This article summarises findings from a DevOps maturity survey conducted jointly by Atlassian and xMatters, originally published on atlassian.com in April 2017. All statistics are drawn from that survey unless otherwise noted.
Overview
By 2017, DevOps had been influencing software and operations culture for roughly a decade. To understand how far the industry had actually come, Atlassian partnered with xMatters—a technology partner providing collaboration tooling for DevOps workflows—to survey practitioners across a range of industries. The resulting report highlighted meaningful progress as well as persistent gaps in three critical areas: Culture, Monitoring, and Incident Management.
DevOps Is Working—For Those Who Practice It
According to the survey, 41% of respondents reported actively practising DevOps. Of that group, more than 65% said their DevOps initiatives were delivering the benefits they had expected, including:
- Faster time to market
- Improved customer experience
- Fewer customer-facing incidents
- Faster incident resolution times
- A higher pace of innovation
The remaining 59% of respondents either did not know what DevOps was or were unsure whether their organisation was practising it. Those respondents were excluded from the detailed survey questions. This gap underscores a significant split in the industry: organisations that have embraced DevOps hold a measurable competitive advantage, while many others still treat it primarily as a buzzword.
For additional context, the 2016 Puppet State of DevOps Report noted that high-performing DevOps teams deploy code approximately 200 times more frequently than low performers—a figure that illustrates the scale of the performance gap between adopters and non-adopters.
Culture: Shared Tools, Not Shared Knowledge
Tool sharing between development and operations teams was the strongest indicator of cultural alignment found in the survey. 80% of companies reported that dev and ops share at least some tooling—a foundational step toward a genuine DevOps culture. Many of those organisations also shared information generated by those tools, though often with restricted access.
However, tool sharing alone does not equal a healthy DevOps culture. Genuine knowledge sharing—where cross-functional teams can access information dynamically and in real time—was far less common. Only 17% of companies reported having open information available to cross-functional teams in dynamic formats such as wikis and chat platforms. The large majority indicated that information is shared reactively and in static formats only when requested.
This gap matters because shared context is what allows teams to respond quickly to change. Without it, the cultural promise of DevOps remains only partially fulfilled, even in organisations with strong tooling adoption.
Monitoring: Extensive Data, but Still Missing Issues in Production
Organisations surveyed had invested heavily in monitoring. Key findings include:
- 75% monitored both infrastructure and applications/services
- 60% monitored transactions
- 51% monitored user experience
- Nearly two-thirds said their monitoring solutions could predict potential issues before users were affected
Despite this monitoring coverage, 50% of surveyed companies still reported finding issues in production after code was released. The likely cause is a combination of gaps in transaction and user-experience monitoring and the fact that barely half of surveyed organisations practised automated testing as part of their QA process.
The takeaway is straightforward: monitoring is necessary but not sufficient. Without automated testing integrated into the delivery pipeline, monitoring alerts act as a reactive safety net rather than a proactive quality gate. Organisations that combine broad monitoring with robust automated testing are better positioned to catch issues before they reach production.
Incident Management: Process and Information Bottlenecks
Incident management is often considered the most rigorous test of DevOps maturity. How quickly a team can detect, communicate, and resolve a major incident reveals the true state of its processes, tooling, and culture. The survey results showed that most organisations still had significant room for improvement:
- 50% of companies had to wait for the operations centre to formally declare a major incident before a response could begin
- 43% used manual processes to update customers and internal stakeholders during incidents
- 34% experienced delays in resolution while waiting for subject-matter experts to become available
The common thread across these bottlenecks is information and process overload. Monitoring tools generate large volumes of data, but translating that data into actionable decisions under pressure requires clear runbooks, automated alerting, and pre-defined escalation paths. Organisations that rely on ad-hoc communication and manual handoffs during incidents will consistently struggle to meet resolution time objectives.
Key Takeaways for DevOps Teams
The 2017 Atlassian/xMatters report painted a picture of an industry in transition. DevOps was clearly delivering value for its practitioners, but most organisations had not yet realised its full potential. Based on the findings, the most impactful improvements organisations could make were:
- Move from tool sharing to knowledge sharing. Adopt dynamic, open documentation platforms (wikis, persistent chat channels) so cross-functional teams can access context without waiting to be told.
- Pair monitoring with automated testing. Monitoring production is essential, but automated testing earlier in the pipeline prevents many issues from reaching production in the first place.
- Automate incident response workflows. Reduce reliance on manual communication and expert bottlenecks by implementing on-call automation, pre-defined escalation policies, and status-page tooling.
- Define what a major incident means before one occurs. Clear, agreed-upon criteria for declaring incidents remove the delay caused by waiting on operations centres to make a call.
DevOps practices were well-established enough by 2017 to show measurable competitive benefits for adopters. The challenge for the years ahead was—and in many organisations remains—moving beyond surface-level tool adoption to genuine cultural and process transformation.