October 18, 2014 | Written by: Vittorio Della Rossa
Share this post:
Cloud computing is a trending topic, and although it is much debated, few speak about how it will affect information technology (IT) organizations. Years ago I drew a diagram of a generic IT organization’s structure, but upon revisiting it today I realized how many changes in IT departments I had left out. This led me to speculate about what changes cloud will cause to happen in the near future.
We can start by taking a look at the messy picture that I drew:
At the top left side of this graphic we can see the symbol of the ever-present lines of business. This is where I put all revenue-related enterprise activities, including enterprise functions that rely on IT and the related requirements and needs. I know, this is an oversimplified representation, but, as we are focused on IT organization, it is acceptable for this purpose.
At the bottom left, we have the users. In this diagram I do not distinguish between internal users and customers, but the Information Technology Infrastructure Library (ITIL) will forgive me.
At the top right, we have application development, which includes all the people involved in designing, defining and developing the applications needed to conduct business and used by users. Their deliverables are given to the operations department. This division is in charge of service delivery. In IT organizations, users usually have a stronger relationship with user’s support, because an internal help desk is where users usually raise issues or requests within the IT organization. The last section in the diagram is demand, which functions as a kind of ambassador between IT departments, lines of business and users. This function has to collect requirements and coordinate the deployment of new services, as well as improve the quality of existing ones.
IT organizational structure in the 1970s and 1980s
IT departments have not always been structured as they are in my first diagram. When I first started working, the IT organization was linked to the business only through development departments. These development departments were in charge of translating business requests into programs capable of making employees’ daily work more efficient and cost-effective. Users were mainly internal (in the same company), but they belonged to different departments.
The operations department was in charge of hardware management—the operating systems and the middleware software. This department also handled print execution, tape mount, the daily start and stop of all systems, and system monitoring. The shift supervisor could be contacted by an agency’s managers when they encountered problems. So, even if the function of user’s support was not defined, the operations department was implicitly in charge of supporting users.
1990 to early 2000: IT becomes service oriented
The first strong transformation in the structure of an IT organization arose with the introduction of the service concept. Processes models (like ITIL) and the automation of activities in the data center reshaped the IT department. Some activities disappeared; others were reduced; and new ones were introduced.
It is important to note that I will not speak about ITIL or use its terminology to describe the organizational structure, for two reasons. First, I don’t want to spread the habit of using ITIL terminology in naming organizations’ departments (which I find to be a misuse of the model). Second, I want my description to be as generic as possible to better represent a synthesis of all the organizations where I have worked, ITIL-inspired or not.
The service perspective introduced two innovations. The demand management function was introduced with the goal of interpreting business requirements and translating them into IT initiatives and operational plans. This function is accountable to the business functions for service quality. Furthermore, the users (both internal employees and external customers) changed their roles to align to the service perspective. Users now needed an explicit point to contact IT support. In the early stage small user support groups were started, and then a well-defined department with these specific goals was drawn into the organization’s chart.
These innovations resulted in people leaving the data center and populating call center rooms and service operation centers instead. New roles that would be able to handle a customer issue or operate on monitoring and automation software were created.
Could cloud adoption further change IT organizations?
Now the cloud is no longer at the horizon. It is a computing model that is already changing the way services are delivered. It is (or it should be) used to give better and faster answers to both the business and the users.
Variation, speed, reliability and safety are fundamental requirements for modern users. They enter the marketplace of Internet services with one hand (the one holding their mobile device) and can quickly decide whether to buy or engage with a product or service. For companies, it is easier to sustain this kind of consumerism with cloud than to try to size their IT to be capable of responding to any kind of situation. But how will IT organizations be affected? Do they need to change in the near future, or can they handle such a model with a traditionally structured organization? Is it only a new technology aspect or will it require new professions and a redesign of activities and responsibilities?
I want to start the discussion around cloud innovation and the change in the IT organizational structure by proposing some hypotheses about possible organizational scenarios.
One way to approach the cloud will be as a technology facilitator. With this approach, the change will appear mainly as infrastructural integration and be hidden to developers and users. Cloud planning and management will be confined to the operations department, which will be the main beneficiary of advantages related to efficiency and speed. In this scenario, the operations department becomes a kind of internal cloud provider; it can easily provide environments for development departments using internal resources or those obtained from an external cloud provider.
This approach should not affect the overall IT organization, but it will require a substantial transformation of the operations department. It must be more independent in planning and spending capabilities. More people skilled on provider management, as well as capable in architectural cloud service integration, will be required. The capability to manage a hybrid environment will be essential to avoid an internal siloed model and the consequent wasting of resources.
I believe this first cloud approach will be pursued by many companies as a tactical phase. It may also be adopted where IT is an independent company in a holding structure.
Another scenario is where public cloud will satisfy the majority of a company’s infrastructure needs. In this case, IT will no longer need an operations department. This will also imply focusing more on users than on infrastructure. User support will be most affected by this transformation and will become the core of service delivery. It will be able to quickly modify infrastructure when needed or start a recovery by leveraging the simplified management of the cloud.
The development department will increase its resource acquisition capabilities, observing only the constraint to acquire resources from the same cloud provider or from one compatible with the provider used for the production environment. This department, together with the demand function, will operate technology choices during the service definition.
Demand could increase its role in defining, maintaining and innovating the enterprise architecture for the company. This department will be staffed with people who have more holistic and architectural skills because it will have the knowledge of the whole company’s IT.
A deeper transformation could occur if the use of software as a service (SaaS) or platform as a service (PaaS) becomes pervasive in the company. An internal development function will no longer be needed. Instead, the demand function will be staffed with integration and customization skills. The core of service design will be shifted from functional design to an aggregating design. The demand generation will act more as a broker, building services based on collected requirements by integrating the default services delivered by cloud providers. At the very most they will customize some software with integration and control capabilities.
I am aware that these could appear to be risky considerations. But the cloud is no longer just knocking at the door; it has already become part of our IT. “Suffer” or “embrace it because it’s the current trend” are extreme attitudes that don’t allow you to make correct choices and take advantage of the potential of this new technology.
Innovations in cloud computing are clearly starting to change the structure of IT departments, and in this post I’ve considered just a few possibilities for what that might look like. What do you think about my initial attempt? How can we further refine these diagrams to reflect the new structure of IT organizations? Please leave a comment below or connect with me on Twitter @vidierre.