Driving Change and Managing Innovation in a Cloud-First Business Reinvention

“Without change there is no innovation, creativity, or incentive for improvement. Those who initiate change will have a better opportunity to manage the change that is inevitable.” – William Pollard

I discovered the future of enterprise computing five years ago, when I attended QCon in San Francisco. And, surrounded by some of the world’s leading architects and engineers from companies like Amazon, Facebook, Netflix, and LinkedIn, I kept picking up on two major themes at the conference — the move to micro-services and distributed systems, and the use of AWS to host these micro-services architectures. Coupled with AWS, the new architectures provided unprecedented levels of scale, elasticity, and availability. But even more astounding was the amount of innovation and change that sites like amazon.com and Netflix were able to drive into their platforms without downtime.

Back in 2011, when I attended QCon, I was a managing director in Accenture’s IT organization, responsible for Enterprise Architecture, Agile Delivery, and Innovation. Before that, I spent 15 years driving global, large-scale IT implementations and transformations in many areas of Accenture’s business functions. Today, however, I’m an Enterprise Strategist at AWS, and my role is to share experiences and strategies with enterprise technology executives so the cloud can help them increase speed and agility while lowering costs.

Returning to my technological epiphanies at QCon — once the conference ended, I was eager to put my new learning to work. I immediately changed our architecture principles at Accenture so they were services-first and cloud-first. We then looked at ways to apply AWS services to meet our challenges.

There was an application which leveraged a proprietary scanning and imaging technology to ingest millions of documents. But the system took days to scan and process the documents and have them available for the user to see. With low — and unacceptable — levels of customer satisfaction, my team created an architecture component that allowed our custom applications to use AWS S3 as a storage mechanism. The team then created a mobile application that allowed users to take pictures of the receipts and store them in S3. The benefit was quicker storage and retrieval achieved in seconds with no additional load on the legacy application.

But the truly amazing result was cost; what was implemented was actually 100 times cheaper than the proprietary system because of the move away from the need for a physical scanning solution and leveraging S3’s low storage costs.

Another example of how the power of AWS was leveraged was in the management of development and test servers. Despite having virtualized over 95% of server hardware in our data centers, the elastic usage pattern of development and test environments meant they were only utilized a low percentage of time; even for a well-run operation like this, that was still too much waste. So, the development and test environments were shifted to AWS and put on a schedule that shut servers down during weekends and nights. The benefits of the cloud were clear at this point, but we still had internal debates as to whether the cloud or on-premise machines were cheaper to run. Through these projects, however, we began to invest in training and enabling architecture, engineering and infrastructure resources on cloud technologies.

Three years ago, we had a changing of the guard with a new CIO and new boss — and Accenture IT shifted from an era of consolidation, outsourcing, and cost take-out to an era of digitalization. Like so many organizations, we felt the pressure to quickly adopt cloud, mobile, analytics, and other capabilities to create end-user-focused digital services.

In order to move at speed, and with agility, we launched a cloud mass migration program with the goal of having 90% of our workloads in the cloud in three years. A year into the program, all workloads in our data center on the east coast were migrated to the cloud. We also provisioned 90% of new infrastructure in the cloud, particularly when it was aligned to new investments. For example, when we re-platformed the accenture.com website, it was provisioned from the get-go on AWS. As a result, we achieved savings of $3.6 million in annualized benefit because the cloud enabled optimized server schedules and server sizes. With our services running in the cloud, we also saw better performance, better uptimes, and lower mean time to resolve when incidents arose. Toward the end of 2016, over 60% of workloads were in the cloud. Meanwhile, Accenture was on track to shut down its primary data center by August 2017, and it was also on target to meet the 90% goal. This put to rest any debate as to whether the cloud was cheaper.

Despite Accenture’s rapid migration to the cloud and AWS services, it still didn’t have a Netflix or amazon.com-like story. This changed in 2015, however, when there was direction from the C-suite to implement a critical new capability in less than a year’s time.

Once we got to work on this, we decided to build a micro-services-based architecture on AWS because the business had to leverage an iterative design process and we needed the ability to adapt rapidly to changing requirements. Through the development of this capability, new features and changes were continually deployed into production. When I reviewed the final stats with the team, I was blown away by the results, which were pretty awesome and far-reaching. In less than a year, we had —

• Deployed over 12 major releases
• Developed 20 micro-services
• Deployed over 4,000 times to our environments, with zero downtime
• And delivered a successful service and experience to a global population of almost 400,000 employees

The business was absolutely thrilled with these major breakthroughs. And, during our post-mortem, I asked the team if we would have been as successful without the power of AWS. The answer was no.

I have many more examples of how the public cloud and AWS have enabled new capability, lowered cost, and averted capacity and scale challenges. And I’ll share all this — and more — in future blog posts.

One last note — as I reflect on Accenture’s journey to the cloud, I see a strong alignment with the AWS Stages of Adoption mental model. Accenture started with a few projects so it could begin to leverage and understand AWS services. It then moved to the foundation stage by growing a group of people focused on cloud-centric architecture and engineering. And, once the transformative decision was made to become a digital organization, Accenture implemented a mass migration to the cloud. Finally, with the latest services we developed, Accenture was able to re-invent how it architected and engineered services with new levels of capability, speed, scale, and availability.

I believe every enterprise has the opportunity to deliver stories like this — and stories like those of Netflix and amazon.com, too — and that’s why I’m super-excited to be a part of AWS and help enterprises in their journey to the cloud.

Never stop innovating,

Joe
chung@amazon.com
@chunjx
http://aws.amazon.com/enterprise/