Share this post:
Genuine digital transformation requires cloud modernization strategies that involve and affect people, processes and technology. In part one of this two-part series, I focused on cloud-based technologies that enable organizations to modernize infrastructures, platforms and applications.
This time around, I’ll examine the people aspect of cloud modernization and include some thoughts on process. I’ll cover:
- Why it’s necessary to refresh and reskill your workforce
- Why organizational restructuring is important
- How to reengineer people and process
- How to measure employee value in an era of transformation
The most effective digital modernization initiatives are holistic efforts that coordinate people, processes and technologies to improve business outcomes.
Refresh and reskill the workforce
Running IT operations in a cloud environment requires a different set of skills and knowledge than a traditional application development world. This is because application and platform architectures are so dissimilar. Therefore, different engineering practices and design patterns will become relevant.
Consider the microservices architecture, in which a single application is made up of many small, loosely coupled and independently scalable and deployable components. If one service goes down, the same service can pop up 10 different times. Or what about 12-factor application methodologies for building cloud-based services — developers and designers need to understand these and ingrain them into the design.
Concepts like these weren’t that relevant in traditional application development. Understanding of a traditional infrastructure doesn’t necessarily translate to a cloud environment. For a cloud modernization initiative to succeed, enterprises must teach employees how to design, test and deploy native cloud-ready applications — or hire new employees who can.
The cloud model also calls for consuming more and more best-of-breed software-as-a-service (SaaS) capabilities provided and managed by third-party vendors. This also exacerbates the notion of consolidated service-level agreements and integrated service management skills in a multicloud environment. These are very hard-to-find skills.
Restructure the organization
The notion of development best practices and operational excellence isn’t new. So why are concepts like DevSecOps so highly talked about now? The answer has more to do with organization than technology.
Five years from now, we may be laughing at the technology we have today. But whatever technology we’re using, we’ll want development and operations teams to be tightly integrated. In the microservices architecture, you own the end-to-end lifecycle of the service or domain. That’s very different from how traditional infrastructure IT teams work.
Cloud computing is increasingly managed by self-sustaining independent teams, each with their own specialties and organized around billing, claims or similar domains. There’s a squad lead. There’s a business analyst. There are a few designers and a few developers. There are probably some testers. There’s a DevOps engineer, or maybe a site reliability engineer (SRE). Each is part of the total squad — and squads can be replicated. As the industry matures, it’ll continue to trend toward that model.
Another trend in organizational restructuring is the growing adoption of centers for enablement. Centers of excellence or competencies are nothing new; centers for enablement are more hands-on and dynamic, focusing not on process but on performance and reuse of assets.
Adopt new rules and process reengineering
Site reliability engineering is all about building highly resilient applications and platforms. If a service suddenly goes down, other services remain working because they are independent. So, if a feature on an airline app stops working, users can still look at current flights or their airline bonus points because the problem is isolated.
The widespread use of automation and robotics has spawned a discipline called AIOps, which is revolutionizing service management. Enterprises initially took a reactive approach to monitoring systems and handling events. With new technology available, enterprises can be proactive about monitoring and management. AIOps allows organizations to be prescriptive: They can know what your systems and instrumentation will look like six months from now because the AI perceives trends and patterns.
Measure people and performance
In a traditional IT development, developers are evaluated using metrics such as the number of defects in a kilo line of code (KLOC). But is this metric relevant in cloud-based development? Wouldn’t a more valuable metric be the amount of time from ideation to deployment? And what about the percentage of a developer’s time spent on building automation, or the number of reusable IT assets produced by individuals or squads? The idea is to use metrics that truly measure business value.
In a nirvana stage, mean time to recover a system should trend closely to time taken from ideation to deployment. This means the method and processes used for new requirements are exactly the same as those being used for bug fixes or other types of defects. They’re all treated as items in a prioritized list and will be consumed, built, packaged and deployed in exactly the same way.
Half measures simply won’t cut it. Digital modernization requires a cloud-first and automation-first strategy that emphasizes coordination between people, processes and technologies.
Learn more about IBM organizational change management services, as well as how guidance, migration, modernization, cloud-native application development and managed services from IBM professionals can help your journey to cloud.