Technical debt

In software development and other information technology fields, technical debt (also known as design debt^[1] or code debt) refers to the implied cost of additional work in the future resulting from choosing an expedient solution over a more robust one.^[2] While technical debt can accelerate development in the short term, it may increase future costs and complexity if left unresolved.^[3]

Analogous to monetary debt, technical debt can accumulate "interest" over time, making future changes more difficult and costly. Properly managing this debt is essential for maintaining software quality and long-term sustainability. In some cases, taking on technical debt can be a strategic choice to meet immediate goals, such as delivering a proof-of-concept or a quick release. However, failure to prioritize and address the debt can result in reduced maintainability, increased development costs, and risks to production systems.^[4][5]

Technical debt encompasses various design and implementation decisions that may optimize for the short term at the expense of future adaptability and maintainability. It has been defined as "a collection of design or implementation constructs that make future changes more costly or impossible," primarily impacting internal system qualities such as maintainability and evolvability.^[6]

Origin of the concept

The concept of “technical debt” was first coined by Ward Cunningham in 1992.^[7] After reading Metaphors We Live By, Ward devised this "debt metaphor" to explain to his boss the need to refactor the financial product they were working on.^[8][9] He wrote that:

"Shipping first time code is like going into debt. A little debt speeds development so long as it is paid back promptly with a rewrite... The danger occurs when the debt is not repaid. Every minute spent on not-quite-right code counts as interest on that debt. Entire engineering organizations can be brought to a stand-still under the debt load of an unconsolidated implementation, object-oriented or otherwise."^[10]

— Ward Cunningham

Similar concepts had existed before this. In 1980, Manny Lehman had published a similar law using an "architectural metaphor" for the deteriorating nature of software. Manny's Law states that:

"As an evolving program is continually changed, its complexity, reflecting deteriorating structure, increases unless work is done to maintain or reduce it."^[11]

— Meir Manny Lehman

It is important to understand that software architecture has been contrasted with civil engineering since the 1960s.^[12]

Causes

Frequent causes of technical debt

• Pressures by businesses to release sooner
• The implementation of last-minute specification changes or changes that are insufficiently documented or tested,^{[13]: 4 [14]: 22 [15]}
• Gaps in knowledge or skills, which may manifest as a lack of process understanding, insufficient knowledge, poor technological leadership, or inadequate mentoring or knowledge sharing practices.^{[14][14]: 17}
• Issues in the development process, such as
  • sub-optimal solutions
  • insufficient requirements (from process inefficiencies)
  • conflicting requirements on parallel branches
  • deferred refactoring, or
  • delayed upstream contributions.^{[15][14]: 29}
• Non-compliance with best practice, such as
  • insufficient software documentation
  • poor collaboration practices,
  • lack of ownership,
  • rewrites for outsourced software
  • inadequate attention to code quality
  • tightly coupled components
  • the lack of a test suite, or
  • failure to align to standards (including ignoring industry standard frameworks).^{[13][13]: 7 [14]}

Consequences

By increasing the cost of ongoing maintenance, technical debt makes it harder to predict release schedules. "Interest payments" result from incomplete work and escalating integration costs due to changes in the upstream project. Increases in complexity and the amount of uncompleted work make it increasingly difficult to accurately estimate effort, resulting in delays, missed deadlines, and stress on engineering teams, which can result in higher staff turnover, compounding the problem.^[16] Carrying technical debt into production increases the risk of outages, financial losses, and potential legal issues due to breached service-level agreements. Future refactoring becomes riskier and costlier, with modifications to production code introducing greater chances of disruption.^{[citation needed]}

Failure to address technical debt can cause productivity to decline and slow down the delivery of features. The cumulative effects of technical debt result in increasingly fragile systems that can make bold improvements difficult. The domination of incremental changes, along with delays to critical refactoring, can result in stressed systems with inconsistent design, causing users to suffer from degraded performance and limited functionality while developers struggle to maintain quality.^[1][17]

Planning

Kenny Rubin uses the following categories to help manage technical debt:^[18]

• Happened-upon technical debt—debt that the development team was unaware existed until it was exposed during the normal course of performing work on the product. For example, the team is adding a new feature to the product and in doing so it realizes that a work-around had been built into the code years before by someone who has long since departed.
• Known technical debt—debt that is known to the development team and has been made visible using one of many approaches.
• Targeted technical debt—debt that is known and has been targeted for servicing by the development team.

Limitations

The concept of technical debt assumes that an expedient design saves present costs at the expense of higher future costs. While often valid, this premise relies on assumptions that may not always hold, such as the assumption that the product must survive long enough for the deferred work to matter,^[19] or that future events or advancements may render expedient and "long-term" designs obsolete,^[20] or that new tools and techniques might reduce the cost of future rework, challenging current debt assumptions.^[20]

Given the uncertainty of the future, what appears to be technical debt today may ultimately prove to be an example of savings. Furthermore, traditional calculations of technical debt tend to focus only on development time, overlooking broader costs such as training and onboarding when debt affects code readability,^[21] licensing, tools, and infrastructure needed to manage or resolve the debt,^[22] and opportunity costs related to delayed features or lost market opportunities.^[22]

Without accounting for these factors, technical debt assessments risk oversimplifying complex trade-offs, leading to suboptimal decisions.

See also

• Code smell
• Big ball of mud
• Bus factor
• Escalation of commitment
• Manumation
• Overengineering
• Shotgun surgery
• Software rot
• Spaghetti code
• SQALE
• Sunk cost
• TODO, FIXME, XXX