Readiness a Key Driver of Enterprise DevOps in Age of Pandemic

The 451 Take

Ready, Set, DevOps

The current pandemic has instantly offered real-world examples of how critical it can be for an organization to be agile, responsive and proactive. Some manufacturers are reworking designs and processes to produce medical equipment. Restaurants and retail chains large and small must now try to accommodate take-out and delivery only. Sometimes an event like the current pandemic can act as a catalyst for market disruption and the need for digital transformation. In the case of automobile dealers, for example, organizations were already supporting or competing with the ability to buy or lease a car almost entirely online without visiting the dealership. That ability has suddenly become not only one part of the market, but essentially the whole market for the time being. Longer-lasting effects are also likely as highlighted by our recent VotE Corona Virus flash survey, which indicates that a sizable number of enterprise organizations expect some of the changes that have arisen from the pandemic to remain in place long-term or become permanent.

According to the VotE: DevOps 2H 2019 survey, complexity is the overall biggest challenge to deploying DevOps, followed by cost and concerns about governance, security and compliance (see Figure 2).

Organizations that have had a DevOps strategy in place for five years or longer are more likely to be concerned about governance, security and compliance risks compared with organizations that have been executing a DevOps strategy for less than three years. These results highlight that early DevOps implementations, which are often limited to pilot projects and testing and development, are less likely to encounter production-release concerns such as governance, security and compliance.

Our research also indicates that many organizations have fairly equal amounts of IT automation and manual process (35%), while nearly a third (32%) have mostly automated processes with some manual processes. Digital transformation leaders are far more likely to have mostly automated (38%) and completely automated (7%) IT processes compared with digital transformation laggards (mostly automated at 15% and completely automated at 1%).

The Business and Management of Enterprise DevOps

We continue to track the stakeholder spread of DevOps, whereby leadership, security, data analytics and other teams are pulled into DevOps as it broadens to more of the organization. Our VotE: DevOps 2H 2019 survey results provide further evidence of top-down DevOps adoption involving management and leadership, which was ranked as the top additional stakeholder beyond developers and IT operators. We also see evidence of more sanctioned DevOps implementations involving central IT administrators, which are ranked as the second-most-important stakeholders.

Our survey indicated that most enterprises are combining business units and DevOps teams to deploy and spread DevOps processes. Organizations with a DevOps strategy for five years or more are much more likely to say both the business units and DevOps teams report to corporate management (52%) and that the DevOps teams drive business unit initiatives (41%). Among organizations with a DevOps strategy in place for less than three years, both of these are the case for only 30%. This may indicate that the longer a DevOps strategy is in place, the higher the likelihood that company management and leadership are overseeing it, rather than business units. It may also show that longer DevOps implementations translate into DevOps teams running more of the show, including business unit initiatives.

Enterprises continue to measure DevOps success by both technical metrics such as quality and performance, and business metrics such as customer satisfaction. There is a nearly equal share of responses for software quality (44%) and application performance metrics (41%), as well as business-level metrics such as customer satisfaction and user experience (41%) measuring and proving success.

Crossover with Data Analytics and Security

Our survey also highlighted the growing crossover of DevOps with data analytics and security. Data processing, analytics and business intelligence represents the top application type for DevOps and also the top technology trend deemed important to DevOps (see Figure 3). Public cloud and automation were also deemed important technology trends for DevOps, and we expect both to grow alongside newer and broader DevOps deployments.

The DevSecOps trend, whereby security tooling is integrated into DevOps workflows, continues to evolve, according to our research. Security people and teams are also ranked among primary stakeholders beyond developers and IT operations. Organizations view a range of security elements as critical to DevOps. Vulnerability assessment seems to be the baseline for security in DevOps for most organizations, followed by integration of additional security elements such as dynamic analysis, software supply chain validation, SCA and interactive analysis. We anticipate more elements will increasingly be included in enterprise DevOps releases.

Conclusion and Outlook

Going forward, we expect 'flexibility to quickly respond to changes' may become the main driver and benefit of DevOps given the current state of the world. Efficiency and managing larger-scale infrastructures and traffic spikes with smaller teams will likely also remain a top priority for enterprises amid the current pandemic. Despite changes in the world, market and the DevOps trend itself, speed will no doubt continue to be a significant driver and advantage of DevOps. We expect software releases will increasingly be daily and weekly as monthly and quarterly releases diminish. As we’ve seen with previous economic downturns, we also expect to see more open source used in enterprise IT in general, including DevOps. It may not represent more paid use of open source software, but we expect enterprise teams will increasingly be looking to open source, particularly cloud-native software such as Docker containers, Kubernetes orchestration, Istio and Linkerd service mesh, and Knative serverless.