Measure your ‘dark matter’ to pay your tech debt
Sven Blumberg
Senior Partner, Istanbul
For all the difficulties and suffering during the pandemic shutdowns, it was also a time when technology took center stage and allowed companies to pull off incredible feats that kept businesses running—and innovating—at a pace that had previously seemed unimaginable. Surprisingly, the pace continued at many companies through 2022. But achieving those feats required workarounds and quick fixes that in many cases added to organizations’ existing technical debt—the “tax” a company pays in the form of resources needed to address legacy technology when developing new tech. For example, updating core enterprise resource planning (ERP) systems might require engineers to do multiple rounds of fixes and testing to address such complexities that have accrued to the system over time.
CIOs estimate that such technical debt can add up to 20 to 40 percent of the value of their entire technology estate (before depreciation).¹ This coming year is going to be one when much technical debt comes due for CIOs and business leadership, as realizing the full benefit of critical technologies such as cloud and AI often requires resolving that tech debt. Migrating an application with significant amounts of tech debt to the cloud without refactoring it, for example, means you’re simply moving your tech debt issues from one place to another, without getting the cloud’s full benefits.
Paying down the debt starts with a task that sounds simple but isn’t: identifying and costing it out. Tech debt is a little like dark matter in the universe—people know it exists, but finding and quantifying it is challenging. This quantification effort has to happen at the application level to be meaningful and requires a cross-functional team to do it because tech debt in one application has effects on the business and also on technology performance at the infrastructure level. Unless the business understands the dollars-and-cents value locked away in its tech debt, making the right trade-off decisions and maintaining support for them becomes challenging.
Once an organization has quantified its tech debt, it needs to think in terms of managing future tech debt in all projects going forward by adding 15 to 20 percent of resources to the cost of applications to clean them up. While this focus will be difficult given pressures to build resilience and, in many cases, to reduce costs, the value to the business can be so large that the CIO will need to hold the line. This is why undertaking the value analysis of what tech debt lies where in a company is so important.
¹ Vishal Dalal, Krish Krishnakanthan, Björn Münstermann, and Rob Patenge, “Tech debt: Reclaiming tech equity,” McKinsey, October 6, 2020.