Yesterday, at the Boston Cloud Services Meetup at the Cambridge IBM Innovation Center, Amalgam Insights (AI) attended a Cloudyn-based event on “Overcoming the Challenges of Multi-Cloud Financial Management.” This presentation was headed by Account Executive Marcus Benson and focused on the challenges that Fortune 500 companies and managed service providers have in managing both complex single-vendor and multi-vendor cloud infrastructure environments.
Cloudyn is a cloud business and financial management solution founded in 2011 and set up as both a multi-tenant and multi-cloud solution running on AWS, Microsoft Azure and Google Cloud. Cloudyn supports a single pane of glass view for consolidated management and a real-time and continuous support of cost optimization for multiple vendors including Amazon Web Services, Microsoft Azure, Google Cloud, OpenStack, and Docker. Cloudyn has raised over $20 million in venture capital and seed funding and currently targets large enterprises, managed service providers, and companies with over 1 million dollars in annual cloud spend.
As a starting point, Cloudyn emphasizes the role that cost plays in demonstrating an anomaly that can potentially define the success or failure of a cloud project. Because cost is the “canary in a coalmine” that can show whether utilization and peak traffic are associated with ongoing functionality and ROI or not, effective cost monitoring of cloud resources is an important starting point for understanding whether a project can be profitably supported through cloud resources. An example Cloudyn brought up was a company with significant video streaming that made a series of code and query changes to improve performance. Based on these tweaks, the CPU and data rates went up 6x leading to a monthly bill that was 2.5 times greater than the month before. Performance may have remained stellar, but at a significant, unexpected, and unmanaged increase in cost.
Cloudyn tends to see three main challenges for cloud cost management. First, throughput-hungry NoSQL databases such as DynamoDB that can be set up to ingest massive amounts of data without careful management. Second is the challenge of IT governance in tracking the origination of new services that are spun up over time. Third is the problem is dealing with “forgotten” servers where hundreds of disks may be unattached, but still active to result in the active billing of abandoned computing resources.
The last of these challenges is a familiar one for anyone who has ever managed multi-million dollar IT budgets, especially in telecom or wireless expense management. In our world of telecom expenses, it has been common to find network elements going into a now-closed branch location or to find boxes of unused cell phones or pagers hiding in a closet that have never been disconnected. And just as in telecom, where these ignored services led to thousands of dollars in unnecessary charges, enterprises are also paying the price when it comes to these ungoverned cloud spending resources.
Because of these unmanaged environments, Cloudyn estimates that it typically saves 20%-40% of a company’s cloud costs when it first comes in. Within these savings, Cloudyn estimates that approximately 40% savings come from operational optimization such as repurposing and right-sizing resources. The other 60% of savings come from the financial optimization to look at improved pricing models, better discounting and tiering terms, and the potential utilization of private cloud or dedicated resources. To support this right-sizing of CPU usage, network input/output, and choosing more optimal pricing models across multiple clouds, Cloudyn typically charges companies based on a percentage of cloud spend.
In starting with new customers, Cloudyn described a relatively fast implementation where the initial setup could be done in a matter of minutes based on Cloudyn’s multi-tenancy and use of standard usage APIs to track cloud resources. Based on this setup, a customer can expect to gain insight into their cloud activities in approximately 24 hours and about a week to benchmark the environment and start identifying savings opportunities. Cloudyn also provided a live example of a customer that had optimized its cloud spending environment to a steady-state approximately two months after first implementing Cloudyn.
Cloudyn also painted an alternative picture of the challenges of managing cloud infrastructure as a Do-It-Yourself project: 25-30 GB of billing information on a monthly basis broken up into tens of millions of line items across a bundle of virtual machines. With this type of scale that is very reminiscent of telecom billing, it becomes clearer why large cloud customers end up choosing a dedicated cloud cost management solution rather than conduct their own management tools.
One of the most interesting announcements Cloudyn provided was in announcing the imminent release of the CloudynDex: The Cloud Efficiency Index. Based on visibility to what is currently 17% of global cloud spend and over 400,000 virtual machines, Cloudyn will be creating an anonymized industry benchmark to demonstrate cloud efficiency on a multi-cloud provider perspective including IaaS vs. PaaS, runtime distributions, and other basic measures of performance. Customers will be compared to an anonymized peer group to identify relative performance and to provide recommendations for improved cloud efficiency and resource consumption. Amalgam Insights believes that this benchmark will both provide Cloudyn customers with a competitive advantage in managing the Total Cost of Cloud and improve Cloudyn’s ability to support improved business management of cloud infrastructure and platform resources.
Overall, Amalgam Insights saw a wide variety of similarities between this newer world of cloud cost management and the more established world of telecom and technology expense management. AI is aware that large telecom expense management solutions are increasingly asked to manage cloud infrastructure and platform spend based on existing relationships with the CIO office. Based on this initial presentation, AI provides the following recommendations for enterprises, telecom expense vendors, Cloudyn, and other Cloud Cost Management vendors seeking to catch up to Cloudyn.
1) Enterprises with current or planned annual cloud spend of over $1 million should consider partnering with Cloudyn or another established cloud cost management partner. The level of detail that Cloudyn provides for CPU, service, input/output, runtime, and peak usage for cloud infrastructure exceed the detail that any technology expense solution has shown to AI. In the short term, it would be better to work with a market leader rather than re-create the wheel.
2) Cloudyn should partner with Global TEM vendors, potentially as an arms dealer to support cloud optimization practices and to provide an add-on product to accompany existing network, telecom, and mobility spend. Cloudyn should be an especially powerful partner for TEM vendors that either provide reseller capabilities for cloud services or have a strong network optimization offering, as Cloudyn is accretive to both environments. In general, vendors seeking to enter this space are recommended to partner with existing CCM vendors because it will be more cost-effective and provide enterprises with better outcomes over the next two-three years.
3) Cloudyn, although a market leader, still has room to improve. The product itself represents modern use of analytics and design layout and the implementation and ROI results should be the envy of any enterprise application company. However, the financial management outputs do not quickly lend themselves to integration with higher-level IT finance or enterprise performance management solutions. In addition, Cloudyn currently lacks the direct access needed to check for potential vendor-based billing errors at the time of invoice input, a capability that is standard in the telecom expense world. This is not a weakness of Cloudyn, but rather an indication of the less mature world of cloud billing where AWS and Microsoft are the “new” carriers.
Overall, AI believes the cloud cost management space is still in its infancy as enterprises are just starting to learn how to source and support the multi-vendor environments that have become standard for data centers, networking, mobility, and other core IT environments. As cloud continues to mature from its current estimated hold of ~5% of enterprise data to an early majority range closer to the 20-25% range, AI expects that cloud cost management will explode as a new market and that enterprise targets for CCM will be defined as those with $5 million+ in annual spend rather than the current threshold of $1 million. Ultimately, the velocity of growth for CCM will be defined by the speed of data gravity shifting from on-premises to the cloud over time and will be driven by an inexorable combination of increased DevOps, video streaming, and the Internet of Things.
If you have further questions on Cloudyn, the Cloud Cost Management market, or the Telecom Expense Management market, please feel free to contact us directly at info@amalgaminsights.com