Working in a multitasking environment makes it challenging to deal with context switching on a daily basis, and developers are no exception. It’s well known that distractions are one of the biggest contributors to a developers’ reduced performance. GitHub found out that reducing work interruptions leveraged developers' daily progress from 7% to as much as 82%.
When juggling too many tasks engineers are unable to achieve the flow state directly related with Efficiency according to the SPACE framework. Short attention spans caused by constant context switching means higher probability of error, overlooking important information and producing more bugs. If not addressed during the code review, it may eventually lead to increased Change Failure Rate, one of the four core metrics in the DORA framework created by Microsoft.
GoodRx and Postman pay attention to the average amount of time lost when measuring developer experience. Identifying time lost for context switches necessary to combat obstacles in the work environment, their dev teams get an insight on where exactly to focus their efforts to improve productivity and time-to-delivery. Equipped with precise data, they have a deeper understanding of the root causes behind process bottlenecks and inefficiencies to make smarter decisions and deliver faster.
78% professionals asked by Atlassian said they couldn't get their work done due to too many meetings, which was very often leading to overtime work. Meeting overload hinders productivity of tech teams as well. For engineers at GitHub
going from 2 to 3 meetings per day lowered the chances to make progress toward daily goals from 74% to 14%. And developers who average just one meeting per day have a 99% chance of knocking out high quality daily work.
The more meetings slots with just short breaks in-between sit in your engineers’ calendars, the less they can actually deliver. Why? More fragmented coding slots make the iteration cycles longer and eventually, results in more time needed for high quality code to hit production. Constant context switching and not enough focus time hinders the understanding of how new tasks refer to the code already written, increasing the risk of redundancies and over-engineering which compound the technical debt.
How reducing meetings and promoting deep work can leverage productivity? Uber started measuring deep work streaks’ number and length across dev teams during the pandemics. Using anonymised real-time data about meetings and chats they’ve observed a 1.3x drop in focus time after turning to fully remote work. The data showed a collaborative overload “trap” in which people schedule and participate in more meetings to be more productive, but in turn lose focus time by reaching a disturbing number of short breaks - too short to actually get the work done. To address this issue, they have decided to organize team calendars by blocking a minimum of two consecutive hours for deep work and - as a result - observed a 20% boost in focus time directly correlated to long term productivity.
Tech industry leaders striving to improve engineering teams’ productivity with the Developer Experience (DevEx) approach are already tracking daily/weekly focus time per engineer as a top-level metric correlating directly with long-term and sustainable performance.
The first step taken by the majority is adding questions about self-reported qualitative measures on deep work time into their developer survey programs. The Developer Insights team at LinkedIn used a quarterly survey asking approximately 30 questions to get a deeper understanding of their developers’ needs. But in order to pulse-check changes and iterate on learnings faster, LinkedIn has also developed a real-time feedback system collecting quantitative data. It is actually a common approach across leading tech companies to use both qualitative and quantitative measures in order to get a full picture of what drives tech teams’ performance.
Quantitative metrics on for example time lost on context switching can easily speak to the business with precise estimations of monetary savings. As presented by Gergely Orosz from The Pragmatic Engineer:
For example, if an organization with $10M in engineering payroll costs reduces time loss from 20% to 10% through an initiative, that translates into $1M of savings.
Translating how cutting on unproductive meetings and deep work protection onto business value helps teams get buy-in and trigger initiatives on changes.
Leading engineering teams like Uber or Stripe pushed it even further by utilizing continuous feedback from anonymised data to track ie. “Number of Days with Sufficient Focus Time” and “Weekly Focus Time Per Engineer” and secure extra uninterrupted focus work hours for their dev teams each week. In Network Perspective we also strongly believe in leveraging real-time data from company’s systems to reduce inefficiency at scale in order to help tech teams achieve more with the same resources. We think about productivity holistically using AI algorithms to analyze data from calendars, chats and emails to inform teams’ not only about their deep work, meetings and context switching time, but also about their intra and cross team collaboration habits. As the key to effective productivity analytics is to do it ethically, we are processing metadata with full anonymization and present results on a team level for individual privacy protection.
Working with global champions like Adaptavist, we’ve proven that to drive actual change, data needs to be turned into knowledge, and knowledge into action. That is why data-informed insights encapsulated in the WorkSmart App are followed with carefully customizedsuggestions aimed at resolving each team’s inefficiencies. Having reliable ongoing feedback from data and actionable best practices on how to work smarter, 20k professionals already using the app are motivated to take small steps to protect their focus time and reduce interruptions. In the company-wide scale their efforts are adding up to a large and quickly observable impact by saving hundreds of hours monthly over the first 3 months.
And now, with another seed investment, we’re heading towards the next phase of transforming work efficiency by ie. building an AI workload assistant!
