Survey shows gap in incident management between DevOps and ITSM

A survey of 505 software developers, IT professionals, and IT decision makers (ITDMs) in the U.S. found that 70% of respondents work for companies where developers are responsible for deployments, while only 22% conduct non-blame-based post-mortems in practice.

The survey, conducted by Atlassian, also found that only 60% of respondents work for companies where developers are part of the team that handles IT incidents. 57% said developers are on call when needed. However, a total of 70% said it is easy to bring in the right teammates when needed.

Most respondents (86%) also conduct post-mortem or post-incident reviews (PIR) after an incident. A full 97% also report having incident management procedures, processes or runbooks in place, with 78% using war games or some other form of incident management training.

Overall, the survey shows that IT incidents are almost exclusively handled by an IT operations team (95%). In addition to developers, engineers (53%), managers (43%) and site reliability engineers (37%) are also involved.

However, the survey also found that only 15% of respondents have cross-functional teams with developers, operations staff and other IT professionals, and only 35% said their company views DevOps as a cultural shift that brings development and operations teams together. Only 30% emphasized the use of automation and continuous integration/continuous delivery (CI/CD) to streamline workflows, reduce manual processes and ensure faster and more reliable software delivery.

Furthermore, only a quarter (25%) associate DevOps with improving the speed and efficiency of software development, and even fewer (20%) focus on the role DevOps plays in improving software quality and reliability through automated testing, continuous monitoring, and real-time performance tracking. Only 15% said security and compliance are a top priority in their DevOps practices.

Yet 96% said development and operations teams have the visibility they need to do their jobs effectively and with minimal disruption. A full 70% said their organization can access incident history, recent deployments, or recent changes to see context while responding to incidents. However, only 55% have access to live information on service health.

More than two-thirds of respondents (68%) work for organizations that proactively manage IT incidents, the survey found. Almost all (99%) use monitoring tools, with 86% relying on them to detect incidents. Nearly three-quarters (73%) said they use these tools to proactively detect incidents before they are reported by end users or customers.

The most commonly used incident resolution tools, according to the survey, are capacity planning (80%), artificial intelligence (AI) to identify incident trends (74%), and monitoring user transactions (73%).

Kate Clavet, product marketing manager for IT service management (ITSM) at Atlassian, noted that as artificial intelligence (AI) advances, it will become easier to incorporate simulations into incident response training.

In terms of metrics tracked, most respondents track time to resolution (80%), while 71% track time to acknowledgement and 55% track time to response. Overall, 61% use tickets created by ITSM platforms as their only reliable source of information for managing incidents, while 39% rely on some sort of chat tool. A full 60% said they do not use a configuration management database (CMBD) to manage their IT environments. However, a full 97% said they have change management practices in place to minimize potential disruptions.

Despite some notorious outages recently, incident management seems to have reached an advanced level of maturity. However, in the absence of tighter integration between DevOps and ITSM teams, there is still a lot of room for improvement.