How to Stifle Open Source Without Banning It
Benedikt Langer
6 Min. Read The sharpest argument against China’s top open AI comes from a man at OpenAI. Dean Ball, ...
Read Article
7 min read
Technical debt doesn’t appear on any balance sheet, yet it costs every corporation dearly. According to McKinsey, 10 to 20 percent of new-product budgets is swallowed by cleaning up legacy baggage, and for one-third of CIOs the figure is even higher. Treating this debt as an engineering housekeeping chore turns a capital question into a maintenance issue – one that deserves a seat at the budget table alongside product investments, platform strategy and risk appetite.
Key Takeaways
- Technical debt is a balance-sheet item that never appears on the balance sheet. McKinsey estimates it at one-fifth to two-fifths of the entire technology estate and says it determines how quickly a company can still build anything new.
- A blanket budget slice isn’t enough. Ring-fencing 15 percent for legacy fixes fizzles out. The debt only shrinks when every investment decision explicitly accounts for it.
- Three roles must sign off. A repayment envelope is capital allocation by CEO, CFO and CIO together – not a leftover IT slush fund.
Related:Zero Trust needs process knowledge/What supervisory boards must demand when 89 percent of AI strategy is improvised
What technical debt really costs the board
Technical debt is not a metaphor – it is frozen decision-making power. What is technical debt? It is the accumulated effort required when systems are built fast instead of right: outdated architectures, unmanaged dependencies, patches that were never undone. Like a loan, it accrues interest in the form of slower development, higher failure risk and effort that never converts into new value.
The numbers make the scale tangible. McKinsey puts the burden at 20 to 40 percent of the entire technology estate before depreciation. At the same time, one-tenth to one-fifth of every new-product budget is spent not on new features but on fixing old ones. For roughly one-third of CIOs surveyed, the share exceeds 20 percent. That skews every growth plan before the first roadmap is even approved.
20 to 40 %
of the value of the entire technology landscape is locked in technical debt before any depreciation is applied, McKinsey estimates.
Source: McKinsey Digital, Tech Debt
Many companies reflexively delegate the problem to engineering. That is where the mistake begins. Without board-level visibility into its size, there is no mandate – and no budget – for paying it down. The debt then grows quietly until a project suddenly becomes unexpectedly expensive and nobody can explain why.
From suspicion to control in three steps
Technical debt can be managed once it’s translated from developer diagnostics into investment logic. Three steps are all a management team needs to demand.
The third step is the uncomfortable one. A one-off cleanup project goes nowhere because debt grows back immediately. McKinsey draws the line clearly: a 15–20 % IT-budget carve-out isn’t enough. The trend only turns when every budget explicitly shows new debt and repayment together.
What carries a repayment framework – and what stalls it
A fixed framework for retiring technical debt sounds simple. In practice, its anchoring inside quarterly and investment processes decides success.
What carries
- A measurable debt KPI in the quarterly report
- A named owner at C-level, not buried in a team
- Repayment baked into every investment, not treated as a side project
What stalls
- Debt reduction that only happens in “quiet” quarters
- KPIs nobody on the board reads
- A dedicated pot that vanishes at the first budget squeeze
The braking mechanism is the same in many houses: repayment is the first thing cut when budgets tighten, and only reappears when roadmaps start to topple.
Who ultimately decides
The most common misconception is that technical debt is a CIO question. The CIO can quantify it, but cannot retire it alone. A repayment framework is capital allocation and needs the CFO as much as the CEO’s priorities. McKinsey names the cast clearly: CEO, CFO and CIO together. The CIO brings the debt map, the CFO supplies the framework, and the CEO sets the priority against every other investment.
The payoff is concrete. Where technical debt is actively managed, McKinsey finds development teams gain up to 50 % more time for work that feeds business goals. That time is the real return on repayment – it shows up not in the debt reduction itself, but in the speed with which new value is created afterward. The question of mandate decides here too whether insight becomes impact.
Frequently Asked Questions
What is technical debt?
The accumulated effort that arises when software is built quickly rather than cleanly: outdated architectures, unmaintained dependencies, and workarounds without follow-up fixes. Like a loan, it incurs ongoing interest in the form of slower development and higher risk.
How big is the problem in numbers?
According to McKinsey, it accounts for one-fifth to two-fifths of the entire technology stock before depreciation. Additionally, one-tenth to one-fifth of new product budgets is absorbed by eliminating legacy baggage, and for one-third of CIOs, the figure is even higher.
Why isn’t a fixed budget share enough?
Because once a separate pot is set aside, the debt doesn’t stop growing in parallel. Repayment only becomes effective when every investment decision books the technical debt it creates or reduces – rather than outsourcing it to a special project.
Who should decide on reducing it?
The CEO, CFO, and CIO together. The CIO provides the inventory and the interest costs, the CFO sets the financial framework, and the CEO sets the priority against other investments. If anchored solely in IT, repayment remains inconsequential.
What does active management actually deliver?
Where technical debt is managed, McKinsey reports that development teams gain up to 50 percent more time for work on business goals. The return therefore shows up less in the reduction itself than in the speed with which new initiatives can resume afterward.
Read more on Digital Chiefs
Digital ChiefsData Spaces: Where Smart Industry and Smart City ConvergeDigital ChiefsZero Trust Requires Process Knowledge, Not Just ToolsDigital ChiefsCloud and AI Development Act: Sovereignty Becomes a Requirement for ProcurementMore from the MBF Media Network
cloudmagazinFinOps sees everything but is powerless to act mybusinessfutureTool hygiene in SMEs: five hard lessons securitytoday14 malicious npm packages in four hoursImage source: AI-generated (June 2026)