To this day I remember my high school driver's education instructor (Mr. Rabinovitz) telling me, "There is no substitute for seeing." In that case he was talking about backing out of a driveway into traffic, but I have since found that his advice is true in so many other areas. That wisdom came back to me in a study we just completed of DevOps professionals.
We fielded a survey in advance of the Atlassian Summit in Las Vegas. We heard from 200 senior DevOps professionals at US-based companies. They were very much DevOps pros— the typical respondent reported that their organization employed 351 to 400 people who used DevOps to get their work done.
SmartDraw has spent the last 25 years helping the world visualize, well, just about everything. So, it comes as no surprise that we wanted to focus on visualization in this survey. But what does visualization for DevOps mean? Here is how we defined that term for the respondents:
The DevOps methodology makes use of a large ecosystem with hundreds of companies that help DevOps teams such as Atlassian, Jenkins, Puppet, Splunk and Basecamp. It can be difficult for DevOps to provide real-time visibility into development projects with so many disparate applications and ecosystem silos.
The reason it is difficult is that such visibility requires contextualizing the raw data that comes from all the disparate DevOps sources to create actionable insights. An example of contextualizing data would be taking raw input from your various DevOps tools to visualize dependencies between tasks:
To summarize, DevOps visibility is defined as:
- Pulling the data DevOps needs from a variety of DevOps tools such as Jira or Basecamp
- Transforming the data into actionable insights by understanding the complex data relationships that exist between the data
- Presenting these insights in the most intuitive manner (usually using a visual chart or diagram)
We wanted to know how interested these seasoned DevOps managers were in getting real-time visibility into their DevOps ecosystem. We started by asked them to tell us what their top goals were for their DevOps methodology.
Two of their top three goals were unsurprising: They wanted to further build DevOps expertise within their organization and how to spot which projects are (or are not) a good fit for DevOps. That makes sense. But the third goal struck a chord with us here at SmartDraw. 84 percent said it was somewhat to extremely important to "provide real-time DevOps visibility across your DevOps ecosystem."
How important? Well, we found that 100 percent were involved in trying to produce DevOps visibility. But why? The top reasons cited were:
- Improve DevOps efficiency
- Improve the quality of their software releases
- Make better decisions
What really interested us, though, is that there was a huge gulf between those DevOps professionals that were doing the best at providing visibility and those doing the worst.
To better understand this, we separated the survey data into three tiers. We scored each respondent on each response having to do with how well they were doing with either general DevOps goals or specific real-time DevOps visibility goals. Those whose aggregate score was in the top 33 percent we called "Top -Tier" and those in the bottom 33 percent we called the "Bottom-Tier."
What we found was the "top-tier" professionals were achieving significantly better responses, both in how they did at DevOps in general as well as how they did in terms of providing real-time DevOps visibility.
For example, 59 percent more top-tier respondents say they are doing somewhat to extremely well at achieving strategic and operational goals. 64 percent more top-tier respondents say they are doing somewhat to extremely well at providing a single operational view.
Future posts will dive into those fine-grained details, but the top-level findings of the 2019 SmartDraw DevOps Visualization Survey are simple to describe. There is no substitute for visibility. Thanks Mr. Rabinovitz!