After six to seven years of pushing the technical stack for growth, many companies reach a breaking point. Outages multiply. Minor changes consume weeks. Bugs appear with every release. The pattern is sharply worse in companies that experienced a pivot, where architectural decisions made for a different product are buried in the foundations of the current one.
What Is Happening
The root cause is typically a historical leadership choice: pushing for velocity quarter after quarter, deprioritising the CTO's calls for refactoring in service of a deal, a deadline, or survival. This is often driven by leaders longing for the speed of the early days, when new features could be developed in an afternoon.
Each shortcut — skipped tests, hardcoded configurations, deferred refactoring — is a loan against the technical stack. The interest compounds. Eventually the stack reaches bankruptcy, where even small changes carry unpredictable risk.
This is not a CTO problem. It is a capital allocation decision that the board should own. The tech debt was accumulated during a period of board-endorsed growth-at-all-costs thinking. The remediation requires board-endorsed investment.
The board endorsed the growth strategy that created the debt. The board should now endorse the investment strategy that resolves it.
What the Board Should Monitor
Leadership Behaviour
Are leaders consistently pushing for speed of execution while ignoring engineering calls for refactoring? Are senior technologists resigning shortly after taking the measure of the technical debt? Both are leading indicators that the stack is in worse condition than the board is being told.
When a newly hired senior engineer leaves within six months of measuring the tech debt, that resignation is a data point. When it happens twice, it is a pattern. When it happens three times, the board is being told the stack is healthier than it is.
DORA Metrics
Four measures tell the board the health of the delivery pipeline: deployment frequency, change failure rate, lead time for changes, and mean time to recovery (MTTR). These should be reported to the board quarterly.
What Are DORA Metrics?
DORA (DevOps Research and Assessment) is an industry-standard framework developed by Google's research team. It measures software delivery performance across four dimensions:
- Deployment frequency — How often the team ships code to production
- Change failure rate — What percentage of deployments cause a failure in production
- Lead time for changes — How long from code commit to code running in production
- Mean time to recovery (MTTR) — How long it takes to restore service after a failure
Elite performers deploy multiple times per day with less than 5% change failure rate. Struggling teams deploy monthly with failure rates above 45%. The gap between these two profiles is the gap between a healthy stack and a distressed one.
Resource Buckets
Each product team should operate with agreed resource allocations: a fixed percentage for tech debt reduction, for strategic roadmap initiatives, and for tactical or reactive work.
Typical Starting Ratios for a Distressed Stack
These ratios should be treated as a top-down constraint, not left to product teams to negotiate locally. Without protected allocations, tech debt always loses to visible feature work. Within the agreed tech debt bucket, the Engineering Lead owns prioritisation based on the team's technical strategy and direction.
Without protected allocations, tech debt always loses to visible feature work. The loudest stakeholder wins every quarter, and the infrastructure that enables sustainable delivery erodes invisibly until it fails catastrophically.
Investment Level
A distressed stack may require up to 40% of revenue allocated to product and technology, at least temporarily. The board should understand the business case clearly.
"We can ship at the current rate and risk a major outage, or invest 18–24 months and double sustainable shipping speed."
This is the business case the board should be evaluating. The question is not whether tech debt investment has a return — it is whether the company can afford to defer it any longer.
Questions for the Board
- 1. What are our current DORA metrics? Are they trending in the right direction?
- 2. Do we have resource buckets in place, and are they being enforced?
- 3. What is the business risk of another 12 months without significant tech debt investment?
- 4. Have we presented investors with a transparent recovery plan?