Cloud and Converged Infrastructure Economics (Cont’d)


When reading about organizations considering or moving to the cloud the underlying reasons for this move are cost saving, competitive and financial advantage and increased agility.  Cost savings includes consolidating or converging physical resources and virtualizing infrastructure and lowering labor cost.

Competitive advantage can be the result of either a planned or unplanned change in the business that’s having a huge impact on IT resources.  This could be due to changes in underlying business conditions, expansion, downsizing or to meet other organic needs.

Financial advantage includes improving ROI, lowering TCO, and reducing time and cost to serve while improving overall time to value.

Agility in cloud computing allows for easier innovation with lower risk simply because you take away the need for capital investment in new or additional equipment, resources and technology while minimizing exposure and risk.   If the effort fails for whatever reason you don’t need to figure out a way to recoup capital outlay.

Last but not insignificant is necessity.  Necessity could include replacing outdated methods and improving the means of doing business or reaching a new market.


Today companies and their MIS teams are very adept at building solutions by providing the ideal technologic solution matched perfectly to customer needs meeting all technical requirements.  With the right customer input and customer access a provider can go about gathering technical data, analyzing it and providing a complete bill of materials and proposal.   This usually includes professional services to help with software, hardware and infrastructure installation, support, licensing help and assistance operationalizing the solution.

What’s Needed

We need to understand how and why not only the technical decisions are made regarding cloud services but also how the business decisions are made.  As important is how you measure and monitor ongoing results.  I have found that an effective way to achieve these goals is through engaging in a cloud economics discussion which likely will lead to important considerations regarding the cloud or converged infrastructure decision.  Economic information and detail can be more important to the success of the effort than you think.

The benefit of the economic discussion from a business perspective to the business is clear: the person, group or team responsible for articulating not only the reason for moving to, not moving to or leaving the cloud puts themselves in a advantageous position to help shape vision, direction and thinking, building a deep understanding of the situation, and establishing credibility.

If you have established a position of being able to help across a broad range of business process issues – not just with the technology itself but also with the underlying processes – then we begin to advance towards the ability to sell ourselves, services and products over time, again and again.

Cloud Economics

Before explaining CES some background on Cloud Economics.  Classic economics defines a service as an intangible commodity.  Services are an intangible equivalent of economic goods.

Providing cloud services along with it being a technical effort is also an economic activity where the buyer does not generally, except by contract, obtain exclusive ownership of the services or underlying infrastructure.

By designing, delivering and orchestrating the appropriate level of resources, expertise and knowledge for effecting specific benefits for the service consumer, cloud providers participate in by delivering service solutions that meet agreed to customer’s requirements.  These services, however, require large investment in physical structure, equipment, infrastructure, support and consume large amounts of resources, such as Capital and FTEs.

Cloud Economy of Scale Characteristics

In order to explain Cloud Economy of Scale and how we can use it to affect the success of our relationships with the services provider or the service consumer there are some characteristics or factors of the relationship that need to be explained that will have an impact on the outcome.


It is important to remember that cloud services are perishable.  The relevant resources, processes and systems are assigned for service delivery during a definite period of time. If the designated or scheduled service goes unused it’s wasted.  Time cannot be recovered.   (Resources x Time Based Units) x $ = Cost of Service).  From either a provider or consumer view what’s not used is not generating value or revenue but could still be costing money to maintain.

When the service has been completely rendered to the requesting service consumer, this particular service irreversibly vanishes as it has been consumed by the service consumer.


Cloud providers are responsible for service delivery as they must promptly generate and render the service to the requesting customer.  In many cases the service delivery is executed automatically but the service provider must prepare and make available resources and systems and actively keep up appropriate service delivery readiness, support and capabilities.

The Customer is inseparable from service delivery because they are involved in it from requesting it up to consuming the rendered service(s).


Services are rendered and consumed during the same period of time. As soon as the customer has requested the service (delivery), the provider is obligated to deliver the particular service without delay and in good order and the service consumer consumes the delivered services for executing their upcoming activity or task.


Each service can be unique. It can be one-time generated, delivered and consumed.  Cloud services can also be repeated as the point in time, location, circumstances, conditions, current configurations and/or assigned resources can be the same.  Most of what a CP provider provides are regarded as heterogeneous and are typically modified for each service consumer or each new situation.

Each of these factors can complicate consistent service delivery make service delivery a challenge in each and every case.

Mass generation and delivery of services is difficult. This is more of a problem of meeting inconsistent service needs, requirements and quality.  Both inputs and outputs to the processes involved in providing services are highly variable, as is the relationships between these processes, the provider and the customer making it difficult to maintain consistent predictable revenue generation, lower costs and profitability for the provider.  For many providers there is the challenge of allocating possibly scarce resources.

Due the elastic nature of cloud services it is difficult to achieve economies of scale or maintain expected revenue generation and margin.  There are demand fluctuations and it can be difficult to forecast demand.  Demand can vary for many reasons including season, time of day, business cycle, etc.  There can be customer involvement as most service provisioning requires a high degree of interaction between service consumer and service provider. There is a customer-based relationship based on creating long-term business relationships.  Most if not all providers desire is to maintain long-term relationships with their customers. These customers serve as referrals for other potential users of the providers services.

So considering these factors you can see that it’s extremely important for the provider to be able to not only provide services on demand but also more importantly to know when to…

It’s just as important from a consumer perspective that the service …

Some of the largest players in the cloud space recently lowered some of their charges for the services they render.  Recent public cloud price cuts include:

How have they accomplished that while meeting consumer needs and maintaining margin on these offerings?  The answer is through Cloud Economy of Scale.  The agile service provider needs to have the ability to work with consumers of the service to demonstrate and help them achieve the same outcome through the use of various forward leaning tools, applications and solutions.

Consumers or customers of cloud or converged infrastructure services need to understand this in order to put themselves in a more advantageous negotiating and operational position.

Cloud Economy of Scale

Everybody understands utilization and that’s great.  But, how do we know when to make a change in the service, which direction to go, expand or contract, scale up or down based on just technical factors.  What is the appropriate tier that a user should be assigned to?  Most providers and consumers you talk to would like to know the answers to these questions but are not sure how to get them.  What’s needed is the ability to look at the service(s) being delivered and consumed in terms of Cloud Economy of Scale.

Before explaining Cloud Economy of Scale for services lets discuss utilization from a different perspective.

The following graph is an example of resource utilization over some period of time from a providers view.   This customer signed up for Tier 1 services from the provider.   That includes a number of servers, CPU, memory, storage but not networking.   I could have left it in but left it out to simplify the example.

The horizontal line represents what the provider has defined in his service catalog as the maximum level of resource utilization agreed to and defined as a description of the provider’s Tier 1 offering.   They may or may not use this service description with their customers.   The definition of Tier 1 services might remain internal to the provider.  Their customer facing service description may be totally different, either somewhat under or over the internal service description

The blue shaded area above the support level line is the area of greatest risk to the provider.   He is using his resources beyond what might be considered optimum which can result in a service failure of some type, leading to customer satisfaction issue at a minimum, loss of the customer, loss of revenue and the resources not being used at all.   The provider would still bear the cost of maintaining resources without being able to apply revenue to them.   Depending on the scope of the impact on the provider’s infrastructure and on how long it takes the provider to get a new customer or user to replace the lost business affects the severity of risk of loss to the provider.

The white shaded area below the Support Level line represents an area of revenue loss and\or increased cost per unit of measure (CPU, memory, GB, MB, etc.).  If these resources are dedicated to a customer then he has to somehow encourage the customer to use them generating revenue to offset the costs.  If they are not dedicated he may be able to move other users to that resource group improving his revenue or lowering his cost.

utilization_vs_timeFigure 1 – Utilization vs. Time

So, how does the provider know what to do to mitigate his risk of loss and\or increase his revenue?  His choices seem to be acquire additional resources to support the customer’s SLAs when then exceed the ability of the current collection of resources to support them or re-negotiate the SLA and move the customer to a higher or lower tier.  But how many more resources will he need to do this and what’s the timing and cost for doing that.  Do it all at once, in stair step fashion or smoothly, gradually in small incremental steps.

In order to make the right decision at the right time the provider needs to evaluate his choices by looking at Cloud Economy of Scale.  The depiction below oversimplifies the process but should help you understand how it works.

Cloud Economy of Scale really means doing things efficiently through scale of purchase; resource specialization; better use of financial resources; full utilizing technical resources; operational and technology advantage.  If a provider has control over all of these factors he can reduce his long run average cost and improve his ROI and margins.

It’s been argued that Cloud Economy of Scale for services will come from improving the flow of service, from the first customer request to satisfaction of that request.  If we can demonstrate to our provider or client friends that not only by choosing the right technology and solutions but helping them with evaluating and establishing their own Cloud Economy of Scale model we put ourselves in a position of creating a long term relationship, becoming a trusted adviser and supplier.


Figure 2 – Cloud Economy of Scale

In economics, a cost curve is a graph of the costs of producing something as a function of total quantity produced. In a service business efficient firms can use these curves to find the optimal point of service delivery (cost minimization point), and can use them to provide deliverables at quantities to maximize profit or lower cost.  There are various types of cost curves, all related to each other, including total and average cost curves, and marginal cost curves, which are the equal to the differential of the total cost curves. Some are applicable to the short run, others to the long run.   For discussion purposes in this paper we will use the long run average cost curve (LARC).

In the cloud or converged infrastructure world Cloud Economy of Scale (CES) is important so far as it’s an indicator of a change.   For example, a provider has built out an environment of some size (capacity, performance, power, storage, management capability, etc.).  He markets his services based on this environment either internally or externally.  He’s at point A.  The provider’s environment has known capabilities beyond which the ability of the environment to deliver the agreed to services diminishes, (refer back to Figure 1 – Utilization).  Prior to maximizing the environment the provider’s environment is considered underutilized.   Some of his service delivery mechanisms are not involved in generating revenue.  He certainly isn’t making a profit and might be losing money.    His per unit costs are high.  As he gets more customers or users into his environment his per unit costs diminish to a point where he is maximizing the return on his investment and possibly margin, point B.  He’s achieving full use of all of his resources and reached optimal scale.   Then as the environment becomes saturated, when his customers start to really consume his services, his costs rise to a point where he’s experiencing rising cost, increased support issues and is likely losing money and operating with little or no margin, he’s moving to point C.  This area of the curve is called diseconomy of scale.

From a provider perspective when the data starts ascending from point B to point C it is time to look at a change in process, policy, procedure, operations, cost or the underlying technology for a refresh or upgrade.  Reaching this point leads to price, cost and services discussions with customers.

From a consumer of the services perspective it’s important to recognize what’s happening with the current solution or provider.  Hopefully as a consumer you have been tracking this yourself and are not caught unprepared in evaluating the current situation and are able to make a decision whether to stay with the present solution or look at alternatives.

In Conclusion

What the Cloud Economy of Scale discussion attempts to do is start you thinking about where things are and which direction they’re moving or trending.  By using a tool like this and discussing the implications with all of the actors it can lead to an understanding of and give reason to making the necessary changes in the infrastructure, the services delivered or costs. Understanding Cloud Economy of Scale and how it can help you make the right decision is as important as making the right business, operations and technology choice.