The ever-evolving need of tech-savvy end consumers to explore and experience new product options and innovations has affected brand loyalty. It became necessary for large organisations to launch new offers or product changes in the market quickly to attract and retain customers. And, in agile methodology, they saw a possible way out of these challenges. Although agile adoption definitely helped these large organisations gain early time-to-market benefits, the next set of challenges appeared quickly enough.
Apart from changing customer demand, large brands had to face competition from nimbler organisations that were able to implement changes to systems with greater frequency and ease. With no legacy baggage, modern technologies, and extreme automation, new-age digital-native opponents pose tough competition to large organisations, which earlier enjoyed a monopoly. The speed at which the new-age organisations delivered brought about a fundamental disruption in the marketplace and further enforced changes to the buying patterns of end consumers. Large organisations realized that while the benefits of agile methodology were limited, those could be enhanced if they adopted DevOps engineering automation to increase their speed to market.
While DevOps certainly instilled a culture of automation, it did not disrupt business as usual. It complemented the IT systems well with the support of available tools (commercial or open source) in the market. Some organisations became early adopters about a decade ago, while others were forced into it later when it became necessary.
DevOps pilots proved successful as adoption was easier for web technologies, and benefits were realized quickly. However, three primary hurdles persisted and needed to be overcome before these large organisations could become as nimble as their new-age digital competitors.
Hurdles to adoption
Challenge 1: Scalability and uniformity of DevOps adoption
Firstly, large organisations were challenged by how to adopt DevOps across the board. Typically, these organisations have multiple stand-alone lines of business (LOB) with different stakeholders. Each LOB differed in its choice of technology regarding package and coding language and portfolio composition – legacy or digital. Often, they adopted different strategies to define coding standards or automation as governed by their respective set of enterprise architects. As a result, building a standard DevOps solution, which could cater to diverse technologies and be adopted uniformly across the organisation, remained a challenge. Added to these technical challenges was the fear of change, or rather, a mindset that resisted change. This resulted in federated DevOps adoption in each LOB leading to higher investment in DevOps tooling, strategy definition, and hardware, along with effort duplication by multiple teams to build DevOps solutions.
Challenge 2: Continuous evolution and management of DevOps
The second challenge for large organisations was the continuous evolution of DevOps tooling and practices. Better tools or newer versions of tools are now available. Certain technologies, such as packages, started supplying their own tooling to ensure end-to-end DevOps within the package. DevOps practices are adopting more life cycle stages in CICD pipelines like security testing. Managing tooling changes, currency upgrades, and enhancing DevOps practices across thousands of DevOps pipelines has become a mammoth maintenance effort.
Challenge 3: People and process transformation
The third challenge that accompanied automation was that it changed the way teams operated. High-speed teams enabled with DevOps engineering automation could not work in silos such as development, testing, and operations. Automation enabled faster releases which meant frequent handovers, and it became difficult when separate teams worked across these functions. An added cultural change was needed when teams were to be merged and work as a single, harmonious unit. Combining such teams added process challenges as the disparate processes such as scrum for development and ITIL for operations could not be continued within the unified team structure. Teams needed unified processes such as Scrumban to successfully transition into a single team, with each individual serving cross-functional roles. Over time, the challenges to DevOps adoption that initially pertained to technology and engineering automation became the challenges of the process and people transformation too. Today, the problem is aggravated with newer ways of working like Site Reliability Engineering (SRE).
The journey turned out to be much slower for large organisations owing to multiple factors such as the size of operations, vast range of technologies, various stakeholders, and complex organisational structures. In contrast, adopting DevOps in its entirety was easier for new-age digital companies with modern technologies and smaller teams.
The Road to Mainstream DevOps
These challenges are now things of the past. Today, entire digital portfolios of large organisations are almost entirely transformed with DevOps engineering automation and a steady wave of people and process transformation by replacing multiple teams with a single, unified DevOps team.
Centralized DevOps solutions that solve scalability and uniformity challenges are also helping organizations course correct their previous federated DevOps adoption approach. Such solutions reduce both the implementation and maintenance costs of DevOps. Further, cloud migration journeys of organisations use DevOps as a key lever to improve cloud benefits. There is also growing interest in adopting security in DevOps (DevSecOps) and DevOps for packages such as CRM and SaaS to achieve DevOps at the value stream level.
This transformation is the result of a cultural shift in organisations in recognizing DevOps as a necessary, mainstream adoption as opposed to being a ‘nice to have’ choice. Enterprise-scale DevOps adoption patterns such as DevOps as a service and DevOps as a platform may not be fully centralized, but they are definitely less federated. These adoption patterns have also eased the challenges, including wide technology spread, multiple siloed units, and too many stakeholders.
Needless to say, DevOps has become mainstream and an integral part of the software development lifecycle. Over the years, DevOps has proved indispensable for businesses to survive competition amid ever-changing market dynamics and end consumer demands. Evolving technology, cultural transformation, and stakeholder buy-ins have ensured there are no more hurdles in adopting DevOps at an enterprise scale, even for the large ones.