This article is based on a GitHub data science report published in early 2020 and originally summarized by Dice Insights. The original GitHub report analyzed platform activity data from Q1 2020.
Overview
As COVID-19 forced millions of knowledge workers into remote work arrangements in early 2020, GitHub’s data science team examined how the shift was affecting developer behavior on the platform. Their analysis of activity during the first quarter of 2020 revealed patterns that have significant implications for engineering teams, engineering managers, and DevOps practitioners thinking about sustainable delivery practices.
Overall Activity Held Steady — or Increased
Despite the disruption caused by shelter-in-place orders and sudden transitions to remote work, GitHub’s data showed that developer activity during the first three months of 2020 was “largely consistent or increased” compared to the same period in 2019. On the surface, this might suggest business as usual. But a deeper look at the data reveals important nuances in how that work was getting done.
Enterprise Issue Activity Tracked Outbreak Timelines
One of the more striking findings was the correlation between COVID-19 outbreak timelines and GitHub issue activity in enterprise repositories. According to the report:
“Our analysis shows that GitHub issues in enterprise repositories rise and fall around COVID-19 outbreaks and shelter-in-place orders. The flux seen in issues is likely due to the move to distributed work, which disrupted the ceremony, structure, and coordination of enterprise software development. On GitHub, that coordination is often done with issues, where teams track enhancements, tasks, and bugs.”
This is a meaningful signal for DevOps teams. Issue trackers serve as a coordination layer for distributed teams. Spikes and drops in issue activity can indicate teams struggling to establish new remote communication norms — or, conversely, teams adapting to async collaboration tools more effectively over time.
Work Days Got Longer
GitHub’s analysis also found that developers’ working days had grown — by up to an hour per day, on both weekdays and weekends. The report acknowledged that longer days are not purely a measure of increased productivity:
“Longer days may also be a result of non-work interruptions (like family or childcare) now that many are working from home.”
This is a critical distinction. Raw commit volume or hours logged does not necessarily translate to productive engineering output. Interruptions from caregiving responsibilities, anxiety about the pandemic, and blurred home-office boundaries all fragment concentration and inflate wall-clock time without proportionally increasing meaningful work.
The Pressure Behind the Numbers
GitHub’s report identified several factors that may have been driving the increased push frequency observed in the data, particularly in March 2020:
- Economic uncertainty: Fear of layoffs or reduced hours may have prompted developers to signal productivity through higher commit volume.
- Work as distraction: Staying busy with coding as a coping mechanism during an anxiety-inducing lockdown period.
- Management pressure: Organizational urgency to ship products in a volatile market.
- Team delivery norms: Expectations around frequent commits as a marker of a healthy, fast CI/CD pipeline.
Understanding these motivations matters for engineering leadership. If increased commit frequency is driven by fear rather than improved process, it can mask underlying team health problems and inflate metrics that teams use to evaluate delivery performance.
Burnout Was Already Rising
GitHub’s findings aligned with concurrent survey data on burnout in the technology workforce. A survey by Blind conducted around the same period found that 73 percent of technologists reported burnout by the end of April 2020, up from 61 percent in mid-February. Among those reporting burnout:
- Approximately 20.5 percent attributed it to an unmanageable workload.
- Approximately 19 percent attributed it to concerns about job security.
It is worth noting that remote work and burnout had a documented relationship even before the pandemic. A 2019 DigitalOcean survey found that 66 percent of technologists working from home reported burnout, compared to 64 percent who worked regularly in an office — a gap that, while small, pointed to additional stressors unique to remote work arrangements.
GitHub’s Own Caution: Longer Hours May Not Be Sustainable
The GitHub report itself included a direct warning about the sustainability of the trend:
“If additional work is happening at the expense of personal time and breaks to replenish, ponder, and maintain healthy separation, we caution this tradeoff may not be sustainable in the long run.”
This is a point well worth heeding for engineering managers designing team processes during distributed work. Delivery velocity metrics like deployment frequency and commit rate can improve in the short term even as engineer wellbeing deteriorates — until attrition, quality issues, or health events force a correction.
Practical Takeaways for DevOps and Engineering Teams
While the GitHub report was a snapshot of early 2020 conditions, the questions it raises remain relevant for any team managing distributed engineering work:
- Monitor leading indicators of burnout, not just delivery metrics. Cycle time, PR review lag, and unplanned work can surface team stress before it becomes a retention problem.
- Distinguish coordination overhead from productivity loss. A spike in GitHub issues may reflect teams adjusting to async communication, not failing to deliver.
- Protect psychological safety. If engineers feel that longer hours are expected or that commit frequency signals job security, engineering culture — not individual discipline — is the thing to address.
- Re-evaluate meeting and ceremony cadences for distributed teams. The pandemic forced many teams to redesign standups, retrospectives, and planning sessions; those lessons remain applicable to any remote-first or hybrid team.
- Use DORA metrics thoughtfully. Deployment frequency and change failure rate are better health indicators than raw commit volume or hours online.
Context and Limitations
GitHub’s analysis covered only the first quarter of 2020 — the very beginning of pandemic-era disruption. At that point, many businesses in the United States and elsewhere were only just implementing work-from-home policies. The longer-term effects on developer workflows, team culture, and software delivery quality would only become visible over subsequent quarters.
Additionally, GitHub’s data reflects activity on its platform specifically and may over-represent open source contributors and enterprise teams using GitHub as their primary development platform. Teams using GitLab, Bitbucket, or internal source control systems would not appear in this data.
Nonetheless, the 2020 GitHub analysis remains a useful reference point for understanding how large-scale external disruptions propagate into observable patterns in software development activity — and why engineering leaders need measurement strategies that go beyond raw throughput.