8 min read

The State of DevOps report is a widely referenced body of DevOps research which helps organizations achieve high organizational performance and productivity. The DORA (DevOps Research and Assessment) team have published the 2019 Accelerate State of DevOps Report which is an independent review of the practices and capabilities of DevOps.

Teams across the globe can take advantage of the survey findings and identify specific capabilities to improve their software delivery performance.

Key findings in DevOps 2019 report

Increase in Elite performers

The DevOps 2019 report uses a cluster analysis method to identify the software delivery performance of the participants involved in the survey. All the respondents are categorized into four distinct groups called as Elite, High, Medium and Low Performers.

The report states, “In this approach, those in one group are statistically similar to each other and dissimilar from those in other groups, based on our performance behaviors of throughput and stability: deployment frequency, lead time, time to restore service, and change fail rate.”

According to the survey, the proportion of elite performers have jumped to 20% from the 7% last year. The percentage of medium performers have also increased, leading to a drop in the percentage of low performers. Hence, the cluster analysis method depicts a continued shift in the industry, as organizations continue to transform their technology.

Image Source: 2019 Accelerate State of DevOps Report

Read Also: Does it make sense to talk about DevOps engineers or DevOps tools?

Strategies for scaling DevOps in organizations

The Accelerate State of DevOps 2019 report have framed out several strategies for scaling DevOps in an organization. The strategies are created based on commonly used approaches observed by the DORA team across the industry.

Training Center (DOJO): Employees are taken out of their usual work routines to learn new tools or technologies. Next, they are required to implement the new learnt methods in their work and also inspire others to do so.

Center of Excellence: This strategy uses all the available skills for consultation purpose.

Proof of Concept but Stall: A central team is given the freedom to build in any possible way (often by breaking organizational norms). However, the effort stalls after the PoC.

Proof of Concept as a Template: It begins with the PoC but Stall project which is then replicated in other groups using the same pattern.

Proof of Concept as a Seed: This approach ensures that the PoC members stay active in the new groups indefinitely or just as long to ensure that the new practices are sustainable.

Communities of Practice: Groups sharing the same common interests in tooling, language, or methodologies are encouraged within an organization to share knowledge and expertise with each other and across teams.

Big Bang: When an entire organization is transformed as per DevOps methodologies at once.

Bottom-up or Grassroots: Small teams are put together to transform resources and share their success throughout the organization in an informal way.

Mashup: When an organization implements several of the above described approaches with partial execution or with insufficient resources.

The distribution of DevOps transformation strategies as per performance profiles are shown below.

Image Source: 2019 Accelerate State of DevOps Report

Cloud computing usage is the driving force behind elite performers

The DORA team uses the five essential characteristics of cloud computing, as defined by the National Institute of Standards and Technology (NIST) to understand cloud usage patterns by the categorized performers. The essential characteristics of cloud computing are given below.

On-demand self-service: DevOps users can use the cloud on an on-demand self-service premise to use the computing resources whenever needed, without any human intervention. This is the least used characteristic by users with only 57% of the survey respondents agreeing  on using it.

Broad network access: Such networks can be used through heterogeneous platforms such as mobile phones, tablets, laptops, and workstations. 60% of the survey respondents agreed on using the broad network access.

Resource pooling: The provider resources in a cloud are pooled in a multi-tenant model such that the physical and virtual resources are dynamically assigned on-demand. This allows the  customer to specify a location at a higher level of abstraction such as country, state, or datacenter. Among all the survey respondents, 58% of them agreed on using the resource pooling characteristics.

Rapid elasticity: The cloud capabilities are elastically provisioned to the extent that it can be released rapidly to scale outward or inward on demand. Hence, the cloud capabilities seem to be unlimited and can be appropriated at any point of time in any quantity. When compared to 2018, this characteristic saw a growth of +15% this year, with 58% of the survey respondents utilizing it.

Measured service: This is the most used characteristic of cloud computing as 62% of the respondents agreed on using the cloud measured service. It allows the cloud systems to be automatically used to control, optimize, and report resource usage based on the type of services like storage, processing, bandwidth, and active user accounts.

The DevOps 2019 survey found that only 29% of all the respondents agreed on utilizing all the essential cloud computing characteristics. The survey also revealed that the elite performers in the survey used the cloud characteristics 24% times more than the low performers.

Importance of psychological safety to SDO performance

The DevOps survey found that a culture of psychological safety where team members feel safe to take risks and be vulnerable in front of each other are able to better deliver software delivery performance, organizational performance, and high productivity. This enables an organization to come up with new products and features without impacting the existing users, hence leading to desirable levels of software delivery and operational performance (SDO performance). The report concludes that due to psychological safety elite performers can twice achieve or exceed their organizational performance goals.

Size of an industry does not correspond to its success

This year, the retail industry saw significantly better SDO performance, in terms of speed and stability of software delivery. However, the DevOps report states that, “We found no evidence that industry has an impact with the exception of retail, suggesting that organizations of all types and sizes, including highly regulated industries such as financial services and government, can achieve high levels of performance.” The report suggests that size is not a factor for poor performance of other industries and high levels of performance can be achieved by adopting DevOps practices.

Read Also: Listen: Puppet’s VP of Ecosystem Engineering Nigel Kersten talks about key DevOps challenges [Podcast]

Proposed steps to improve performance and productivity in DevOps

The Accelerate State of DevOps Report of 2019 proposes two research models to improve DevOps performance and productivity.

Performance model

The DevOps report states, “A key goal in digital transformation is optimizing software delivery performance.” To improve software delivery and operational (SDO) performance and organizational performance, teams can first start with basic automation such as version control and automated testing, monitoring, clear change approval processes, and a healthy culture. The DevOps 2019 survey finds that low performers use more proprietary software than high and elite performers.

Image Source: 2019 Accelerate State of DevOps Report

Productivity model

The report defines productivity as “the ability to get complex, time-consuming tasks completed with minimal distractions and interruptions.” Teams can use the useful and easy-to-use tools along with internal and external search to accelerate productivity. The DevOps survey reveals that the highest performing engineers are 1.5 times more likely to use the easy-to-use tools. An improved amount of productivity also helps employees have a better work/life balance and less burnout.

Image Source: 2019 Accelerate State of DevOps Report

DevOps teams can use the above two models to locate their goal, identify the dependent factors and increase their overall performance and productivity output.

Demographic makeup of the DevOps 2019 survey

This year, the DORA survey had almost 1,000² participants. Among all the responses, 26% responses came from employees working in very large companies (10,000+). Compared to last year, there has been a drop in response from employees working in companies with 500-1,999 employees. In contrast, more responses have been received from people working in company sizes of 100-499 employees.

The majority of participants worked in the Development or Engineering department. The DevOps or Site Reliability Engineering (SRE) department came next, followed by Manager, IT Operations or Infrastructure, and others. Half of the participants in the 2019 research are from North America, followed by EU/ UK at 29%. This year also saw a fall in responses from Asia, which is only 9% when compared to the 18% last year.

The percentage of women on teams have also reduced to 16% (median) from the 25% reported last year.

Image Source: 2019 Accelerate State of DevOps Report

Developers across the world love the Accelerate State of DevOps 2019 report and are thanking the DORA team for major takeaways like cloud being a key differentiator, how to be a elite performer in software delivery performance, and more.

Here’s a two minute summary video of the Accelerate State of DevOps 2019 report, published by Google Cloud.

Interested readers can check out the report to see a detailed comparison of tool usage, according to low, medium, high, and elite profiles. Also, you can read the full 2019 Accelerate State of DevOps Report for more information.

Read Next

5 reasons poor communication can sink DevSecOps

7 crucial DevOps metrics that you need to track

Introducing kdevops, a modern DevOps framework for Linux kernel development

A born storyteller turned writer!