Fighting Pandemic Fatigue by Safely Experimenting With Engineering Team Structures

Posted by

In this fourth installment of our Tasty Dog Food Series, we continue to share stories of how we at Tasktop are using our own product, Tasktop Viz™, to boost our growth and happiness. Tasktop Viz is a value stream management solution that measures and optimizes the flow of business value in software delivery organizations. 

This time, I spoke to John Simoes, our Engineering Manager on the Tasktop value stream responsible for Flow Fabric™, which connects Tasktop’s VSM platform to the enterprise toolchain.  

According to John, he’d been waiting for a tool like Viz for the last ten years. “I use Viz on a near-daily basis to safely experiment with team structures and process changes. My goal is to improve our flow while maintaining high employee engagement, which has been especially challenging during Covid and the transition to remote work.”

John joined Tasktop in March 2020, just as lockdowns began, which meant he had to onboard and get to know his team entirely over Zoom. That went surprisingly well, but around December pandemic burnout set in, as captured in the team’s declining eNPS scores. 

John had to get increasingly creative to find ways to generate the sense of ownership, autonomy and impact that motivates engineers. “We wanted to mitigate burnout and minimize turnover, while remaining productive and competitive. As we began to experiment, Viz provided me with constant feedback on whether our changes and tweaks were working.” 

Experiment #1: Engineers Self-Select the Type of Work They Do 

John’s first idea to reinvigorate the engineers was to split them into two smaller teams. As a group that develops and maintains connectors to 60+ Agile and DevOps tools, one team would focus on new features and tools, while the other focused on maintaining existing connectors. The engineers could self-select the team they wanted to join. “Self-selection was important for creating a sense of ownership and fostering one’s ability to impact outcomes,” says John. Interestingly, for the most part, the less experienced engineers elected to join the maintenance team, perhaps lacking the confidence to take on completely new undertakings.  

As John implemented the new structure in mid-January, he kept a keen eye on both teams’ Flow Metrics. “This is how Viz was immensely useful right away, serving as a safety net for us once we implemented these experimental changes. It allowed us to observe the impact of the changes, so we could feel comfortable enough to experiment more.” 

Tasktop Viz illustrates declining eNPS scores (left) in December as pandemic burnout sets in, followed by John’s first experiment to split the team in January (right)

In February John noticed a decline in overall Flow Velocity, caused primarily by a decrease in Defect throughput. While Feature flow was good, it appeared that Defects were taking longer to close due to the relative inexperience of the team. 

“Viz was what brought this to my attention. It provoked the analysis that led to us having the right conversations with the team,” says John. As it turned out, the engineers needed more time to familiarize themselves with the connector in question before they could resolve a defect. That overhead manifested in two ways: 

  • Defect Flow Time was increasing. Since each connector is like a mini-product, before an engineer could dive into its code, they needed some time to learn and understand the context and perhaps consult a more experienced team member. 
  • Defect Flow Efficiency was dropping. While waiting for answers or help from a peer, the engineers would pick up another defect to ‘stay busy’, which increased context-switching and slowed things down even further. 
Tasktop Viz illustrates a Flow Time increase (left) and Flow Efficiency decrease (right) for defect work in February, following the split into two teams

Tasktop Viz illustrates a Flow Time increase (left) and Flow Efficiency decrease (right) for defect work in February, following the split into two teams

Clearly, this new structure was working for Features, but needed some tweaks on the maintenance side to optimize for the breadth of knowledge required to support so many tools. 

“This is where Viz proved incredibly valuable. Looking at the dashboards prompted us to have insightful conversations with the engineers as we tried to understand what was happening in our value stream.” 

Experiment #2: Introducing Specialization within the Maintenance Team

Flow Metrics had made it quite clear that despite the engineers self-selecting to be on the maintenance team, the lack of specialization and too much context-switching was impacting their productivity and speed. Through conversations with the engineers about what they were experiencing, the team surfaced ideas on how the scenario could be improved. 

“Once we’d uncovered the reasons for the problem, we could make adjustments to allow Flow Velocity for Defects to recover,” says John. He devised a second experiment, this time involving self-selection within the maintenance team itself. The goal was now to find an arrangement within the team that addressed this issue, while giving each engineer a sense of purpose and tangible impact. 

John asked each engineer to rate the connectors where they felt they could have the biggest impact considering their experience and proficiency. Once all engineers had ranked the connectors, tasks could be matched to an engineer intentionally, rather than randomly, creating a greater sense of ownership. 

John tracked the impact of this new arrangement on the Flow Metrics, and it seemed to have done the trick! Flow Metrics for Defects began to trend positively again, with velocity increasing, Flow Times getting shorter, and Flow Efficiency improving.  

And the best news of all: Employee Happiness was looking really good. eNPS scores shot up, reflecting the team’s renewed sense of purpose and engagement. 

Tasktop Viz shows eNPS scores rising (left) and Flow Velocity increasing (right) in March following the second experiment with maintenance team specialization

“The fact that we were doing so much to listen to the team and improve things for them, ultimately made them more open to sharing their ideas for further improvements. We had triggered a virtuous cycle of continuous improvement, owned by every single team member,” says John.  

More Ownership and Autonomy Equals Higher Engagement

“In the three months since we implemented these changes, everyone on the value stream has become more comfortable voicing their ideas about what we could try and where we could improve,” John sums up. “Most of those ideas have been about increasing autonomy and ownership. 

“The great thing is that Viz is always there to give us fast feedback on the impact of our experiments on flow and business results. If and when we make mistakes, we’re able to correct them quickly. In many ways, that defines the culture of our team. We’ve become a much more experimental group, able to take risks because we know we’ll be able to see the impact of our changes within just a few weeks.”  

More Blogs in the Tasty Dog Food Series

Taste Our Product For Yourself

This series of blogs has hopefully given you a real taste of what is possible with Tasktop’s VSM solutions. Why not take a personalized product tour to see where you can improve flow and your teams’ growth and happiness?


Get Software Delivery trends and insights in your inbox. Subscribe now.

Leave a Reply

Your email address will not be published. Required fields are marked *