Often-times, the three words: value, cost and price, are considered somewhat interchangeable and sometimes that might be correct. However, let’s make a distinction for now and look at cloud computing.
Value is tied to the utility rendered from a product, service or technology.
Cost is tied to the consumption of resources (financial resources, man hours, energy or other) needed to perform or produce a product or service.
Price is the, by the market, agreed upon amount needed to acquire a product or service. The price determines who of the market participants that will capture the value rendered.
There are surely more precise and correct distinctions of these three concepts to be found elsewhere, but the ones above will do for this analysis.
From fundamentals to market specifics
The order by which the three concepts were mentioned above was not random. When trying to determine the possible benefits of IT as a Service for the system of industries associated with IT, the analysis should start by determining the increased value it brings. Secondly, the cost implications needs to be considere. After looking at the big picture and establishing that the increased value outweighs the increased costs (if any), the next question is how the excess value is devided among the market participants. The following paragraphs will outline the main points of the value proposition, the drivers for decreased and increased costs and how that might affect the current and future price.
Value
- Allowing firms to focus on their core activities – estimates indicate that around 80% of firms’ IT budget is spent “keeping the lights on”, or at MOOSE – maintanance and ongoing operation of systems and equipment. These activities are not developing the enterprises’ competitiveness. IT as a Service means that these activities will be carried out and improved by a service provider according to best practice. It is important to recognise that it is not only ownership and operation of the bottom of the application stack that is reallocated to the cloud provider. Firms are also freed from performing support, research, development and managing other related issues such as compliance.
- Reduced complexity – A core component of cloud computing is a smart interface between the underlying infrastructure and application developer or user. With such an interface, the underlying complexity is abstracted away from the top layers of the stack. The sophistication and complexity of the underlying technology of leading cloud providers is also unattainable for most firms by themselves. This in combination with not having to acquire any physical hardware results in fast deployment of IT projects.
- Elasticity – IT as a Service enables firms to increase and decrease the scale of their computing resources at a pace and to an extent not possible with proprietary hardware.
Cost
- Scale advantages – The main factor behind the favorable cost implications of ITaaS is the sheer scale of the cloud providers. The scale enables more efficient procurement, energy usage, hardware (at scale, hardware failure becomes a certainty which makes “enterprise grade” and commodity hardware equally advantageous), etc. In total, the increased efficiency is estimated to halve the cost per computing instance.
- Multi-tenancy – a contributing factor to that increased efficiency is multit-enancy with uncorrelated load levels, because it means lower total need of redundancy.
- Leveraging scale of other web scale applications – Many major actors in IT as a Service are leveraging the scale and technology developed for their core business (the e-business of Amazon and Google’s search, advertising technology etc.)
- No initial investments – When IT is acquired as a service, no initial investment is necessary. However, the investment is not eleiminated, they are simply transferred to the cloud provider but the total cost of capital can still be reduced because of better solidity and access to financing among the cloud providers.
- Transactional costs – Purchasing ITaaS instead of performing the activities in-house could entail increased transactional costs, but since cloud computing also results in dis-intermediation and the technology involves a well defined interface between the underlying infrastructure and application/user, it is hard to determine whether transactional costs will increase, decrease or remain unchanged. However, current scepticism involving legal and lock-in uncertainty can be interpreted as transaction costs.
As a comment on the overall cost, John Keagy, CEO and founder of GoGrid predicts that cloud computing will decrease the total cost of the IT economy from $1.5 trillion today to about $500 billion in a decade.
Price
The price is ultimately dictated by the value and cost, but in the short term, especially in still emerging and oligopoly-like markets like ITaaS, the price and price structure is controlled by the dominant players.
Since the suppliers are contributes with the initial investments, the price structure can be of a pay-as-you-go kind. This together with elastic scale means that firms can pay for actual demand for capacity instead of projected peak demand for capacity, and they don’t have to pay a big chunk of it in advance. The typical current usecases for IaaS and PaaS emphasize these two key benefits by involving either variable demand, high costs of capital or both. Inertia and uncertainty have made cloud adoption most common among young companies lacking previous investments in infrastructure, test and development of applications and applications with very variable load level. Subsequently, the current price level is not primarily meant to compete in stable settings.
Thus, it is not very suprising that the cost of ITaaS have sometimes been found to be higher than traditional alternatives. (Examples: Real-world cloud computing, Could Cloud Computing Cost More?) The pay-as-you-go strucutre and the current levels have even been grimly interpreted: “the main attraction of the cloud to investors and entrepreneurs is the idea of making money from you, on a recurring, perpetual basis, for something you currently get for a flat rate or for free without having to give up the money or privacy that cloud companies hope to leverage into fortunes“. Granted, there are issues such as potential lock-in (an issue to be discussed in a separate post) that could give cloud providers enough power to keep prices high. However, since there are fundamental factors for increased value and lowered costs for the whole eco-system, it seems more likely that ITaaS has the potential to offer lower prices and total costs for the demand side. John Keagy of GoGrid speculates around halved prices within a decade, adding that “today, there is a bookstore and GoGrid who provides IaaS [...] hardly a competitive market“.
Another issue related to price, is how it is still very hard to compare service offerings from different suppliers. There are currently multiple projects working with standards for cloud computing, many of which are dealing with technical issues or approaching the price issue from a technical point of view (like defining a Universal compute Unit, 1 UCU). However, standards, standard metering and standard prices are also important steps towards really making computing a commodity. Like John Keagy said, it is hardly a mature industry, but the potential for improved value and costs are promising.
