IT Consulting Services - Cirrient Technology Partners

Is the economics of the Cloud getting lost in translation?  

By: Scot Moir | April 27, 2020


When organizations think of the Cloud they’re focused on an outcome, what they want to get from moving to the Cloud. But they also need to be thinking about the Cloud vendor’s business model, what fuels it and how that translates into what they value? To get the most value you need to leverage what fuels the economics of the model. If you don’t understand the economics, it’s like playing a game without knowing the rules, it makes it difficult to win.

The value in the Cloud comes from sharing. The business model for Cloud vendors is to serve customers with shared resources. Their revenue is tied to how many resources they have to share, and their profit is tied to how well they share those resources. It is sharing that enables scaling and economies of scale are part of the Cloud economics. The more you’re willing or able to share, the better your goals are aligned with the Cloud’s economic model the more value opportunities become available to you.

What organizations value will vary. Your ability to capture value from the Cloud is very much dependent on the characteristics of your business, the capabilities in your systems and your views on sharing. However, if you’re talking about the Cloud but not talking about what you can or are willing to share, your Cloud efforts may miss their mark.

The Cloud and sharing.

When we talk about Cloud, we’re talking about Cloud computing. Cloud definitions vary a bit based on the perspective of the defining organization. The National Institute of Standards and Measures (NIST) has a good definition1. I’ve trimmed it down to, "Cloud computing is a model for enabling … on-demand network access to a shared pool of … computing resources … that can be rapidly provisioned and released … ‘’. “Shared” is the critical word in the definition and is consistent across definitions. Clouds come in different flavors, broadly, there are two Cloud models2. A Cloud that is shared within one organization is a private Cloud. A Cloud that is used and shared by multiple organizations or “tenants” is a multi-tenant Cloud. The most popular variation is public Clouds.

The goal of sharing is better resource utilization.

Why is that significant? To answer that we’ll talk about a hardware gap, virtualization, complementary workloads, and the public Cloud. The foundation of computing resources is hardware. Hardware has a fixed capacity and a fixed cost. For example, a server has a preset number of resources; processors, memory, etc. that define its capacity. An application (software) uses a portion of a server’s capacity. The amount of work an application needs to do and the time it takes to do it, varies based on demand. Demand could be driven by the number of users being supported or the number of records to be processed, etc. Depending on demand an application may use a little, a lot or all a server’s capacity. If the server hardware capacity is not fully or more appropriately, optimally utilized, then organizations are paying for capacity they are not using and the cost of running the applications is higher. This is the hardware gap, hardware cost is fixed, but workloads vary which often leaves servers underutilized. This is the issue and the opportunity where sharing comes in.

If one application doesn’t use all the capacity of a server then the server can potentially be shared with other application(s) to improve server utilization. Provided the applications don’t interfere with each other which is a significant concern. An application can use too many resources impacting the performance of the other application(s) or it may compromise the security of the other application(s). This makes finding applications that can share a server challenging and often considered too risky. Which is where virtualization comes in.

Virtualization technology helps address issues with sharing.

Virtualization is software that simulates hardware. Virtualization that simulates a server is known as a virtual server aka a virtual machine or a VM. Essentially, it creates an artificial environment that protects the application and keeps it from interfering with other applications. A software supervisor manages the virtual servers and controls access to the physical resources. The software supervisor enforces resource thresholds to keep the virtual server from using too much capacity. Applications run on a virtual server just as they would a physical server. Multiple virtual servers can run on one physical server. Which means through virtualization, applications can safely share a physical server with each application running within its own virtual server. Virtualization makes sharing a server less risky and provides better resource management but to improve utilization you need complementary workloads.

What is a complementary workload?

Applications or workloads that can share a resource and improve resource utilization without impacting each other are complementary workloads. To find complementary workloads you need to understand the usage pattern of applications when they are in use and how many resources they use. The goal is to get the server to its optimal utilization by finding workloads that keep the server consistently busy without exceeding its capacity. Even with virtualization organizations may not have the usage patterns to make a significant improvement on resource utilization.

Organizations’ workloads often follow similar patterns. A typical pattern has systems busy during the day and slower at night. They can work to improve utilization; however, the cycle of their business usually leaves them with peaks and valleys and consequently underutilized resources. Regardless of how well this is done they are not sharing costs. A public Cloud adds another dimension, it allows for the sharing of computing resources across organizations.

In a sharing model customers pay for what they use (or think they’ll use). They will pay a premium for the service if the value is greater than owning and operating the resource. Sharing improves resource utilization. For Cloud providers, good utilization maximizes the revenue generation of a resource. Public Clouds have multiple clients and operate on a large scale. Their scale helps them improve utilization by providing more opportunities for sharing, more organizations and workloads. The scale does more than help with sharing it creates economies of scale.

Economies of scale.

Economies of scale allow vendors to provide services at a price point that is both attractive to customers and makes a profit. Economies of scale drives down the costs that Cloud vendors pay to produce their services. As an example, they pay less for hardware because of the volumes they purchase. But it is more than buying things for less, they operate differently.

To operate efficiently and effectively at scale they need to do things differently. Their scale allows them, and forces them, to specialize. They create new methods and tools to manage and run their operations and reduce the cost of delivering their services. As they scale and mature the specialization becomes part of normal operations and a competitive advantage. An advantage that is hard for those not operating at their scale to emulate.

Economies of scale also help with revenue. Their model is based on how many resources they have in use and how well they utilize those resources. To grow revenue, they need to add resources, increase the number of customers, and increase spending from existing customers. They do this by continually innovating on new services. Scale helps here as well. They have a large, invested, customer base. They have an ecosystem of partners to fill gaps in their services. They have extensive information on how their services are used. All, this lowers the cost and risk of developing new services, reduces the cost of selling and accelerates the adoption and profitability of new services. A continuous stream of new value-added services improves their value proposition to their customers driving growth and revenue.

A few more points.

Value may come from sharing, but customers don’t want their operations to be impacted or to incur more risk because they are sharing. A primary focus of the new services is breaking down the barriers to sharing.

Many of Cloud services are only viable because of sharing and scale. Services like on-demand computing and elastic computing. To provide these services Cloud providers need some level of capacity available (not in use) to handle these requests. They can do this by keeping utilization lower on active resources and/or by having standby resources. Both solutions reduce overall resource utilization and impact the return Cloud vendors get on their assets. It is the reduced cost from scale, the number of customers requesting the services and the ability to share the resource across customers that creates an economic model that makes the services viable.

Good resource utilization relies on predictability and complementary workloads. The Cloud vendors influence utilization and demand using pricing models to encourage usage patterns that improve predictability and the use of underutilized resources. That is why you pay less when you make capacity commitments than you do for on-demand capacity, and why they offer substantially reduced spot pricing when you make short term use of idle resources.

Finally, not everything in the Cloud is based on sharing, but if their customers aren’t sharing, if the Cloud vendors stray too far from their business model their profitability suffers, and they get into direct competition with software and hardware vendors, commercial and private data centers.

Wrapping it up.

Cloud computing, especially public Clouds deliver value through sharing and scale. Sharing is about improving resource utilization, spreading out costs and improving margins. Scale creates opportunities for sharing and new services. It reduces the cost and risk of bringing the new services to market. Scale is also about economies of scale to drive down the cost of resources and operations. Sharing and scale is what makes many Cloud services like on-demand and elastic computing viable. Sharing and scale together drive profitability and are foundations of Cloud economics.

We’ll finish how we started – Your ability to capture value from the Cloud is very much dependent on the characteristics of your business, the capabilities in your systems and your views on sharing. The more you’re willing or able to share, the better your goals are aligned with the Cloud’s economic model the more value opportunities become available to you. If you need or want some level of isolation in your Cloud operations, you’ll pay more for those services and limit your value proposition. But, if you are sharing, leveraging Cloud services and evolving your operations to capitalize on the Cloud vendor’s business model then you can maximize the value potential of the Cloud.


1NIST definition, "Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction." Mell, P. & Grance, T. (2011) The NIST Definition of Cloud Computing. (NIST Special Publication 800-145). National Institute of Standards and Technology.

2NIST identifies two other cloud models, hybrid, and community. For the purposes of this article I don’t address hybrid Clouds and I consider community Clouds a variation of a multi-tenant Cloud. Mell, P. & Grance, T. (2011) The NIST Definition of Cloud Computing. (NIST Special Publication 800-145). National Institute of Standards and Technology.