We've worked in numerous types of organizations, from multi-billion dollar corporations to non-profits and small startups. We've worked with Operations teams, been part of the Operations team and comprised the entire Operations team. We've learned what works and what doesn't work in various types of organizations. What we've found is that while advances in Cloud Computing and automation continue to surge, many Operations teams still operate like they were 20 years ago. There is a better way…a MUCH better way: Continuous Delivery in the Cloud. This means 100 % automation of the delivery process (build, test , deployment and release) and the utilization of Cloud resources. Cloud by itself is just a buzzword. Cloud coupled with 100% automation is where you get huge productivity gains enabling you to delivery software as quickly and as often as the business desires.
While the Cloud ROI Calculator we developed is geared toward how much costs are reduced (for example, over $3 million USD per year for a 20+ application organization), the most considerable value, in my opinion, is when implementing Continuous Delivery with 100% automation in the Cloud, you can release more quickly and more often.
How to Enter Information
There are four items of data entry: Average Engineer Hourly Rate, Number Of Applications In Portfolio, Average Size Of Projects and Average Technical Architecture Complexity. Since all data entry uses sliders, you can use the default values if you want to see how it works and then go back and modify values to see different results based on your organization. A medium sized project is approximately a 25K-100K SLOC code base, small is less than 25K and large is greater than 100K. Keep in mind that it's an average of all of the projects in your organization. The technical architecture complexity considers the number of application servers, number of database tables, configuration and other technical complexity. Again, it's the average technical complexity of all your applications.
Explanation of ROI Results
Once you click the Calculate ROI button, you'll see six rows of information, which are explained below. For each row, you'll see the number/cost when using 100% Automated Operations (what Stelligent provides for our clients) vs. the Traditional Operations team that manually performs and queues human tasks.
Number of hardware instances - For medium size projects and complexity, we assume 10 ephemeral instances per application and 20 fixed instances.
Hardware costs per year - Because of commodization and economies of scales, cloud instances are approximately 1/4 the cost of managing your own data center. Source: A combination of The Economics of the AWS Cloud vs. Owned IT Infrastructure and work performed by Stelligent.
Number of engineers – This is where the real cost savings is: number of engineers required to create and support a 100% automated cloud environment. using 100% automated cloud, you pay 1/3 the costs in terms of the human capital required to create and maintain the infrastructure. Source: Work performed by Stelligent.
Engineering costs per year - The total cost based on the average engineer rate in your organization, the number of applications and whether it's automated cloud or traditional Operations.
Organizational cost per year - The hardware plus the engineering cost per year.
Total savings per year – The amount in savings between automated cloud and traditional Operations per year.
About ROI Calculator Development
We used the Platform as a Service (PaaS) offering, Google App Engine (GAE) to develop the calculator. My friend Andy Glover developed the first release of the calculator using GAE, Groovy, Gaelyk, etc. We've since provide new features. The application is essentially stateless, but we're using BigTable to manage certain configurable values. Because it's a PaaS offering, we don't worry about hosting, uptime, etc. It just works. GAE also provides a comprehensive dashboard that has extensive logging, etc.
Commentsif ( $comments ): ?>
Trackbackselse : ?>
_e('No comments.'); ?>endif; ?>
Post a comment
Comments are moderated, and will not appear until the author has approved them.
Comment below or sign in with