Cloud Computing: Where to Start When You Have Already Started

The issues and opportunities surrounding SOA, SaaS, and cloud computing effectively amplify the conversational divide between business and IT, in part because of the newness of the architecture and the technology. Find guidance on linking the technical opportunities presented by these new models to the needs of the organization.

Ric Merrifield and Dennis Stevens

Comparing a major business change to repairing an airplane in mid-flight is a pretty good comparison in terms of complexity, risk, and disruption, and with increased focus on the opportunities related to service-oriented architecture (SOA), software-as-a-service (SaaS) and cloud computing, the comparison can begin to look like a gross understatement.  In fact, most organizations today are experiencing the collision of two crucial, but oftentimes competing objectives:

As organizations mature, usually through a blend of organic growth and acquisition, the different departments, divisions, groups, and companies become more standalone, like silos, and that results in enormous duplication of effort and fuels confusion when trying to explain everything to customers, partners, and employees.  Most organizations today seek greater transparency across their organization to present a more unified effort and message for everything from mission, to their value proposition to customers, to how they conduct their day-to-day activities.

While new technology pops up every day, major shifts are rare, and many people recognize that the opportunity presented by cloud computing is one of those major shifts, so many organizations are rushing to fulfill the promise of cloud computing, ahead of competitors.

As you might expect, if an organization is silo’d and lacks visibility across its various groups, opportunity maximization becomes more complicated.  Do you first unify and then maximize the opportunity? There isn’t time for that in many cases.  The right answer right now is to continue to aggressively pursue both goals, but with an eye on intelligent prioritization.  Unfortunately, that’s a lot easier said than done today – most organizations lack a view of themselves, a “lens” that enables them to efficiently and objectively prioritize on the unification side, which makes the opportunity maximization efforts almost doomed from the start.  Why is this?  There is a gigantic communication divide that spans the various organizational groups, including the information technology (IT) department.  At the heart of this communication divide is something that has been called ‘the “how” trap’ and that is discussed in greater length in Section 1 of this white paper.

Now back to SOA, SaaS, and cloud computing.  The issues and opportunities surrounding SOA, SaaS, and cloud computing effectively amplify the conversational divide between business and IT, in part because of the newness of the architecture and the technology.  The business is going to look for ways to offer new capabilities and services to both new and existing customers that will be great new sources of revenue, and create further competitive differentiation.  At the same time, many will consider migrating legacy technologies to cloud services to give new and existing customers more options in such areas as security, speed, access, and personalization.  The business is likely to provide the IT department with detailed “requirements” that include significant customizations to packaged or already custom software.  As we explain in the body of this white paper, if an organization hasn’t yet gotten itself out of ‘the “how” trap’, IT will typically create overly customized, more-expensive-than-they need to be services which will erode the potential profitability of these new models and ultimately slow the organizations efforts to achieve its strategic goals. 

Yet just as there needs to be clearer more objective articulation of the needs and priorities of the organization, there also needs to be a clear understanding of how these new models (SOA, SaaS, and cloud computing) will complement existing legacy IT solutions and architectures (back to fixing the plane in mid-flight).  Linking the technical opportunities presented by these new models to the needs of the organization is vital, and that is the guidance provided in this white paper.

In Section 1, we talk more about getting out of ‘the “how” trap’ through capability modeling and heat mapping.  Section 2 then offers guidance with respect to the role of heat mapping in business requirements and project prioritization. In Section 3 we get into what SOA, SaaS, and cloud computing mean, and the kinds of opportunities, and risks, they present to IT and the business.  Section 4 includes a cloud computing case study showing the basic hosted solution benefits a business should expect. .  The Conclusion section proposes that if you are thinking of cloud computing as just a technology solution/opportunity you are aiming too low for success in terms of value add to the business.  Until you apply the objectivity of analysis that you get through the rigor and discipline of business capabilities, the risk that you are not explicitly linking the strategies and tactics of the organization to the technology goes up significantly.  We close with some specific thoughts on next steps and resources that are available for this discussion. 

We believe that this paper offers clear guidance, as a first step, to help organizations like your own, continue to focus on unification and opportunity maximization, in a much more structured, objective, and disciplined way than what you have experienced in the past.

Getting out of ‘the “how” trap’ with capability modeling and heat mapping

Before we leap into these opportunities in technology and architecture let’s first call out a reality that exists in many organizations; the business side of the house isn’t interested in the specifics of the technology (even though they know more and more about it every day), yet they set specific strategies that translate into specific tactics, and quite often those require technology.  However when the discussion about the specific role of technology in supporting the tactics and strategies of the organization, much like a patient describing a pain to a doctor, the business will often include details that may or may not be relevant to the specific problem/opportunity.  In the case of the doctor/patient relationship this typically works itself out because the doctor is trained to filter out what is and isn’t likely to be relevant to what the patient really needs.  By contrast, in the business/information technology (IT) conversation, because the IT people aren’t experts in the business, they aren’t as able to see which details aren’t relevant and as such will often do many things that the business asks for that don’t translate into much value. This isn’t because people aren’t smart or they aren’t working hard, it’s because they literally often don’t speak the same language. 

The How Trap

At the root of this conversational divide, is something that has been called ‘the “how” trap’ and it affects us all.  People often become so attached to “how” they do something (like sending a fax), that there description of their work often masks “what” they are doing (communicating the status of something is more likely to be “what” is being done, and the fax is the “how”).  The “whats” that make up an organization are what we also call business capabilities (this is examined in greater depth in an article we co-authored with Jack Calhoun in the Harvard Business Review called The Next Revolution in Productivity).  We have found that identifying the various business capabilities in an organization to be an excellent first step in getting to a far clearer, more objective view of the work that makes up the organization, and this work is fast and most people enjoy it.  From there adding pieces of information about the areas that are most and least valuable, performance, and maturity can lead to highly objective and efficient discussions about work prioritization, especially when business value is an element of the conversation.  This is a first step that we recommend for most organizations. Rather than embarking on a months long mapping of the entire organization, we recommend that organizations start smaller at a divisional or departmental level to get a sense of how this approach differs from, and complements, other methods they may use such as process reengineering, lean, and six sigma to name a few.

Any conversation related to SOA, SaaS, or cloud computing that starts with technology, or even architecture, is aiming too low for success.  An important first step for both IT and the business is to get out of ‘the “how” trap’ and the good news is that it doesn’t take very long.

Most of us have had the “why are we going this way?” conversation in the car, and that is the result of people getting attached to “how” they get to a favorite destination when it rarely matters “how” you accomplish the outcome, or objective, of getting there on time.  It’s not that people are stupid, it’s simply that we often overlay “how” we do things in a way that starts to mask “what” the actual objective is, making it hard to think of any other way to achieve it.

Nowhere is ‘the “how” trap’ more common than in the workplace, and examples are the best way to illustrate that.  If you are trying to collect business requirements for a specific part of an organization, and not knowing anything at all about the organization you decide to walk up to a person at the fax machine sending a fax and ask them “what” they are doing.  The response you are most likely to get is “I am sending a fax” and in that situation most people are likely to have follow up questions like “Is sending a fax a necessary part of the work you do?  Do you have to send a fax to successfully complete your work?” and the answers are most likely to be “yes” which will lead the person to capture “send fax” as a requirement.  

But it isn’t.  The requirement, “what” they are doing, in this case is something more along the lines of “communicate status” or “confirm order” and “how” they are doing it is with the fax machine.  So if you then go back to the person having disentangled the “what” from the “how” and ask if it matters “how” it gets accomplished, the worker will typically see that it doesn’t matter, and which has already transformed the conversation about requirements.  The “whats” that make up the requirements of the organization are what we call its business capabilities and capturing the business capabilities is an effective and efficient way to get people out of ‘the “how” trap’ and that is an important first step in getting to the thermal diagrams we call heat maps.

The Value Performance Map

So then what is the first step out of ‘the “how” trap’? Large organizations are made up of thousands of business capabilities, and over time that inventory of business capabilities should take more shape, but right now, an organization should identify at least one of two things:

Identify the business capabilities that have the greatest business value, as defined by three equally weighted tests (and as an example, the business capability “pay employees” fails all three of these tests – it is necessary, it has to be performed successfully and be compliant, but it fails all three tests):

  • Does it contribute to the brand or identity of the organization in terms of why customers, partners, and employees do business with you.  Is this one of the things people associate with your organization?  (can be yes/no, high/medium/low, or 1-5)
  • Does the performance of this work link directly to an organizational key performance indicator (Yes/No)
  • Is there value in improving the performance of this particular business capability (Yes/No)

Identify the business capabilities that have the least business value. While this may seem surprising, this is where many of the greatest opportunities for cost cutting, consolidation, and outsourcing emerge.  Even if this is not a major undertaking at first, we recommend some work here initially, since working with any new approach or method will include some lessons learned, and getting lessons learned in a low value area of work is much lower risk than a higher value area of the organization.

So look at the work and strip away the “how” verbs from the work, the graphic below from an insurance company that has to create insurance quotes helps illustrates identification of “how” verbs.

Fig. 1 - Identification of How Verbs

As an aside, notice that the verb “Automate” at the left maps to the right with “None” - that is because automate is neither a “how” verb nor a “what” verb, it is a secondary description of a “how” verb and should be used with caution in these discussions.

From here, the next step would be to document the view of the business capabilities, the “whats” for the work that makes up creation of an insurance quote:

Fig. 2 - Create Quote

This by itself isn’t going to “wow” many people in an organization, but when you ask people about the business value of each block of work, including the parent “Create Quote” for an insurance quote, and then also ask how each is performing, you can color each business capability, where the shades of red (pink and red) are flags for your attention (high value, and poor performance in this case) whereas the shades of green suggest the opposite (low value, performing well). 

Fig. 3 - Create Quote: Business ValueFig. 4 - Create Quote: Key Table

This is where the conversation starts to get significantly more objective and interesting.  Now you can objectively ask which “child” business capabilities will cause the parent to perform better, and in this case, where the parent has a yellow “Fill” color that indicates medium business value, just because the child “Create Certificate” is high value and the lowest performing, you may be able to ignore that problem because the parent isn’t very valuable.  This is where the big impact of doing a business capabilities analysis comes in, which opens the door to more objective business prioritization, which then leads into the maximization of an opportunity such as cloud computing.  But we are now getting ahead of ourselves.

How to Use Heat Maps as Inputs to Prioritization Discussions

In our analogy of repairing an airplane in mid-flight, we highlight that it’s impractical to fix everything at once while you are still doing all of the work. We need to focus on the most important issues to maximize our results. Only after we have prioritized the issues, should we develop and scope solutions.

In Section 1, we discussed getting out of ‘the “how” trap’ using capability modeling and heat mapping. The heat map presents your business in terms of “what” you do and shows the business value and performance for each business capability. A critical aspect in developing these heat maps is that both the business and technology people should have contributed. So we have already started the conversation. How do we use the heat maps in our prioritization discussions?


Prioritization is challenging. The heat map literally presents a big picture, holistic view to share while making prioritization decisions. There are a number of considerations when prioritizing your efforts and investments. Most considerations fall into one of three categories:

  • Improve performance
  • Reduce costs
  • Address business risks

First, how can the heat maps show us where to invest to improve performance? Review the parent capabilities. Identify those business capabilities that are most valuable but underperforming. This shows a value performance gap where investment will result in improved performance for the organization. This includes identification of potential new business capabilities, such as cloud computing. These are the hotspots on your heat map where both the fill (value) and border (performance) are shades of red.  A minor point of clarification, it actually doesn’t matter which is the fill color vs. the border color, but we are being consistent here to avoid confusion in this article. The hotspots reveal the most important performance gaps. These areas require attention and investment to improve performance and typically should be only 10-20% of the list. Very quickly, by “listening” to the owners and practitioners of the work we have objectively prioritized our list of parent business capabilities to only these critical few business capabilities.

Now we need to scope our requirements. Take a deeper look into each “parent” business capability on the prioritized list to find the “child” business capability(ies) most in need of attention.  In the simplest terms, this is about causality – which child/children has/have the greatest influence on the performance of the parent? Our focus here will be on performance. Why? Most business capabilities can be categorized in one of three ways:

  • Value-added
  • Controlling
  • Supporting

Generally, only the value-added contribute directly to value (and they are quite often the “cause” of the performance of their parent). The controlling and supporting capabilities are necessary to ensure the value-added business capabilities perform at a high level. Poor performance in any of these types of “child” business capabilities could lead to poor performance at the “parent” level.   As an aside – as important as it is to find the causes of performance, most organizations have to start out at the assertion level (where there isn’t absolute certainty as to the “cause” of performance), but in our experience, because there is so much good knowledge and experience in organizations (waiting to get unlocked with a lens such as this), the initial assertions tend to be very accurate.

Using the Value Performance Heat Map

The heat map also reveals areas where we can add value through cost cutting, consolidation and outsourcing. The best way to do that is the identification of the business capabilities with the lowest business value. These are the capabilities have green fill (value) on your heat map. If these are performing well or not – they are low value, so not core, so consider outsourcing them to someone who can do this better, someone who makes this a core competency so you can focus on what is most valuable to your organization. If these are performing well or even over-performing, look to improve efficiency, cut costs – and narrow your focus to what’s most important. Also, consider consolidation of services. Is there really a need to have multiple systems and departments to “pay employees”?  This is much easier to see once we overcome ‘the “how” trap’ and describe “what” we do in terms of purpose and outcome. And because we have determined this to be necessary but lesser business value, the implementation details are less critical – it literally doesn’t matter who does it (or how many people do it), where it happens, what the technology is, or what the process is as long as it meets performance goals (including compliance).

Your heat map provides a framework to address risk as well. We discuss that in greater detail further in section 3.

We have presented an approach to use your heat map to prioritize and scope the issues, using the “what” lens to inform the needs of your organization. The next step is to consider the new or future state “how”. This requires dialogue, disciplined analysis and problem solving. The result is now a list of initiatives that address the major issues and achieve the future state. This list of initiatives may involve process improvement, training or technology. In the next section, we present how SOA, cloud computing and SaaS offer new tools to provide often superior solutions.

It is important to understand that issues and solutions are different. This sounds straightforward but can get messy in the course of discussions. In this section, we have described a top down analysis. Our prioritized list of issues as described in this section becomes a major input in the prioritization of solutions. This becomes more important when we recognize that other initiatives are already in play or under consideration.

How do we prioritize your existing list of initiatives? The heat map helps us relate issues and solutions. Identify the business capability that each solution is designed to improve. Does it address a hot spot? Remember, these are areas that typically require performance improvement. So if the focus of the initiative is cost cutting, we may want to examine that more closely – for example, while cost cutting is good is the cost cutting going to deliver enough value to justify the investment and distraction it causes?  Are we sure we aren’t cutting costs out of a high value business capability and if so, are we sure that’s smart? Another may address a green business capability (in terms of business value). We are looking for cost reduction here. If the focus of the initiative is performance improvement, we may want to revisit our business case. Do we really need to have the ability to “pay employees” in five minutes instead of one day? There are certainly more important areas to improve performance.

The heat map provides important input into the prioritization discussion. We can identify and prioritize those business capabilities that have performance, cost, or risk issues and that require attention. Further business capability analysis helps scope the issue. Now, we know where to focus our attention (and where not to) and can consider how to use SOA, SaaS, and cloud computing to deliver superior solutions.

In the executive summary we talked about the competing goals of unification and opportunity maximization.  In our experience, use of the heat maps is a way to drive conversations to a more tactical or granular level so that the trade-offs between unification and opportunity maximization can be evaluated in clear and objective terms rather than some of the more individual agenda, more subjective sorts of conversations that go on in the absence of something so objective.

Identifying the minimum scope in terms of what is required to be delivered to achieve the business strategy can greatly improve the business return on your SOA, Saas, and cloud efforts.

SOA, SaaS, and Cloud Computing Opportunities, and Risks, for the Business

First let’s take a look at what these terms mean. SOA, or Services Oriented Architecture, is an architectural style or way of thinking about the business and IT based on loosely-coupled, black-box components orchestrated to deliver a well-defined level of service.  SOA is architecture, it is not technology – many different kinds of technologies can support a SOA architecture just as many varieties of building materials can fulfill the requirements of the architecture of a building.

SaaS, or Software-as-a-Service, simply refers to a software enabled service that is offered via the internet.  SaaS is a way of delivering technology solutions that fulfill the architecture vision of SOA. Typically SaaS offered on a subscription basis (an operational cost) versus an up-front capital investment for the software license and the It infrastructure.  

Cloud computing, in its purest form, is simply the use of computer technology via the internet.  Cloud computing enables users and developers to utilize computing resources without knowledge or control over the IT infrastructure for those resources.  Resources are virtualized and provided via the internet. 

This model of software consumed as a service can be delivered by the vendor to the customer in many different ways; within the vendor’s own on-premise IT environment, in a virtualized hosted environment, in a cloud environment, etc.  SaaS does not need to be delivered via the cloud to be considered SaaS (though the cloud makes business sense for many SaaS vendors as it allows them to focus on building business value into their applications versus building and maintaining IT infrastructure).

SOA, SaaS, and Cloud Computing Benefits

SOA, SaaS, and Cloud Computing enable your IT group to support the business in four ways:

  1. Improve Efficiency and Drive Focus - SOA is the first opportunity we have had where, for the most part, the technology implementation can be talked about at the same level of thinking as the business. Understanding and building more precisely what the business needs in quantifiable, measurable chunks, leads to profitable delivery. Delivering on SOA through SaaS and cloud computing helps organizations focus on what’s most important, empowering IT management to deliver services and value to customers (inside and outside of the organization) efficiently. While most organizations expend significant resources in building and managing their own technology infrastructure, organizations leveraging cloud computing don’t have to focus precious financial, development and IT resources on deploying, managing, and scaling technology infrastructure.
  2. Improve Agility - When designed and implemented effectively, services have a high level of interoperability. This means that they can be composed with other services to deliver new services. New solutions can be delivered either by integrating existing services from outside providers or by bringing together previously segregated platforms. The ability to federate historically segregated platforms and to compose new services leveraging existing services enables IT management and development teams to respond quickly to changes in business and customer needs. This results in the ability to take an idea to market faster. Infrastructure can be quickly allocated to meet increasing traffic / adoption demands via simple request. This provides the ability to scale up and down seamlessly without complex operational procedures as well as the ability to upgrade a service without taking it down. This is far more scalable than traditional on-premises solutions involving procuring, installing, testing and delivering hardware, software, network, and storage infrastructure.
  3. Deliver Higher Performance and Availability - A true cloud platform delivers worldwide, geographically distributed datacenters, resources and platforms that dramatically exceed in scalability, performance, availability, redundancy, best practices and security what any organization can reasonably accomplish with their own resources. Just at using the cloud allows the business to focus on what is most important, cloud computing providers can provide best of breed service management and highly available solutions that individual companies cannot deliver.
  4. Help organizations balance flexibility and control - A cloud platform should empower organizations to select the best deployment model for applications – whether hosted on their own servers, hosted by a cloud provider or a combination of the two; helping  developers and service managers to combine on-premises and cloud resources to solve business problems.

These technology benefits provide new opportunities for the business. First, companies may plan to protect and grow revenue by improving or extending offerings through federation across legacy systems. For example, a financial services firm has been able to acquire new customers because it is able to present consolidated solutions of best in breed solutions unavailable in the market. They are able to do this, not by writing entirely new systems, but by federating existing applications through a SaaS implementation. Second, companies may improve their business agility through assembling targeted offerings or rapidly adding new offerings through partnerships and interoperability. Finally, companies can incrementally reduce IT costs by moving relatively expensive capabilities and environments to the cloud and by eliminating redundant implementations of capabilities within the enterprise.

SOA, SaaS, and Cloud Computing Risks

In addition to the new opportunities, the new technology and architectural approach bring about risks involved with effectively designing, building, and delivering services.

  1. Business Risks – The conflict between unification and opportunity maximization highlights the importance of the effective identification of boundaries and focus of the service efforts. A failure to have a focus on specific business benefits for each service acquired or developed can actually increase costs without delivering the expected business benefit. For example, building or acquiring a SaaS solution that provides a broad-range of Human Resource Management capabilities would not be valuable unless there were under-performing capabilities or dramatic cost improvements delivered by the new services. Unless the solution helps to protect and grow revenue, improve business agility, or reduce IT costs it may not be the right area of the business to focus on. The heat map provides a proven way to work with the organization to mitigate business risks.
  2. Design Risks – Designing services that don’t deliver on the promise of loosely-coupled, black-box components orchestrated to deliver a well-defined level of service will limit the eventual benefits the business can achieve. Building services that don’t align with the business model will limit the IT business alignment. A failure to follow design principles that support interoperability, autonomy, loose coupling, and Composability will limit the agility, availability and performance potential from the services. Consider how tightly interconnected a business capability is to others. The higher the level of interconnectedness at either the business or the technology implementation level the more difficult it will be to create an autonomous service.  Also, understand if there are significant compliance requirements with exposing a business capability as a service. For example, services that access PCI data may be designed to improve the level of compliance with privacy laws – but if this isn’t considered then exposing these services can be problematic.  As part of producing the heat map the business and technology risks associated with capabilities should be evaluated.
  3. Development Risks - There are new skills and development processes associated with building services well. To gain the business and technology benefits of SOA, SaaS, and cloud computing the development teams must have the appropriate skills in analysis, design, and development. There are new tools, new deployment patterns, and new implementation risks that must be understood and mitigated. Ensuring the appropriate training, environments and partnerships are established early on to establish the appropriate competencies and infrastructure are necessary to mitigate development risks.
  4. Delivery Risks - Finally, implementing SaaS and cloud computing requires a significant investment in infrastructure, new technology processes, and new development techniques. Leveraging cloud computing can dramatically reduce the initial costs while simultaneously reducing the associated risks.

A Cloud Computing Case Study

Now that we have talked about business capability modeling, and heat mapping and how helpful they can be in helping to prioritize decisions related to both unification and opportunity  maximization, now we want to step through some of the specifics of an actual company who has realized some of the benefits of a cloud computing strategy along the lines of what we talked about in Section 3.

In our case study, we have an actual company based in India with thousands of employees that offers software development services that help enterprises and public sector customers enhance their offerings.  We will call this company Contoso, and if at some point it becomes important to learn the actual name of the company and the people quoted, let us know and we can approach them to see if they are comfortable sharing their actual name.

Case Summary

In this case, Contoso chose to use the Microsoft Windows Azure™ Platform to deliver its applications on the Internet through a Microsoft® data center.  It is worth noting at this point that while there are other technology options for cloud computing solutions, we cannot speak to whether other solutions would deliver the same success that Contoso achieved in this specific situation.


  • Simplified application deployment
  • Flexible, cost-effective scalability
  • Reduced costs
  • Quick, inexpensive development
  • Enhanced industry specific (in this case government) services


Headquartered in Pune, India, and with operations in Asia, Europe, and North America, Contoso Systems provides software product development services to a wide range of customers in the telecommunications, life sciences, data infrastructure, and government sectors. With more than 6,000 employees, Contoso Systems offers services that help its customers enhance their offerings while reducing overall costs.

One of Contoso Systems’ primary offerings is an e-government solution that offers regional and local governments and agencies the ability to deliver services and interact with citizens and businesses electronically through a set of four Web-based applications that the company calls its e-governance suite. The suite forms a cohesive solution for public service, and encompasses complaint resolution, roads and infrastructure, census, and election management systems.

The Grievance Redressal System allows citizens to register and track incident reports to any government department. With the Roads and Infrastructure application, citizens can report issues related to roads and infrastructure, identifying specific locations with an online mapping tool. Registered hospitals, doctors, and other authorized personnel can use the Census Department application to register births and deaths. The Election Office application interacts with Census Department to maintain up-to-date voter lists and help authorities manage and schedule elections.

Contoso Systems developed the e-governance suite using Microsoft® ASP.NET and Microsoft SQL Server® database management software, offering the separate components as client-based software applications, hosted in the customer’s own data center. However, the company found that its ability to promote its e-governance solution was often limited by the technology capabilities of local governments.

In India, many local and regional governments lack the IT infrastructure necessary to fully deploy Contoso Systems e-government applications as on-premises software solutions. Even where governments or agencies have the funding available to develop a high-performance server environment, they may lack the technology expertise to adequately manage the networking, redundancy and other infrastructure issues that can add additional costs. Moreover, they may not want the technology capacity or expertise, needing instead to focus on delivering government services.

Contoso Systems knew that its solution would enhance those governments’ ability to provide services. The company needed a way to deliver its e-governance suite to local governments without requiring them to invest heavily in new IT infrastructure and personnel. It wanted to offer customers a way to scale the solution up or down, adding or removing component applications, computing capacity, or data storage as necessary— quickly, easily, and cost effectively.

The company wanted to provide potential customers that didn’t already have a high-performance infrastructure a way to test the solution, and it wanted to offer them a way to pay just for what they used, as they used it. At the same time, Contoso Systems had already made a significant investment in developing its e-governance suite; it needed a new delivery model that it could develop efficiently, without having to significantly re-engineer the solution.


Contoso Systems decided to develop a solution  that would host its e-governance suite on the Internet through a data center  - an application delivery system sometimes referred to as “cloud” computing.  The company chose the Windows Azure™ platform, an Internet-scale cloud services platform that is hosted in Microsoft data centers, for its reliable high availability and scalability to match usage needs.

Contoso Systems is using the Windows Azure cloud services operating system, the development, service hosting, and service management environment for the Windows Azure platform, to provide on-demand computing and storage capacity for its Web applications.  It will use the Microsoft SQL Azure™ database as a service to store and manage application data, and application users will be able to store files and images using the Blob Storage feature in the Windows Azure platform. With Live Services, they can search for information using Bing™, and identify locations with Bing maps for Enterprise.

In addition to the four core components of the e-governance suite, Contoso Systems will deploy its own Tenant Provisioning System (TPS) in the Windows Azure environment. With the TPS, Contoso Systems will provision specific components to individual customers (or tenant applications). Contoso Systems will deploy each tenant application on individual Windows Azure project accounts, automatically isolating each tenant application from the others, enhancing security and scalability for each tenant.

With the e-governance solution deployed on the Windows Azure platform, local governments now can pay only for the applications they need in the form of a monthly subscription, rather than investing up front in an on-premises infrastructure. Contoso Systems will use the TPS to manage auditing and billing for individual subscribers, and customers will be able to provide feedback to system administrators if they want to modify their subscription. “From the customer’s perspective, they have much more flexibility,” says a Senior Project Manager at Contoso Systems. “As they move forward, they can easily purchase additional applications or discontinue applications they do not need.”

Contoso System uses SQL Azure to store the e-governance application database and a configuration database. Log-in details and attachments uploaded by users are stored using Windows Azure Storage Tables and the Blob Storage features. The system uses the Service Bus feature in Windows Azure to connect applications within the e-governance suite and share data between applications.

Because Contoso Systems developed its original e-governance solution using ASP.NET and SQL Server, company developers were able to move the suite to the Windows Azure platform with minimum effort. For instance, developers used SQL scripts to move the existing SQL Server schema to the SQL Azure database. “Because we had worked with traditional on-premises SQL Server software, we were able to save a significant amount of time migrating the existing applications to SQL Azure,” says a Technical Lead at Contoso Systems. “We were able to minimize our learning curve and make the overall transition very smooth.”

With the Windows Azure platform, Contoso Systems can offer its e-governance applications to local governments while reducing capital expenditures for itself and its customers. Governments can quickly test and deploy the applications and scale up and down as necessary, paying only for what they need when they need it. Due to the ease of development, Contoso Systems brought its new solution-delivery model to market quickly, and can now offer its e-governance applications to more customers.

Azure Benefit: Simplified Application Deployment

By hosting its solution on the Internet through Microsoft data centers, Contoso Systems can deploy its e-governance suite to customers that don’t have their own server infrastructures, and the company can offer a hosted solution without having to set up an infrastructure of its own. Prospective customers can evaluate the solution without having to deploy the application on-premises, and Contoso Systems can provision applications to new customers approximately 50 percent faster using its new system.

“With Windows Azure, we can deploy the applications to new customers very easily, while reducing our overhead significantly,” says a Senior Project Manager. “Trial runs are very easy. A customer can simply subscribe as a trial user for one month, and that’s it.”

Azure Benefit: Flexible, Cost-Effective Scalability

Because Microsoft data centers offer high availability and scalability, Contoso Systems can upgrade configurations for customers easily and quickly, adding or removing components of its e-governance solution per a customer’s need. And with the vast computing power offered through the Windows Azure platform, customers have the capability to handle varying loads without a full capital investment.

Rather than investing in server capacity for peak loads and then underutilizing that capacity during other periods, customers can pay just for the capacity they need as they need it. “For instance,” says another Contoso employee, “if an election is coming up, we can add more instances of a customer’s Election Office application, provide them more computing power, and they only have to pay for it during that period.”

Azure Benefit: Reduced Costs

Governments that subscribe to the Contoso Systems e-governance solution will be able to manage their costs more effectively by minimizing their capital investment, reducing operating costs by avoiding infrastructure-maintenance overhead, and paying only for what they use as they use it.

For a given set of features in the e-governance suite, a customer might have to spend U.S. $24,000 in capital costs and as much as $60,000 in annual maintenance overhead. With Azure, it can completely forego capital and maintenance expenditures, paying only service fees that might be less than U.S. $10,000 a year.

“With Windows Azure, our customers don’t have to invest up front in infrastructure or hosting services,” says a Senior Project Manager. “And because they can pay as they go, their budgeting is much easier.”

Because Contoso Systems can provision and manage the solution on the Windows Azure platform, it will increase profitability and reduce costs by as much as 70 percent as it offers its e-governance suite to more customers. “By delivering our e-governance solution through the Windows Azure platform, we will get more business from more customers,” says a Senior Project Manager. “And adding, managing, and billing customers through the platform will enhance efficiency and reduce costs.”

Azure Benefit: Quick, Inexpensive Development

Because they could take advantage of existing skills, developers at Contoso Systems spent less time learning how to work with the Windows Azure platform, reducing the time it took to deploy the e-governance solution with Windows Azure.  And because they did not have to configure an infrastructure to support the deployment, they could focus on the business logic and design of the applications. “Without Windows Azure, we might have spent 25 percent more time in process development,” says a Senior Project Manager.


By subscribing to Contoso Systems e-governance solutions, local governments in India and elsewhere can provide government services and interact with constituents more effectively. They can provide convenient access to services, empower citizens with information, and enhance government transparency and accountability, while reducing costs, simplifying operations, and improving efficiency.

Conclusion and Recommendations

Contoso is a great example of an organization that has realized simplified application deployment, flexible and cost effective scalability, reduced costs, and focused development with cloud services, in this instance using Microsoft technology.  As they demonstrated SOA, SaaS, and cloud computing introduce new technology opportunities that can produce new opportunities for the business. Along with these new technology opportunities come a new set of business and technical risks that must be managed to ensure the technical and business opportunities are achieved.

As we have proven over more than six years of work, the analysis of the needs of the business - through business capabilities analysis and heat mapping and prioritization - snaps to the IT conversation at the technical architecture level. We made this point in the June 2008 article we co-authored in the Harvard Business Review, “The Next Revolution in Productivity”, “With the heat map of activities in hand, managers will have much or most of the information they need to design a new operating model.”

So cloud computing has offered new benefits in speed, cost, and scalability. Business capability modeling has helped align technology with the strategic direction of business.  When coupled together these promise to enable an organization moving to SOA and SaaS to optimize the return on investment and time to value. So our specific recommendations for next step are really made up of two largely parallel paths:

Begin the Business Capabilities Analysis

This is the first step in getting heat maps and drive discussions about unification and opportunity maximization.  It is very likely that your organization already employs complementary methods such as process re-engineering, Six Sigma, Lean, or another, so we recommend that you start using business capabilities in a small, contained area so that you can understand how it differs from, and complements these other methods. There are resources available to help you with this, beyond our Harvard Business Review article and the book Rethink, so please contact us with your questions and comments.

Define a Technology Roadmap

Be clear about what the current technical architecture is, how much of it has clear service definitions, and then look at the various technologies that will complement existing legacy solutions, where appropriate, but also help the organization down the path of cloud computing, in a way that best aligns with goals of the organization and the needs of the customers.  If you have questions about technical architecture and cloud computing, please let us know.

Maintain the Roadmap

To close out our analogy about repairing an airplane in mid-flight – even with the right tools you need a pretty clear idea of how to fix the plane while keeping the plane in the air.  As you progress through the transition different ways to leverage SOA, SaaS and Cloud Computing to deliver business value will become apparent, you will identify different situation specific risks, and your value performance gaps will change.  The Value Performance Heat Map is very low friction to keep current and the effort will help maintain an explicit understanding of what is important to the business and why.  SOA, SaaS, and cloud computing are the right tools and capability analysis provides a clear map that can be used to clarify the value and risk associated with your efforts.


Ric Merrifield

Ric Merrifield leads business architecture endeavors at Microsoft Corporation in Redmond, Washington, and is a co-author of “The Next Revolution in Productivity” June, 2008, Harvard Business Review.  Merrifield is also the author of the book Rethink – a Business Manifesto for Cutting Costs and Boosting Innovation, FT Press 2009.

Dennis Stevens is the CEO of Synaptus, a consulting firm based in Norcross, GA and is a co-author of “The Next Revolution in Productivity” June, 2008, Harvard Business Review. Stevens was also co-author of the July Cutter Consortium report Rethinking the Agile Enterprise and is currently writing Value Driven Agile Adoption: Scaling Agility to the Enterprise for Addison-Wesley due by year-end 2010.