/
All articles

Less good solutions

It is not technical debt – it is information debt

When integrations stall and automation fails due to constant exceptions, we almost always blame technical debt and legacy systems. But the real bottleneck is rarely the code. It is information debt – and brand-new, modern cloud applications become part of it the moment they launch with their own isolated data models.

By Jens Fahlén · Published · 3 min read

A tangled blue monster surrounded by ones, zeros and mismatched data cubes

What is the difference?

When we talk about technical debt, most people think of outdated frameworks, messy architectures, legacy COBOL code, or systems that can no longer be updated without breaking everything. That is a very real debt that costs money and slows down development.

But information debt is about something else entirely: it arises when the information itself is modeled incorrectly, lacks unambiguous primary keys, or is defined differently across the organization.

A legacy system burdened with technical debt can still hold clean, accurate data, just as a technically flawless modern application can produce absolute conceptual chaos.

The modern application that made the debt worse

A common misconception is that replacing a system automatically eliminates debt. An organization retires a twenty-year-old enterprise application and purchases a modern SaaS solution with microservices, clean REST APIs, and an intuitive user interface. Yet what happens beneath the surface?

The new cloud solution introduces its own unique concept of what an "organization" or "unit" represents. It features an isolated user directory with proprietary local roles, rather than recognizing formal organizational process roles and delegations. It ignores standard external registration identifiers and instead introduces arbitrary internal IDs.

Technically, the organization may have reduced its technical debt. Architecturally, however, it has accumulated even more information debt. The new application is simply another modern silo, requiring bespoke integration logic and ongoing manual mapping to function within the enterprise.

Three signs you are suffering from information debt

1. Integration middleware is crammed with translation tables. If your integration layer requires extensive rules to translate "department" in system A into "cost center" in system B and "team" in system C, the issue is not your middleware. Your foundational information structure is fractured.

2. Automation fails over basic definitions. An attempt to automate an approval workflow or application provisioning stalls because no system accurately identifies who is the formal manager, who holds financial authority, or what role an employee performs within that specific process.

3. Shadow spreadsheets keep the operations running. When coordinators and administrators must maintain offline spreadsheets just to reconcile which internal codes match across systems, the software has abdicated its role as a reliable source of truth.

How do you pay off information debt?

Refactoring source code or licensing new software will not fix the issue. Paying off information debt requires conceptual discipline.

1. Establish unambiguous primary keys. No entity should exist without a stable, globally verifiable identifier. Legal entities rely on national business registration numbers, individuals on national identification numbers, and employments on unique employment IDs.

2. Enforce single-source data ownership. Every core attribute must be maintained in exactly one authoritative location. Downstream systems consume and derive data – local recreation of organizational hierarchies or management structures is prohibited.

3. Decouple structure from permissions and tools. Do not tie user roles directly to specific screens or applications. Assign individuals formal process roles within the organization, and allow tool access and permissions to be derived systematically from those roles and the organizational hierarchy.

Conclusion

Before you commission your next modern cloud service or embark on an enterprise integration project: do not evaluate just the APIs, throughput, or user interface.

Examine the data model. If the solution cannot align with a coherent, unified information architecture, you are not investing in the future – you are merely purchasing another vehicle for your information debt.

Tags

  • #information debt
  • #integration
  • #automation

What did you think of this article?