What is technical debt (and its risks for product teams)?
Technical debt (or tech debt) is additional development work that must be done after a release is shipped. Typically, technical debt occurs when developers opt for a more limited solution than what was originally promised, often to save time and ship code faster. But that work — or "debt" — will need to be revisited and completed later, similar to repaying a loan.
Technical debt takes many forms. It can refer to architecture and infrastructure work, bugs, design, documentation, processes, security, testing, and even coding style. Sometimes product management and engineering teams will disagree on what exactly constitutes tech debt. A good rule? Consider any shortcuts or trade-offs that you take to meet customer demands and deadlines as technical debt that will be "owed" at a later date.
Is technical debt good or bad?
The concept of debt evokes discomfort in most people. No one likes to be in the position of owing something. But at the same time, taking on debt can be useful when done wisely. It can help you reach goals faster — like earning a degree or purchasing a home — and can be managed responsibly when you plan for it.
The same goes for technical debt. In certain scenarios, product teams will need to consider if it makes sense to assume technical debt. For example, say you discover a bug in a new product feature that is set to launch the next day. The bug will take more than a day to fix — but it is minor and will not be noticeable to most users. Your team agrees that it is not worth delaying the time to market, so you move forward with the launch and decide to address the bug soon afterward. In this case, accruing tech debt helped your team hit the launch target without major consequences.
The tricky part is that technical debt is often much more complicated than simple bug fixes — and it is not always intentional. Unknown issues that require rework can crop up at any time. These could be symptoms of larger problems, such as outdated legacy code or architecture that cannot scale.
Tech debt also accrues interest, in a sense. The longer it sits around, the more costly and time-consuming it becomes to resolve. And becoming inundated with tech debt can be demotivating. It uses up team capacity that would otherwise be used for product innovation and delivering new value for customers.
Boon or burden? In truth, tech debt is a balance. Many folks view it as an inevitable part of software development. The important thing is to establish a plan for how and when to address it so it does not hinder your product performance over time.
Related: What is included on an IT roadmap?
What causes technical debt?
Running into technical debt is a standard occurrence for most development teams. Here are some of the typical reasons it occurs:
Time constraints: As business needs change, development teams may be pressured to reduce time to market. This is when engineers will have to shift focus from quality to efficiency in some areas. This often means that features must be revisited later to ensure that the delivered functionality matches what was originally planned.
Business or market change: A shift in business goals, customer needs, or market environment can affect development work and even render it obsolete. You might assume technical debt to keep pace and stay relevant when plans pivot.
Lack of effort or expertise: People can unfortunately be a source of technical debt. Developers will sometimes seek shortcuts or stopgaps to save time — but cutting corners can lead to more bugs or flimsy code. Knowledge gaps can cause similar problems. Inexperienced developers could write lower-quality code that will need to be reworked later on.
Lack of resources or process: Whether tech debt is created intentionally or unintentionally, it takes effort to manage. Development teams must build processes to prevent and resolve technical debt so that it does not just pile up.
Technical debt is not a catch-all for messy development work or poor strategic alignment. Issues like persistent conflicting priorities, cross-team miscommunication, and poor quality assurance warrant a more critical eye toward your product development process as a whole.
3 types of technical debt
Technical debt can be categorized by type of work owed, such as security improvements or more testing. But on a higher level, most development teams designate three main types of tech debt:
Type of technical debt
Intentional tech debt Also referred to as active or deliberate tech debt
Development team knowingly takes on technical debt in order to finish a release faster.
Unintentional tech debt Also referred to as passive or inadvertent tech debt
Development team believes work is complete and ships the release — then discovers later that additional work must be done to fix problems that became known after delivery.
Environmental tech debt Also referred to as software entropy
Development team fails to actively manage technology advancements, code maintenance, and updates to legacy code over time — leading to a deteriorated product that requires greater effort to repair.
How to manage technical debt
There is no right way to solve technical debt. The amount and types of debt will vary for every development team, as will the tools and resources available for addressing it. But ultimately the motive should be the same — you want to find a sustainable way to manage technical debt alongside new development work driven by your product roadmap.
These best practices can help set you up to successfully manage technical debt:
Decide what it means. Do not waste time debating your definition of tech debt every time it occurs. Work together as a product team to proactively agree on what qualifies as debt. This will become the basis for evaluating and prioritizing future debt.
Build process around it. Technical debt is unavoidable at times — which means it should not be a surprise. Decide how you will incorporate tech debt work as part of your regular sprint or development periods, including how much time to spend on it and who will lead the effort.
Automate to save time. Automated code testing is a core agile development practice for enhancing release quality. Set up automation to help you efficiently review code and spot bugs and other issues before they turn into tech debt.
Set standards and track. Many development teams use agile metrics and KPIs to set benchmarks and track code quality. Including a measurement for tech debt will help you stay on top of monitoring it. (See some examples below.) Agile tools and extensions offer ways to track quality and throughput — to gain deeper insights into code and team performance.
Prioritize the most detrimental problems. Prioritization matters for all types of product development work by helping you to focus on what will drive the biggest impact. Use this same approach to help organize your technical debt and start chipping away at the most high-priority items.
Find a balance between new and improved. Deciding what to build next is often a point of contention between product managers and developers. While product managers tend to be most concerned with driving new product functionality, developers often vie for technical enhancements (like resolving tech debt). It helps when both teams can take a holistic view of priorities and balance capacity between both areas. This takes effective communication and a unified backlog.
Give recognition. Managing technical debt is sometimes unglamorous. It can feel more like cleanup than creative problem-solving. But it does not have to be the work of unsung heroes. Show that you value the time and effort spent on technical debt — reward engineers who do this work with the same praise and recognition as those who work on new features.
An example of how to manage technical debt alongside product roadmap-driven work in Aha! software.
How to measure technical debt
Many development teams use metrics to keep tabs on technical debt. Metrics like code quality and cycle time help you track where tech debt is stemming from and measure how well you are resolving it. Find some examples of technical debt metrics in the table below:
Simple count of bugs that are fixed or unfixed. This is the most basic indicator for how well your tech debt is being managed.
Count of how many times code has been deleted and reworked. Paying attention to higher code churn helps you identify problem areas before they become tech debt.
Number of lines of code that have been covered by your tests, expressed as a percentage (the higher, the better). A higher ratio of validated code can indicate a lower risk of issues that will cause tech debt.
Measurement of how many engineers have contributed to a given task. Code ownership helps you understand how efficiently you are planning and delegating development tasks, including managing tech debt.
Subjective measurement of "good" and "bad" quality code. Setting and tracking high code quality standards can help you avoid debt.
Time spent actively working on a feature from start to finish, including time spent on reopened issues. If cycle time seems unusually high, it is possible that some underlying debt needs to be addressed.
Technical debt ratio (TDR)
Measures the cost of fixing development work against the cost of building it. Cost can be measured in money or time.
If it feels like technical debt is dragging your team down, try to reframe your approach. Accept that it is often unavoidable and will never totally go away. Turn debt into an opportunity to make your product the best that it can be for your customers and future developers on your team. And remember that it should not consume all your efforts — your engineering team is most effective when you can improve on the past and work toward new goals in tandem.
- Introduction to information technology
- How does an IT team create business value?
- What is enterprise technology?
- Who makes up an IT team?
- What are some IT job titles?
- What is the role of an IT manager?
- What does an IT manager do each day?
- What skills do I need to be an IT manager?
- How can I learn to be an IT manager?
- What are some IT manager interview questions?
- What is a typical IT manager salary?