise interfaces and reduce unnecessary dependencies without compromising essential services.
The commercial implications also extend to technology vendors. Banks seeking to simplify their environments may become more selective about purchasing standalone applications that require substantial integration work, favouring solutions that demonstrate interoperability, support open standards and can operate within existing architecture. Vendors that help financial institutions retire legacy functionality, consolidate workflows or reduce operational complexity may therefore offer greater strategic value than those competing primarily on the number of features available in their products.
The Next Technology Advantage May Come From Owning Less
For much of the digital banking era, technology investment has been associated with acquiring new capabilities. Banks have competed to launch additional services, deploy more sophisticated analytics, modernise customer channels and adopt emerging technologies, often measuring progress through the functionality introduced or the speed at which new platforms could be implemented.
That approach is becoming harder to sustain as financial institutions face simultaneous demands to invest in AI, strengthen cybersecurity, improve operational resilience and control operating expenditure. Every new technology initiative competes for financial resources, engineering capacity and management attention, while the existing application estate continues requiring maintenance and support.
Deloitte’s banking industry outlook has emphasised the need for more sustainable cost transformation, including improvements in productivity, scalability and resilience rather than relying exclusively on short-term expenditure reductions. This perspective is particularly relevant to technology rationalisation because the financial benefits may emerge gradually through lower maintenance requirements, reduced integration work, simplified operations and greater reuse of existing capabilities.
The most significant change may ultimately be in how banks define the value of technology. A new application can create value by introducing a capability that did not previously exist, but removing an unnecessary application can also create value by reducing complexity, improving data consistency and allowing technology teams to concentrate on more productive work.
This does not mean that banks should stop investing in specialised solutions or abandon established systems simply because they are old. The objective is to develop an architecture in which every major platform has a clear purpose, its dependencies are understood and its contribution justifies the operational burden it introduces.
As banking becomes more dependent on interconnected software, the ability to manage complexity may become as important as the ability to introduce innovation. Financial institutions that can simplify their technology environments while preserving the capabilities they need may find themselves better positioned to deploy AI, respond to regulatory change and develop new services without continually increasing the cost of running the bank.
The next phase of banking technology transformation may therefore involve a less visible but potentially more valuable achievement: delivering more business capability through fewer unnecessary systems.
What this means for the industry
- Technology spending needs to reflect total ownership costs. Banks should evaluate applications beyond licensing and implementation expenses, accounting for integration, maintenance, testing, security and operational support.
- Application rationalisation could release capacity for innovation. Reducing unnecessary duplication can allow technology teams to devote more resources to developing new capabilities rather than maintaining overlapping systems.
- AI readiness depends partly on architectural simplicity. Consistent data, reusable services and well-governed integrations can make AI systems easier to deploy and control across banking operations.
- Operational resilience should influence consolidation decisions. Simplifying technology estates can reduce dependencies, but banks must avoid creating excessive concentration risk through poorly planned consolidation.
- Banking technology vendors may need to demonstrate broader value. Interoperability, migration support, open integration standards and the ability to replace redundant functionality could become increasingly important purchasing criteria.
- Technology portfolios require continuous evaluation. Banks should treat application rationalisation as an ongoing management discipline rather than a one-time cost-cutting exercise.
Image Source: Pexels.com
