By Aman Sardana, Application Architect, Capital One
Application modernization – migrating and upgrading legacy systems to modern architectures – often fails when organizations neglect data considerations. According to an industry research, 68 – 79% of legacy modernization projects fail or fall short of expectations. Application modernization is often framed as a cloud migration or technology refresh initiative. New platforms are adopted, legacy systems are decomposed, and architectures are redesigned. Yet many modernization programs stall, exceed budgets, or fail to deliver meaningful business value.
The uncomfortable truth is this: application modernization fails far more often because of data than the complexity of cloud technology.
Application modernization is not just modernizing the application design and the hosting platforms, it’s also a data transformation problem. In my experience, when the data strategy lags architecture decisions, modernization efforts inherit the same constraints they were meant to eliminate. Executives talk about compute, platforms, and vendors. Engineering teams refactor code, but the most stubborn constraint – the one that quietly dictates architecture, cost, risk and agility remains untouched: legacy data.
This article explores why ignoring data imperatives leads to failure, and how robust data architecture and governance practices are critical for success.
Cloud-first thinking misses the real dependency
Most modernization programs begin with infrastructure questions around cloud provider, hosting strategy, and the timelines to migrate the application to cloud.
These are important questions – but they are downstream ones. Applications are not independent units. They are tightly coupled to decades-old data models, schemas, batch processes, reporting assumptions, and regulatory controls.
When data architecture is treated as an afterthought, modernization efforts inherit every historical compromise embedded in the system:
- Monolithic schemas optimized for batch processing
- Shared databases supporting unrelated applications
- Implicit business logic hidden in stored procedures
- Data access patterns that assume low concurrency and static workloads
Moving such applications to the cloud does not modernize them – it simply relocates their constraints. Organizations frequently modernize applications while leaving core data models, ownership structures, and governance mechanisms untouched. Without a clear data strategy, modernization shifts technical debt rather than retiring it. A cloud-smart strategy that centers on the data combines two powerful IT philosophies: data-centricity, which treats data as a permanent, primary asset independent of applications, and cloud-smart, which prioritizes choosing the optimal environment for hosting the applications.
Data gravity is not a buzzword – it’s an operating reality
Executives often underestimate data gravity because it doesn’t appear on architecture diagrams.
In my experience designing the systems and especially the ones where the new system coexists with the legacy, the data placement and structure dictate:
- Latency and performance characteristics
- Cost predictability
- Security and compliance posture
- Failure modes and recovery timelines
Legacy data architectures were built for environments where scaling was vertical, data access was centralized, and failures were rare but catastrophic. Cloud-native systems assume the opposite: elastic scale, partial failure, distributed ownership, and continuous change.
While application teams focus on APIs, scalability, and deployment velocity, data debt quietly accumulates. This includes:
- Fragmented and duplicated datasets
- Undefined data ownership
- Inconsistent semantics across business domains
- Legacy reporting logic embedded deep within applications
These issues directly limit agility. They slow product launches, complicate regulatory reporting, and undermine analytics initiatives. Infrastructure modernization and cloud hosting doesn’t compensate for unresolved data debt.
The result is a paradox many organizations experience firsthand: cloud hosted platforms that are technically modern but operationally brittle.
Why application teams can’t solve this alone
Application modernization initiatives are often delegated to engineering or platform teams. Data, meanwhile, is treated as a governance or reporting concern. This organizational split is one of the root causes of failure.
Application teams cannot safely modernize when:
- They don’t control data ownership
- Schema changes require cross-organizational approvals
- Data contracts are undocumented or implicit
- Downstream consumers are unknown
As a result, teams optimize for minimal disruption rather than long-term sustainability. Legacy data structures remain untouched, and modernization becomes cosmetic rather than transformative.
This is why application modernization is fundamentally a data leadership problem—one that requires executive ownership, not just technical execution.
A data-first modernization mindset
Successful modernization programs reverse the sequence. Instead of asking how applications move to the cloud, they start by asking:
- Which data domains are truly core to the business?
- Who owns them, and who depends on them?
- How should data be structured to support scale, resilience, and regulatory needs?
A data-first modernization approach has a structured process for keeping data central to the overall plan.
- Domain-aligned data ownership
Breaking shared databases into domain-owned data models reduces coupling and enables independent evolution.
- Explicit data contracts
APIs and events replace implicit database dependencies, making change safer and more predictable.
- Modern data access patterns
Designing for concurrency, partial failure, and elasticity—rather than retrofitting legacy assumptions.
- Incremental decoupling
Data modernization does not require big-bang rewrites. Targeted strangler patterns applied at the data layer often unlock disproportionate gains.
Cloud-native and event-driven architectures amplify the importance of sound data foundations. Microservices introduce distributed data ownership. Event streams require shared definitions and trust in data quality. AI and real-time analytics depend on consistent, well-governed data flows.
When data architecture is treated as an implementation detail, organizations experience:
- Proliferation of point-to-point integrations
- Conflicting metrics and dashboards
- Delays in compliance and audit readiness
- Reduced confidence in enterprise data
I have seen a pattern where modern platforms succeed only when data models, governance and lifecycle management are designed deliberately alongside applications.
Prioritizing data integrity means ensuring accuracy, completeness, and security of data is a top priority, not a checkbox. This includes building robust access controls, audit logs, and backup/recovery plans for data as part of the new system’s design, not afterwards.
Data fabric is another modern architecture pattern that complements modernization by providing a virtualized, integrated layer for data access. A data fabric uses technologies like metadata management, knowledge graphs, and virtualization to pull together data from disparate sources (legacy and modern) and make it appear as a unified data layer for consumers
The CIO’s role in modernization success
Successful modernization programs increasingly require close collaboration between CIO and Chief Data Officer (CDO). CIOs with the help of CDOs bring the cross-domain perspective needed to align data strategy with business outcomes.
This includes:
- Defining authoritative data domains and ownership
- Establishing shared semantics and data contracts
- Embedding governance into pipelines, not processes
- Ensuring modernization roadmaps explicitly retire data debt
When CDOs are involved early, modernization shifts from a technology upgrade to a platform for long-term value creation.
When data leadership is proactive, application teams gain the freedom to modernize safely. When it is not, cloud programs become expensive exercises in constraint preservation.
Modernization guidance from Microsoft explicitly advises verifying that policies for data security, privacy, and compliance are in place as part of readiness, and that data management is assessed alongside apps and infrastructure. In practice, organizations might form data governance councils or enlist a Chief Data Officer during modernization to oversee data policies. This governance focus prevents the “wild west” of unmanaged data that can derail new systems.
Modernization that lasts
Cloud platforms are powerful enablers – but they are not magic. Without modern data foundations, application modernization remains fragile, costly, and incomplete.
Organizations that succeed recognize a simple principle: applications can only be as modern as the data they depend on.
The question for leaders is no longer whether to modernize, but whether they are willing to confront the hardest part of the problem. Those who do build systems that scale, adapt, and endure. Those who don’t eventually discover that the cloud merely made their legacy faster – and many times more expensive.
Application modernization fails not because organizations choose the wrong cloud or framework, but because they underestimate the strategic weight of data.
For CDOs, modernization is an opportunity to move from stewardship to strategy—shaping platforms that are not just modern, but meaningfully intelligent.
When data leads, modernization delivers.
The views and opinions expressed in this article are those of the author and are provided in a personal capacity. They do not necessarily reflect the views, policies, or positions of any employer, organization, or professional body with which the author is affiliated.
Aman Sardana is a senior technology leader and architect with extensive experience designing and modernizing mission-critical platforms across financial services and fintech. He is a Fellow of BCS, The Chartered Institute for IT, recognized for his contributions to enterprise architecture, cloud strategy, and resilient system design. Aman is a frequent speaker at international technology and leadership conferences, where he shares practical insights on cloud architectures, building resiliency systems, and modernization of critical platforms. His work focuses on helping organizations balance innovation with reliability, security, and regulatory expectations. He holds an M.S. in Information Technology from Northwestern University and is a member IASA Chief Architect Forum.