Traditional IT infrastructure management methods can stymie the rapid delivery of modern solutions. Agile methodologies can help improve efficiency, speed, and quality.
Many businesses have sped up application development by implementing agile principles and current software engineering best practises like automated testing. Despite the fact that applying these methodologies and tools to IT infrastructure and operations can enhance productivity and speed the time it takes to bring digital products and services to market, it is still unusual. The typical IT infrastructure company continues to place a premium on reliability over speed. Requests for infrastructure services are still frequently handled in an assembly line fashion, with numerous handoffs, considerable delays, and frequent misunderstandings.
Good Read: Embracing Agile: Agile Is More Than Sprinting
It is not easy to modernise an IT infrastructure organisation using agile transformation, but it is worthwhile. Depending on the scale of the business, agile methodologies can help IT infrastructure groups increase productivity by 25 to 30 percent in six to 18 months. As more automated solutions are developed and used, the benefits could grow even more. Improved infrastructure service delivery and faster time to market for digital products and features are common additional benefits. In this post, we’ll look at how infrastructure firms can modernise using agile approaches, starting with a look at three different companies’ experiences. We also look at the four shifts mentioned above, as well as practical suggestions for getting the transition started.
Three ways to make the switch to a more agile IT infrastructure
Three corporations showed how diverse agile transformation techniques, based on similar principles and adjusted to their specific needs, can help update their IT infrastructure organisations while considerably enhancing performance. A global infrastructure team of several thousand individuals oversaw a global footprint capable of managing millions of active users and thousands of log-ins per second at a multinational software and service provider. As the company evolved, the methods it employed to supply infrastructure services became more sophisticated and labour-intensive, so bringing new products and features to market could take months.
When the company’s IT infrastructure leaders modelled the benefits of implementing agile principles in their organisation, they saw a 20 to 25 percent increase in productivity in 12 to 18 months. Given the size of the infrastructure group, the leadership team decided to roll out agile ways of working in stages throughout that time, forming approximately 150 agile teams to introduce new methods and technology to the entire firm. Teams were directed to focus first on enhancing the infrastructure department’s internal operations by simplifying and automating processes, then on establishing self-service tools and application programming interfaces (APIs) that could be used more widely.
Traditional infrastructure-building and management processes were slowing the release of digital products and services, as well as the adoption of more efficient, sophisticated application-development practises and tools, according to a European financial services company with a much smaller IT infrastructure organisation. Within its own business, this company set out to embrace agile processes and build highly automated infrastructure service offerings. However, unlike the software-and-services company, it took the strategy of implementing a new agile operating model throughout the whole infrastructure group at once rather than iteratively. The business also decided to focus from the beginning on developing an operating model and tools that would allow developers to manage their applications’ operations directly.
Another company, a significant financial services firm situated in the United States, used an agile strategy in its 250-person IT infrastructure and operations group. It, like the European financial services firm, implemented a new agile operating model throughout its entire infrastructure group at the same time. The corporation, on the other hand, elected to focus on upgrading its procedures first. It accomplished a transition in six months that resulted in a 35 percent reduction in IT costs and a twofold increase in total productivity. Following the implementation of the new operating model, the company intends to automate up to 80% of its operations activities.
Agile transformation principles
Despite their disparities in transformation strategies, these businesses shared many of the same fundamentals. We’ll look at those concepts in four areas in the sections below: technology, organisation and personnel, processes, and collaboration with developers (exhibit).
| Traditional Organization | Agile Organization | |
| Technology | Highly Customized Infrastructure, Provisioned on requestSignificant manual effort required from infrastructure teams | 1.Standardized Infrastructure service offerings with largely automated delivery. 2.Self service tools let application developers control and configure infrastructure on their own with appropriate guardrails. |
| Organization and Talent | 1.Technology – or – function specific teams 2.Staff with highly specialized skill sets focussed on operations and administrations | 1.Integrated, cross-functional teams ( or squads) build well -defined infrastructure service offerings. 2.Infrastructure engineers with sophisticated development skills |
| Processes | Rigidly sequenced processes, with many handoffs among group of specialistsRepetitive tasks( such as deployment and incident resolution) performed manually. | 1.Squads responsible for end -to-end service offerings 2.Processes in which repetitive work is automated and streamlined |
| Collaboration with Application Development | 1.Infrastructure requests submitted as “tickets” 2. Relationship and service managers deal primarily with application developments on behalf of the infrastructure function 3.Developers not accountable for application code after it is put into production | 1.Application development and infrastructure operations responsibilities become more integrated 2.Self service tools let developers and users handle more operations directly. |
Technology: Using software to define it infrastructure
Traditional infrastructure organisations are slow because their technological platforms necessitate teams manually configuring infrastructure for each new application. Companies can increase infrastructure agility by establishing automated systems that allow infrastructure to be defined by software, as well as providing “guardrails” that allow application-development teams to safely manage more of their own operations. While such systems can be built using existing infrastructure, automation becomes easier as a corporation migrates more of its infrastructure to modern platforms, particularly cloud platforms that provide a wide range of enabling tools and technologies.
Even though the infrastructure team at the software-and-services firm had standardised most of the hardware and virtualization architecture, it still spent a lot of time configuring custom virtual machines and operating systems for product development teams. Solution engineers met with the developers of each application to discuss their requirements, and then built up the requisite environments, which typically required numerous manual processes.
Rapid infrastructure teams deployed automated solutions to streamline server provisioning and setup as part of the company’s agile transformation. One agile team created and maintained a centralised platform that automated server provisioning and was accessible via self-service tools. Other agile infrastructure teams, each aligned with unique software-as-a-service (SaaS) products, used a configuration-management tool to specify the servers’ configurations fully in code, automating the configuration of those servers for the products they supported. This improvement cut environment build times in half, from several months to roughly 10 minutes. After these solutions were established, teams could make the necessary modifications quickly, with low manual work and risk of error, whenever a cluster of servers needed to be upgraded or expanded.
Using similar technologies, the European financial services industry sought to automate its IT infrastructure offerings. It aimed to empower application developers to handle their own operations as much as possible as part of a larger push to adopt DevOps principles. Application developers were trained how to use automated, self-service infrastructure solutions built by IT infrastructure teams. Developers might then write code to inform the system how to configure or upgrade servers based on their applications’ specific requirements.
Building cross-functional teams requires organisation and talent.
Infrastructure organisations have long been organised around teams with strictly defined responsibilities for specific technical functions (such as administering relational databases or operating systems) or stages of the plan-build-run IT service life cycle at traditional firms. Because many teams must often work on each service request, neither this structure nor the specialisation it promotes is favourable to efficiency or agility. Infrastructure firms can arrange their employees into small cross-functional teams focused on providing well-defined services to become more agile. They can also build modern workforces of well-rounded engineers that can quickly pick up new skills and operate across many functional domains to supply infrastructure services end-to-end, as we’ll see below.
Good Read: Confession : Cross Functional Teams or specialized teams ?
Engineers in agile infrastructure organisations usually require more broad skill sets than application developers, therefore CIOs and technology leaders should keep this in mind. This makes agile infrastructure modifications more difficult. By carefully vetting current workers, the infrastructure organisation at the European financial services corporation was able to find some of the well-rounded infrastructure engineers it required. The most capable were assigned to infrastructure squads in charge of developing the above-mentioned highly automated self-service solutions.
To help with capability building, professional development, and standard setting, the executives of the infrastructure division at the software-and-services firm decided to arrange their employees into skill-focused “chapters.” Chapter leaders were given the task of determining which new skill sets their areas required and developing training or hiring programmes to address those needs. The organisation divided everyone from those chapters into two types of cross-functional agile squads, each led by a product owner who identified and prioritised the backlog of activities that their squads would work on. Infrastructure teams concentrated on providing highly automated basic infrastructure solutions (such as server provisioning) that could be utilised by other teams to set up, manage, and decommission infrastructure. Product squads worked closely with SaaS product development teams to create and automate hosting and operations for their apps, utilising infrastructure squads’ services when they were available.
Processes: Minimize delays by simplifying and unifying activities.
The functionally oriented structure of a typical IT infrastructure organisation imposes a specific working style: specialised resources accomplish tasks in a predetermined order, with numerous handoffs between groups. This method of working causes several delays: whenever a request is forwarded to a new group, it is pushed to the bottom of that group’s task list, where it may languish for days. Tasks are frequently referred back to prior groups for clarification, lengthening wait times even more.
Good Read: KANBAN in IT Operations
Many of these delays can be avoided by forming small cross-functional teams, as mentioned in the previous section. By overseeing the end-to-end delivery of certain service offerings, such teams can reduce or even eliminate process handoffs. They should be given the authority to not only deliver service offerings, but also to improve them by streamlining processes and developing fully automated solutions.
As the software-and-services company evolved and introduced new customer-facing products, the infrastructure group’s processes became increasingly complex. As a result, project coordinators were hired to help rush service requests through the system. Waiting periods that had previously preceded handoffs between functional groups evaporated as the corporation organised its infrastructure engineers into agile squads. That modification alone cut the time it took to provide several essential service offerings in half. Common processes, such as approvals, were also overhauled by the company’s teams to simplify workflows or eliminate superfluous stages. For example, the number of processes in virtual server provisioning was reduced by more than two-thirds, and the remaining tasks were mainly automated thanks to better engineering.
To compensate for its infrastructure organization’s low development skills, the US-based financial services corporation mentioned previously took a different method. To begin, it formed cross-functional teams to streamline operations without resorting to automation. Employees were given ample time to develop more advanced engineering skills as a result of the increased productivity. Then they started thinking about how to build automated capabilities to handle frequent requests.
Fostering understanding and accountability through collaboration with application development .The contact between traditional infrastructure groups and application development teams is low.
The initial setup of systems for new applications and the resolution of catastrophic problems are usually the only areas where the two camps collaborate. As a result, conventional infrastructure engineers know much too little about the applications they support to assist in improving their stability. Furthermore, developers lack the operational expertise required to design strong, easy-to-support apps. In contrast, modern agile organisations make a point of enhancing communication between their application development and infrastructure departments.
One cooperation strategy, as exemplified by the European financial services organisation mentioned earlier, is to hold developers accountable for the operation of their apps. Developers become more aware of flaws in their application code when they are involved in the incident-response and post-outage follow-up processes for their applications. Developers that are involved in operations are more likely to write code that is easy to manage and support, as they may be woken up in the middle of the night if an event occurs.
A large software and services company takes a different strategy. Its infrastructure team continued to assist application-development teams’ operations, but in a different way: by tightly aligning agile product squads with application-development teams. Coordination and collaboration were considerably improved as a result of the alignment. Many of the product squads shared space with the application development teams with whom they collaborated, at least in part. Some of the agile ceremonies of the application-development teams would be attended by core members of each product squad. Furthermore, the close alignment allowed infrastructure engineers to get more familiar with the applications they controlled, allowing them to feel more invested in their performance, which could now be better monitored and supported.
An agile approach to infrastructure transformation
Using a disciplined approach to designing, deploying, monitoring, and enabling agile teams, we’ve found that the obstacles of updating IT infrastructure with agile can be overcome. (Applying this method in waves at larger firms can assist the change get off to a faster start.) This can be useful as part of a larger attempt to adopt agile principles to restructure a corporation, or as a stand-alone effort focused on the IT infrastructure division. In any case, the following are the important steps in building an agile transformation of an IT infrastructure function.
1. Develop a vision for the new infrastructure organisation, including how it should function and how quickly it should evolve.
Several essential questions will aid IT and business leaders in defining the organization’s vision.
What kinds of infrastructure services should the company give to app developers and business users? Establishing a catalogue of infrastructure service offerings assists businesses in designing and defining the scope of agile teams, as well as determining who should be in charge of delivering and upgrading those services.
What is the best way for the infrastructure organisation to communicate with application developers, and how should the interaction model change over time? If the infrastructure organisation is responsible for running applications, having teams that are tightly aligned with application development teams might be useful (for example, deploying code).
How soon should the company work on developing automated solutions and implementing cloud technologies? Agile teams that focus on operations may have quite different structures, methods, and skills than agile teams that focus on building infrastructure services.
2. Segment and rank opportunities based on their ability to add value to the company.
It is critical to assess infrastructure demand by gaining a data-driven understanding of past consumption patterns and anticipated future requirements. Knowing how much effort goes into delivering various infrastructure solutions aids in structuring the work into scopes that an agile team can handle. If, for example, the demand for storage-related work necessitates a workforce of 24 individuals, which is too large for a single team, the effort could be split between two teams, one focused on block storage and the other on file storage services.
Demand analysis can also aid in identifying the most promising areas for efficiency improvement and prioritising the deployment of teams. For example, a corporation might get a lot of value out of a transformation by assigning the first agile infrastructure teams to manage and improve labour-intensive services that are used frequently.
How will infrastructure officials and corporate executives assess the transformation’s effectiveness? Business and IT leaders should identify specific targets for enhancing performance and value generation before embarking on an agile infrastructure transformation, so they can track progress and outcomes using well-defined metrics.
3. Create each agile infrastructure team with a focus that matches the working style it will employ.
The teams that work on developing automated infrastructure service offerings are typically modest, ranging from eight to twelve persons. They frequently find that the scrum methodology works best for them, with solutions being developed in two- to three-week development sprints. Longer rosters of up to a couple of dozen employees may benefit teams focused mostly on operations (such as level-one support teams). These teams frequently employ kanban or scrumban techniques, which are better suited to managing an unexpected or event-driven work flow.
It is frequently desirable in the long run to have the same infrastructure team responsible for both planned development and unanticipated operating work for a certain service. This method encourages teams to identify and resolve operational concerns. Separating unplanned operational tasks from newly established infrastructure teams, on the other hand, can assist newly established infrastructure teams focus on developing highly automated solutions at the start of an agile transition.
4. Establish a method for deploying agile infrastructure teams.
Everyone involved in the process should have ample time to prepare for the launch of their teams. Our experience has shown that giving time and advice to train team members, design a good team charter, align key stakeholders, and create an initial backlog is crucial.
Before each agile squad debuted, the product owner and scrum master received two days of role training on how to perform their new duties at the software-and-services company, for example. They then participated in a six-week self-organized programme led by agile coaches, during which they developed their teams’ vision, scope, objectives, performance indicators, minimal viable product for improved delivery, and composition. Product owners had to identify their important stakeholders up front and go over their plans with them and the transformation sponsors to ensure that everyone was on the same page. After the product owner and scrum master had completed these procedures, the agile coach would guide the entire team through a one-week “sprint zero,” during which they would get agile training and create an initial work backlog. Following sprint zero, the agile coach attended major ceremonies during the team’s first few sprints to ensure that it was stable.
5. Concentrate on the transformation’s long-term viability.
Governance bodies (such as a committee comprised of senior IT leaders) will almost certainly be required soon after agile infrastructure teams have been launched to ensure that the teams are progressing toward their goals, refreshing their objectives as the organization’s priorities change, and improving their use of agile practises. Furthermore, many infrastructure businesses soon identify a variety of opportunities to build on the early benefits of the agile transformation. Redesigning career models to accommodate new agile jobs, implementing more flexible budgeting methods, and making strategic planning more agile are just a few examples.
It will take time to address these areas for improvement, but senior IT infrastructure leaders may do so by employing the same methodologies as their newly formed teams. They can work as a team to organise themselves, establish a backlog of opportunities, prioritise tasks, assign owners, and complete work in sprints.
Legacy IT infrastructure processes, which are ubiquitous in businesses that weren’t “born digital,” can stymie the rapid delivery of new digital products and capabilities. Agile approaches can greatly speed up the process, and the benefits can frequently be seen within the first six months of implementing them. A modern IT infrastructure group that works closely with developers and employs automation to speed up configuration and maintenance can significantly improve its own and the company’s performance. This can make the difference between success and demise for incumbents facing disruption from digital challengers.
