The Defects Planning Dilemma

While coaching Agile teams, I’ve noticed a recurring dilemma: whether or not to plan defects for a sprint, essentially pulling defects into the sprint during the sprint planning event. Many teams follow the convention of planning all backlog item types into a sprint, including defects. When asked why, the answer usually revolves around the desire to have control over which defects the team will be fixing. But here’s the question we need to ponder: Does knowing exactly which specific defects will be fixed truly lead to improved quality?

In this article, I will argue that the practice of pulling defects into a sprint during sprint planning can be counterproductive to both quality and velocity. Let’s delve into some of the downsides of planning defects.

The Downside of Planning Defects

1. Quality Prioritization: When you plan defects at the start of a sprint, there’s a risk of prioritizing less severe defects over more severe ones, as the more severe ones might be created during the sprint's lifetime. In practice, this means that the team ends up handling less severe defects before addressing more severe ones.
   If the team decides to tackle the newly created more severe defects first, my experience shows that most pre-planned defects are often left untouched, which turns the pre-planning into a waste of time.
2. Estimating Defects: Planning defects into a sprint requires pre-estimating them. But, as explored in my previous post on “Estimating Defects Leads Teams Down a Rabbit Hole,” this pre-estimation can be a challenging and time-consuming exercise that may not yield meaningful results.
3. Dynamic Nature of Defects: Unlike user stories or tasks, defects are dynamic and change frequently. Some defects that seemed relevant during sprint planning may turn out to be non-relevant as the sprint progresses. This dynamic nature makes defects less suitable for pre-planning compared to other backlog items (e.g. stories and tasks).
4. Board Overload: Pre-planning defects can clutter the Scrum board with numerous tickets, potentially overloading the board management and making it less effective in visualizing the team’s work.

An Alternative Perspective

A more effective approach would be — instead of pre-planning defects, the team should simply leave enough room for handling defects by adjusting their sprint commitment for content delivery.

When a team member is tasked with resolving a defect, they have the liberty to simply select the next defect from the top of the defects backlog — namely, the most critical defect at that particular moment.

This approach not only affords the team with the much-needed flexibility to address the most severe defects promptly but also eases the burdensome pre-planning efforts. In doing so, it opens up the opportunity for marked improvements in both quality and velocity, as valuable time and resources are optimized toward the most pressing concerns.

Balancing Content with Quality — An Example

The following chart unveils the dynamic journey of our team’s velocity across 16 sprints, each spanning a two-week period. Just beneath it, lies a revealing defect trend chart, offering insights into the flow of pending defects from November 2022 to September 2023.

Sprint Velocity of a Team (Feb-Sep 2023)

Defect Trend of a Team (Nov 2022 — Sep 2023)

Let’s delve into the graph, step by step:

• November to March: During this phase, the number of pending defects steadily escalated.
• March to April: As an organization, we set the bar high for quality standards, aiming to limit the number of pending defects to no more than 20. To align with this goal, the team made a strategic shift around March by pulling in less content during each sprint. This allowed us the necessary breathing room to address existing defects effectively.
• Throughout this journey, the team rigorously scrutinized the defect trend. Unsatisfied with its trajectory, we consciously pulled back on content from sprints 35 to 38.
• Upon concluding sprint 38, we finally achieved a satisfactory defect trend, leading us to confidently expand our content intake in sprint 39.
• June: A significant turning point arrived in June when our team decided to embark on a ‘zero defect policy.’ As a result, sprints 42 and 43 witnessed a deliberate reduction in content.
• End of July: Our commitment and dedication bore fruit, as by the end of July, we successfully attained the ‘zero defect policy’ milestone, with the defect count plummeting to below 10. This accomplishment empowered us to ramp up our content inclusion from sprints 44 to 49, ultimately establishing an impressive average velocity of 33 Story Points per sprint, all while upholding our ‘zero defect’ commitment.

Conclusion

In conclusion, the practice of planning defects in sprints, while well-intentioned, may not be the most effective approach to defect management in Agile teams. It can lead to prioritization issues, neglect of pre-planned defects, and the burden of pre-estimation. Embracing a more flexible approach that balances defect handling with commitment to content delivery can enhance both the quality of your software and the content delivery of your teams.