There’s no denying that digital now drives the world in which we live, and like it has every business, that evolution has impacted Accenture.
As I shared in my previous post, our internal IT organization decided to respond to our changing environment by undertaking an aggressive three-year program to create a more scalable and robust IT infrastructure in the cloud.
Upon making the decision, we developed our strategy and got down to business. This strategy included three major phases—adoption, optimization and transformation—that would help us achieve our end goal. Here’s how we executed on each of those phases:
Phase 1: Adoption
We had a rather ambitious timeline to move our business applications to the public cloud. Our first—and most aggressive—target in Phase 1 was to move from nine percent to 50 percent operating in the cloud within one year. Our focus was to accelerate cloud adoption by:
Rapidly migrating business applications out of on-premise data centers and into the cloud (IaaS)
Diligently adhering to a “cloud-first, cloud-only” strategy and principle for all new applications
As far as the rapid migration, we focused on one data center to start, with the goal of moving all business applications out of that one virtual on-premise data center within one year. Whereas some companies start safe with non-critical applications, Accenture took a more impactful approach by migrating our highest-cost data centers in North America first, even though the decision represented a greater risk due to the involvement of production environments that supported business-critical applications. We made this decision to keep up with our aggressive plan and to prove that such a migration could be done at scale and at speed with mission-critical applications.
Next, our cloud-first, cloud-only approach required all new applications to be architected and designed for the cloud and all new infrastructure to be provisioned directly in the cloud. This meant, for example, designing and deploying Accenture’s new website, accenture.com, in the cloud—another bold move.
Since Accenture wanted to move quickly, we took a strong position on cloud first, cloud only for any new applications, and this approach enabled the global team to start with development and testing in the cloud, which led to an effortless public launch.
Initially, our Cloud team had to overcome the challenge of not having a full solution for migrating complex enterprise applications at the desired pace. Additionally, since cloud technology solutions for the enterprise are ever-evolving, the team needed to come up with innovative and flexible solutions and engage business partners to embrace the impacts of the changes and understand the resulting benefits.
Our internal IT organization also had to fill in skills gaps. While the team had expertise in navigating our internal IT governance and stakeholders, they did not have experience as cloud migrators and brokers. To fill this gap, we partnered with the Accenture Hybrid Cloud Services team to create and test technical tools and procedures for the migrations and to standardize and train people on the supporting processes required to execute successfully in a highly matrixed and varied business stakeholder and application environment.
The result was a full cloud migration blueprint with a new methodology designed specifically to address these technical and non-technical enterprise complexities. We are now reusing this blueprint to migrate the remaining virtual on-premise data centers. Perhaps most notably, the teams that are migrating our internal IT organization’s complex legacy applications to the cloud are comprised of the same people who will undertake similar migrations for our clients.
Although our team kicked off the program from greenspace—creating and managing entirely new processes—this approach allowed us to quickly accelerate activities, ultimately delivering with rapid, military-style precision and making significant progress. In fact, we migrated and provisioned new virtual machines at a pace of approximately 250 per month, resulting in a total of 4,500 virtual machines in the cloud over 18 months. Our efforts will continue to move forward at this pace—or faster in some cases—with the start of another major data center move this fall.
Phase 2: Optimization
Once we set our technical foundation and had cloud adoption well under way, our internal IT organization’s focus shifted to optimizing the new cloud model.
We first pursued optimization opportunities in infrastructure hosting and tool automation, and progressed from there by narrowing our focus on streamlining processes and simplifying governance. All of these optimizations led to business benefits that are contributing to our ongoing transformation.
These optimization efforts also created an opportunity to improve our cloud hosting model, as we only need to pay for resources we consume. This flexibility allows applications to request more or less on demand, which means that instead of overestimating to create large contingencies, teams can react and adjust to real-life circumstances in order to proactively and aggressively avoid unnecessary costs.
Specifically, we have reduced the standard “uptime” schedules for non-production servers and enabled our teams to start and stop their cloud machines on their own without having to go through manual tickets and procedures—an early step toward self-service transformation. Similarly, teams have influence to actively manage the actual server size of their cloud machines, and our Cloud team now reports on historical data to recommend an optimal server size every month to increase efficiencies even further. Armed with this data, application teams are responsible for forecasting their cloud hosting spend and creating baselines to continuously track against as part of the regular financial process.
Phase 3: Transformation
In addition to optimizing the efficiency of our growing cloud infrastructure, the Cloud team also recognized the importance of moving forward on our path to true organizational transformation. This transformation is the third phase of our journey.
To deliver the desired transformation and be better equipped to keep pace with evolving cloud technology and capabilities, our Cloud team adapted and implemented Agile delivery practices. Agile delivery is usually implemented by singular, co-located software product development teams that execute daily sync-ups, incremental development and frequent delivery.
In our case, our Cloud team committed to transforming toward Agile principles, but needed to do so in an environment that is not traditionally Agile—a globally dispersed and infrastructure-focused universe with multiple technology functions and owners. The team adapted the traditional Agile approach by using virtual tools for daily sync-ups, combining sprint boards across multiple technology functions and managing the work horizontally (versus vertically within one product), with one lead scrum master having full visibility to the cloud delivery roadmap. The team fully committed to adopting these changes and ceremonies, which included a procedural shift as well as a mindset shift to engage in an intense level of collaboration and to prioritize speed to delivery.
By transforming the way it delivers, our Cloud team can now keep pace with incremental changes that help transform our internal IT organization into a self-service environment. For example, one of the key selling points of the cloud is the speed to provisioning. The team’s challenge is to benefit from this speed while maintaining enterprise standardization and security posture. To meet this challenge, our Cloud team relied on the new Agile delivery framework to reengineer, automate and integrate processes and tools in order to significantly reduce standard infrastructure provisioning effort and lead times while still satisfying standardization and security requirements.
Transforming our pace of delivery has directly impacted our pace of cloud service adoption (both at the infrastructure and platform levels) as well. Layering on to our now-established Agile practices, the Cloud team worked closely with multiple stakeholders to introduce a more streamlined process for “certifying” and adopting new cloud services for business applications. Again, the challenge lies in maintaining enterprise standardization and security posture while still enabling speed of adoption to bring to life innovative solutions that meet emerging business needs. The teams agreed upon different stage gates and requirements based on the extent of adoption. For example, full enterprise adoption and support for new services requires more sign-offs than a service that will only be used by specific applications. This approach allows for faster adoption of lower-impact services, and we can always expand to full enterprise adoption as needed.
Overall, these changes in approaches allow our team to overcome the limitations of legacy technology, processes and governance, which require heavy and segregated services. These advances will continue to create speed and efficiency that will allow us to reskill IT people and incentivize new behaviors that better match the fast-paced digital world in which we now operate.
Preparing for a Smarter Future in the Cloud
With the three phases of our cloud journey complete, Accenture is well on its way to preparing for a smarter future in the cloud.
Currently, our internal IT organization is building on these efforts by working toward the long-term goal of re-dedicating resources from infrastructure management to development of Platform-as-a-Service and other cloud capabilities. We want to do this so that we can re-allocate resource time and capacity to development work and focus on delivering more value back to the business.
The opportunities this change will create to operate with increased flexibility and agility, both technically and procedurally, in the cloud will lead to increased speed, productivity and innovation—and that’s where the lasting business benefits lie. With the optimized technology, processes and delivery we now have in place, the IT organization is primed for just that, and we’re well on our way toward creating a self-service-focused environment enabled by streamlined processes, simplified governance and a more modern technology architecture.
Next up: Lessons learned.