The strategic significance of the Cloud

Written by: Michiel de van der Schueren

Much has already been written about Cloud Computing. And especially in terms of the new business model (because of on-demand software and machine capacity!), the possible cost advantages, the (in)security and the Service Level Agreements. All important things, but we read too little about the strategic significance of Cloud. While that is the first question we should ask ourselves: why do we want to move to the Cloud and what is it’s strategic importance? Most organisations have to constantly adapt to changing market demands or other changing circumstances and the speed with which the end product can be adapted in form, function or numbers determines the speed of this innovation process. If Cloud Computing enables us to speed up that process, then it will be of great strategic value to any organisation. It’s high time to zoom in on the strategic significance of Cloud! We’ll do that in this blog.

Unbundling the Corporation

In 1999 Harvard Business Review published the article “Unbundling the Corporation” by John Hagel III and Marc Singer, both principals of McKinsey & Company. In that article they described the primary process, as it applies to most organisations, with a division into three main processes: (1) product innovation, (2) production and (3) marketing & sales.

Of these three main processes, the McKinsey men found that they each have completely different dynamics:

Product innovation requires a creative process in which the talent of individual employees is central and must be used to the maximum. Economics of Skills is more important here than efficiency or effectiveness.
Production revolves around achieving the lowest cost per unit, predictability of the production process and increasing productivity through standardisation and mass customisation. In short: here the money has to be made with ‘Economies of Scale’.
For the marketing & sales department, the customer is central. The relatively high costs of this department make it necessary to realise as much turnover per customer as possible. All employees must be focused on the market, customers and sales targets. This part of the organisation needs ‘Economies of Scope’.

Due to these differences in dynamics, each of which requires its own type of organisation, it is very difficult in practice to unite the three main processes in one organisation. It would make much more sense to place the three main processes in a separate organisation. The disadvantage of this was that the additional coordination and communication costs are so high that it was more economical to organise the three main processes within the walls of one company. With the advent of the internet, that argument has been dropped. This is evident from the example of Apple, which illustrates what “the unbundling of the corporation” can look like in practice.

Why Apple has outsourced its production

With the introduction of the iPhone and iPad, sales records at Apple follow each other in rapid succession. Steve Jobs has been able to come up with a product that perfectly meets the needs and expectations of consumers like no other. Nice for Apple, but now they face the challenge of actually delivering the numbers that are requested by the market. How do they do that? Where are these devices screwed together? Apple has selected a number of suppliers for this, of which Foxconn Technology Group from Taiwan is the most important. Foxconn is the largest OEM manufacturer of electronics in the world and has factories in China, among other places. The largest factory is located in Shenzhen, a city in China with 16 million inhabitants, close to the border with Hong Kong. Until recently, 140,000 people worked in that factory. Foxconn was commissioned by Apple to deliver no less than 21 million iPads in the first quarter of 2011. To make this possible, Foxconn quickly recruited 160,000 extra employees, doubling production capacity. If Steve Jobs had ever thought of producing his brainchild in America, Apple’s success would have come to a screeching halt. The speed at which Foxconn can scale in China would be utterly impossible in the West.

Unbundling the IT stack

Although this is an example from the ICT industry, these are hard goods that come from a factory and that require a surprising number of hands. What if the product to be delivered is not a product, but a software application or an online service? What does the “unbundling” of the primary process look like? The answer is Cloud: decoupling the seemingly logical bundling of the different layers in the IT stack into the three layers of the Cloud stack: an Infrastructure-as-a-Service layer, a Platform-as-a-Service layer and a Software-as-a-Service.

To explain that, let’s peel off those layers one by one:

Layer 1. Infrastructure-as-a-Service

The Infrastructure-as-a-Service layer is essentially the factory of Cloud. Thanks to automated provisioning systems, machine capacity can be added or disabled here on-demand. The speed and flexibility with which this is possible speeds up the process of taking a new service into use. In addition, virtualisation ensures that all management activities are fully automated, so that many more servers can be managed in the Cloud per system administrator than outside the Cloud. Market research by Intel shows that in the pre-virtualisation period, an average of 35 servers could be managed per system administrator, compared to 42 servers per system administrator after the advent of virtualisation. A disappointing result, when you consider what all that virtualisation has cost. In Cloud, however, IaaS providers appear to be able to manage up to more than 2000 servers per system administrator. The volume allows them to extend the automation of management tasks to infinity. Management must be automated! This results in ‘Economies of Scale’, which an ordinary data center does not even come close to.

Layer 2. Platform-as-a-Service

The ‘Economies-of-Skills’ are realised on the ‘Platform-as-a-Service’ layer, as intended in the theory of “Unbundling the Corporation”. We choose the best components, build what is still missing ourselves and assemble it into a custom solution. This way of on-demand assembly offers for the first time in history the possibility of mass customisation in IT, which in itself is an unprecedented innovation. For large companies with legacy systems, it is possible to integrate new online applications with existing IT systems using a PaaS platform.

Layer 3. Software-as-a-Service

On the Software-as-a-Service layer, the software vendors offer their applications at a price per user per month. These can be stand-alone applications, but also applications that are part of a larger PaaS platform, as described above. In the latter case, the mutual integration will be child’s play, but even in the case of a stand-alone application, integration with other applications is generally easy to realise using web services. That is, if you have a good SaaS provider that publishes open APIs to anyone who wants to work with them and with which web services can connect. The simplicity and speed with which it can be integrated with other applications thanks to the fact that they are all web-based applications is evident. In addition, SaaS providers have the advantage that they only have to maintain one platform as a provider. This makes an upgrade to the next version child’s play compared to the upgrade of a software package that runs on-premises at customers. As a result, the frequency of upgrades can be increased and many small upgrades can take place instead of having to bundle them into a few major upgrades per year, as has been the case with packaged software so far. SaaS providers plan upgrades per week or per day instead of per year, which makes it immediately clear what that does to the speed of innovation. The advantage of Cloud for a SaaS provider is that they can fully concentrate on their core business, namely: developing and managing their application. If they use a PaaS platform for this purpose, this offers additional advantages with regard to the speed of programming and the related cost savings. But even if they only use the infrastructure of an IaaS provider, they no longer have to worry about the availability of the infrastructure.

The benefit of using SaaS services for the end users is also evident. They no longer have to worry about their applications, which are used as an online service via the internet. They no longer have to worry about patches and updates, which are now provided by the SaaS providers. They also no longer have to worry about investments in hardware or the lifecycle of their applications, which they purchase flexibly and on-demand. This way they can concentrate on their own core business. The advantages of ‘Economies-of-Scope’ apply to both parties, buyers and providers of SaaS services.

“Do what you are good at and outsource the rest to third parties who are better at it”

We call the described three layers in the Cloud stack the Cloud enablers, because they enable the customers of the cloud providers to purchase exactly what they need. Flexibility not only in terms of scalability and pay-as-you-go, but also in terms of functionality. With this, Cloud enables us to apply the theory of “Unbundling the Corporation” to the IT stack: do what you are good at and outsource the rest to third parties who are better at it, then you will see that you are more flexible and can innovate faster.

But Cloud adds an extra dimension: communication between the layers in the Cloud stack has not only become simple and cheap, but by using Web services or APIs, there is automated, real-time information transfer. And if information is always available up-to-date somewhere, you naturally want to make optimal use of it: do not repeat what someone else can do faster or better than yourself. Thanks to the ‘Platform-as-a-Service’ layer, information from different sources can be integrated and the process between that information and different applications can also be automated.

To reflect this development, Gartner has added two additional layers to the Cloud stack: Information Services and Business Process Services, which we explain in more detail below:

Layer 4. Information Services

By Information Services we mean those services where only information or data is made available. This can be done once, as in the case of Carspotter, where the value of the car in question can be retrieved by sending a registration number by text message. But this is also possible in a subscription form, which gives access to a database in order to request data from the database during the subscription period. A good example of this is the information from the Chamber of Commerce, which can be queried using web services, for example by a CRM application. The CRM application does not store this information, because that only leads to outdated data, but always reads the most up-to-date data in real time. The Chamber of Commerce collects and produces data and the CRM application retrieves it from where it can be produced best and cheapest: at the source. This is called ‘Connected CRM’ and could lead to a CRM system in the future consisting of a framework without a database, in which all kinds of Information Services are bundled into a total overview of customer data.

“Technique is out, Trust is in”

Layer 5. Business Process Services

Business Process Services refers to services where data is not only collected, but also processed. That operation adds value. This can also be a one-off action. Think of an online survey system, which we have all completed at one time or another, or an online exam system, with which you can take the exam for, for example, the Sailing Certificate. But it can also be repetitive in nature, such as the production of payslips. A good example of such a service is iDeal, an online payment system of cooperating banks that is nothing more than a shell around the existing internet banking systems. By combining the various internet banking systems in one iDeal portal, a universal online payment system was built. The interesting thing about this example is the fact that a new player has emerged in the market, which puts all other players behind in terms of trust with end users. And that at a fraction of the cost for which a whole new online payment system would otherwise have had to be built! That is a typical characteristic of the Cloud: reuse of existing applications, with the added advantage of the relationship of trust with the customer in this case. Technology is ‘out’, Trust is ‘in’. iDeal is now the market leader in the Netherlands and the system is so successful that expansion abroad is being considered. For example, more of these types of disruptive business cases will appear on the market in the future.

The Information Services and Business Process Services are collectively referred to as the Business Enablers because they enable the customers of the cloud providers to build new services using the Cloud Enablers, creating value for the business.

Final verdict

We have seen that Cloud allows us to apply the theory of “Unbundling the Corporation” to the IT stack, distinguishing between “Infrastructure-as-a-Service”, “Platform-as-a-Service ” and “Software-as-a-Service”, which together form the Cloud enablers. This allows us to flexibly organise the primary process of online applications, online services or our own IT department and respond more quickly to what the market or organisation requires in terms of form, function or numbers. Using those Cloud enablers, we can also develop new innovative Information Services and Business Process Services, creating value with the help of IT. All this means that Cloud is of great strategic significance for any organisation where IT plays an important role in the primary process.

Do you want to know more about the influence of Cloud on both business and IT?