tech debt, software engineering, agile, devops

How to Effectively Manage Tech Debt

As discussed recently on Twitter, tech debt is an issue that deserves more attention, given its potential to derail products and projects in IT organizations. This article provides an opportunity to not only announce that Delivery Management is now active on Twitter, but also to delve deeper into this topic. 

Weā€™ll explore what tech debt is, what causes it in agile environments, the challenges in managing it, and why anyone overseeing projectsā€”like project managers, product managers, and engineering managersā€”needs a new perspective on addressing it.

What is meant by tech debt?

Technical debt, often shortened to tech debt, occurs every time we choose quick and dirty solutions now over better-designed, more future-proof options. Like financial debt, tech debt accrues interest payments in the form of extra effort required later to maintain and extend suboptimal systems.

Weā€™ve all been there ā€“ under pressure to ship a feature ASAP, you cut corners and leave clean-up for tomorrow. But that messy code or sloppy architecture sticks around, making each new change slower and buggier. Before you know it, your codebase resembles a giant house of cards in danger of toppling over.

While taking on some tech debt speed development in the short-term, letting it pile up guarantees painful delays down the road.

Technical debt and Agile

Agile organizations are at a higher risk of encountering tech debt, as teams prioritize delivering new features over building a strong technical foundation.

With agileā€™s focus on rapid iterations, developers feel pressure to take shortcuts to meet quick deadlines. These shortcuts not only increase the risk of bugs but also accumulate as debt over time.

Additionally, lack of automated testing and continuous refactoring can cause debt to accrue.

Without oversight, this debt can accumulate silently, setting the stage for disastrous outcomes. For this reason, particularly within organizations that embrace agility, empowering technical leadership becomes vital.

Who Plays a Role in the Creation of Tech Debt

Different people contribute to the creation of tech debt in different ways.

technical debt, stakeholders, product managers, developers
  • With few exceptions, stakeholders usually lack the technical understanding to fully grasp the importance of technical debt. They also have low visibility into technical debt and, even if they had, they do not see any immediate benefit in addressing technical debt. As a result, they pressure product managers and development teams to prioritize delivering new features over fixing underlying issues.
  • Often product managers lack the technical understanding as well. As a result, they fail to allocate enough capacity for addressing tech debt. Also, since tech debt usually manifests as obstacles to user stories rather than user stories themselves, itā€™s often hard to prioritize.
  • The engineering team also play a role by implementing workarounds rather than long-term solutions. Also, proper documentation of technical debt is required and engineers must continually maintain dependencies and document emerging issues in order to log these issues and prioritize fixing them accordingly. Without this upkeep, not only is technical debt created, but it is also doomed to grow exponentially over time.

Sprinting Toward Disaster

Let me share a cautionary tale, this is something I personally witnessed in a startup here in Europe:

  • The backend was built in-house without a clear vision, reacting only to managementā€™s ever-changing strategy and feature requests.
  • Six small development teams (each with its product manager) focused narrowly on the delivery of their own features, without considering holistically the broader impact on the system.
  • At least the company wisely implemented a feature switch to enable / disable features withouth having to touch and re-deploy the code, but the lack of documentation over time made it difficult to decide what could be safely disabled.

As you can imagine, the application grew into a slow, bug-ridden, monstrous, mythological creature no one knew how to tame. Though engineers agreed to dedicate 20% of the time of each sprint to addressing technical debt, product managers failed to prioritize this, pressured by stakeholders to deliver new features at all costs.

I do not know what ultimately happened to that application. The startup is still operating, but I do not know if they had to halt development for extensive refactoring of their backend or even start over from scratch, preserving only the essential elements.

In any case, the moral is clear: ignoring technical debt can quickly become disastrous. Without proactive management, short-term feature gains lead to unmaintainable systems in the long run.

Reframing Technical Debt

Technical debt is often viewed as an afterthought in many companies, especially by non-technical stakeholders. However, it is crucial that organizations start reframing the discourse around it and recognizing it for what technical debt truly is: a hidden cost that accrues interest over time.

If left unaddressed, tech debt can lead to slowed feature development, reduced system stability, and increased costs down the road.

To prevent these impacts, stakeholders must shift their mindset to treat technical debt with the same urgency as financial debt.

Teams should regularly allocate sprints and resources specifically to pay down technical debt through activities like refactoring, and improving test coverage. Just as financial debt can cripple a business if ignored, technical debt will continue to accumulate as a silent tax on engineering efficiency.

By changing perspectives on technical debt and making it a priority in planning discussions, organizations can mitigate rising maintenance costs, improve velocity, reduce bugs, and enable innovation.

Recommended Actions to Manage Tech Debt

The following is a list of 10 actions I would recommend to companies plagued by technical debt:

  1. Get leadership buy-in to prioritize tech debt work
  2. Analyze codebase to identify major problem areas
  3. Automate quality checks to measure tech debt over time
  4. Raise tech debt issues in the daily stand-up and sprint retrospective meetings
  5. Ensure tech debt items are added into product and sprint backlogs
  6. Review overall debt levels during planning meetings and prioritize accordingly
  7. Allocate developer time in each sprint for refactoring and debt repayment
  8. Adjust the Definition of Done to make it ā€œready to releaseā€, rather than ā€œready to be testedā€
  9. Refine the code review process to check for technical debt accumulation
  10. Bring in outside help if needed to objectively assess and address debt

Conclusion

In summary, technical debt is a critical yet often overlooked factor that can dramatically impact engineering velocity and product quality over time.

To prevent runaway accumulation of tech debt, organizations must make its management a priority through leadership buy-in, dedicated resources, and cultural changes that reward technical excellence.

While paying down debt may temporarily slow feature development, it is an investment that prevents exponentially greater slowdowns in the future. Just as prudent financial management enables long-term growth, keeping technical debt in check is essential for engineering teams to maximize their effectiveness and build products that stand the test of time.

Picture of Manfredi Pomar
+ posts

Italian cloud computing professional with a strong background in project management & several years of international experience in business consulting. His expertise lies in bridging the gap between business stakeholders & developers, ensuring seamless project delivery. During his free time, he enjoys fatherhood and immersing himself in nature.

Be a Content Ambassador
Skip to content