Shifting Your DevOps Mindset from Tools to Culture
DevOps is based on continuous improvement and development, with employees as active participants in that iterative process. Enabling self-learning and self-improving among employees is central to this school of thought.
Real-time, data-driven IT decision making is challenging and, until recently, hasn’t been feasible. IT administrators have grown up in this “Tayloristic” work model — where operations, security, development, etc. are fragmented by tasks, goals and skills — resulting in siloed work styles and disparate definitions of success. Furthermore, traditional managerial structures don’t allow for creative thinking and self-starting. For this decision making approach to drive business forward, there must be cross-team collaboration and understanding. We are entering a new digital era where a shift in collaboration will define success. The driving force behind this collaboration is not only a shift in workflow, but more importantly in culture and mindset.
With this new philosophy, all teams work together on activities that impact the entire business, because they know how each team and activities affect the systems and services driving the organization as a whole. It’s about the general principles of working off the same data set, sharing tasks and empathy. It’s also very much about a cultural shift that needs to happen: teams working together to deliver services that customers want and drive business results.
Traditional IT work model
The namesake of the “Tayloristic” work model, mechanical engineer Frederick Winslow Taylor, sought to ultimately change the way people work to be more efficient. In his book “The Principles of Scientific Management,” Taylor urges companies to apply engineering principles to work being carried out on the factory floor. He argued that by reviewing employee performance through an engineering lens, management could ensure that workers completed tasks in the most efficient way possible. He thought by analyzing work, the "one best way" to do it would be found. However, he did not account for the possibility of more than one “best way,” especially across various teams, workflows and technologies.
Similarly, traditional IT was known for applying process to operations to make the daily workload more efficient. Given the process focus of IT, it wasn’t uncommon to have a manager that believed in singling out individuals when things don’t work as planned is good practice, then creating an environment where people put themselves on the line to try new methods will become near impossible. At one business I visited recently, I was told that if a problem was in your inbox when it was escalated to management, you were singled out for exacerbating the problem, not finding the solution. This “blame game” culture was so prevalent that no one ever wanted to own a production problem. Not surprisingly, instead of looking for solutions, the implicit goal for support teams for any incident became to get the ticket out of your inbox, rather than to solve the problem.
A culture of blame can quickly degrade all of the progress made in fostering a culture of openness, collaboration, innovation and accountability. I am seeing more and more companies moving away from this top-down enforcement model, simply because we are seeing clear benefits of working in team-oriented work environments. At one company I have worked with – a traditional 25+ year-old software company based in the US – they changed roles, team structures, and even physical spaces to encourage a more engaged and collaborative environment.
Rather than looking for someone to blame when something goes wrong, they hold "blameless postmortems" using shared data and tools, aimed at fixing the problem now, and preventing it from recurring later. As a result, they found substantial benefits, beyond faster releases, fewer errors, and faster time to resolution; but also better collaboration within and across teams, resulting in higher job satisfaction and staff retention.
IT complexity as an agent for change
Digital transformation is happening all around us. With it comes the complexities of new and segmenting IT innovations. Another organization I am working with – a large financial agency based in New York – is struggling here. It is facing pressure to accelerate delivery times for new solutions to compete with innovative financial businesses – aka "fintech" – but is hamstrung by too many siloed teams owning product delivery. While the fully aligned teams in emerging fintech businesses can plan, develop, and release new capabilities in hours, the non-aligned functional teams in this traditional business cannot approve a completed project for release in less than about four weeks. They have no chance to catch up, let alone keep up, if they retain legacy team structures.
[Check out 10 Tools for DevOps Collaboration]
These fundamental changes demand a new working model for IT and DevOps to stay connected, relevant and leading, rather than struggling to keep up. ITOps, DevOps and the rest of the business need to stay aligned on delivering critical services, systems and information needed to achieve and sustain critical business outcomes. In my conversations with customers, agility, speed, and quality come up as imperatives for digital transformation. But to achieve this, a cultural shift is required — employees need to feel empowered and enabled through leadership, not put down or stressed by redundant processes. How do we change the IT operating model to enable this cultural shift and bring about digital transformation?
The DevOps workflow, by definition, breaks down “Tayloristic” barriers to make way for more streamlined approaches. Whether you are an employee or a manager, working with your team is critical for activities that impact the business, because you each bring a different perspective to how the business operates. This is how one traditional European bank now operates, having swapped a structure with teams aligned horizontally by functional silo (dev, QA, ops) for teams that are aligned vertically by business service (online banking, lending services, etc.). This facilitated a ‘continuous delivery’ model for software as each team has all the specialists required to plan, build, verify, deploy and operate new technologies. Everyone shares common goals while executing on individual responsibilities.
Using this mentality, everyone has insight across all teams, is empowered to make needed changes, and is working on the same page, all while working toward the same goal. This is not unbridled power though - because workflows are open and outcomes are shared, each team and individual is responsible and accountable to each other, to management, and to the business. DevOps is based on continuous improvement and development, with employees as active participants in that iterative process. Enabling self-learning and self-improving amongst employees is central to this school of thought. Processes play a central role as a way to ensure consistent quality and remove redundancies, but the central thought process associated with shifting the cultural mindset is undoubtedly focused on empathy.
DevOps is mainly about people, so we must find ways to measure cultural change too. You can mine various applications for objective metrics, such as variance in team and individual productivity, number of out-of-hours callouts per week and employee timesheets, but metrics will always have subjective interpretations. A common data fabric allows all stakeholders to share visibility into DevOps processes, so they can more easily collaborate and communicate, with access to objective data for faster decision making and code delivery. However, this data is not the end-all-be-all. Shifting the team mindset from vertical to horizontal creates untapped opportunities for teams to succeed not only in a digital world, but also in their everyday lives.
Don't miss Andi Mann live and in person at Interop ITX. He'll be delivering a session on Effective Leadership for DevOps Teams and Programs. Register today for Interop ITX, taking place May 15 to 19 in Las Vegas.
About the Author
You May Also Like
2024 InformationWeek US IT Salary Report
May 29, 20242022 State of ITOps and SecOps
Jun 21, 2022