Seven Best Practices for Cloud Computing

If you are a CIO, business analyst, or project manager responsible for getting your enterprise to the cloud, read these tips before diving in.

by Sharam Sasson

Anyone who has been working for more than a decade understands the cyclical nature of the economy. Prosperous times are followed by market corrections, and a downturn is no time for stifled investment in innovation. As a result, cloud computing is on everyone’s mind and a topic in every boardroom - promising lower costs, robust scalability and autonomy from constrained IT and development resources.

Any enterprise considering a move to the cloud must understand that the perceived benefits can be short-lived without a plan that places cloud computing in the context of its overall business strategy and affects security, performance, and connectivity.

In particular, large organizations will need to be able to integrate cloud computing into existing IT systems and applications. Very few organizations are ready or willing to start from scratch and most will not move all of their business processes fully to the cloud at once. This makes it essential to plan for the integration challenges ahead.

In fact, there may be business processes and applications that never move off-premise. This means we may always need to connect on-premise to on-premise, as well as on-premise to cloud, and cloud-to-cloud, adding more complexity to integration than ever before. With enterprise IT increasingly freed from ongoing infrastructure management to focus on new areas for innovation, business integration demands simplification. With this in mind, let’s look more closely at how we advise our own customers considering the cloud.

Best Practice #1: Create a strategy and a set of (realistic) goals upfront

Some businesses are jumping in without articulating a long-term cloud strategy and how it relates to their overall business. As with any project, establish realistic goals and priorities, a clear budget and deadline, as well as a shared understanding of what resources are available for implementation and maintenance. Although cloud computing promises significant ROI -- productivity gains of 50 percent or more -- keeping complexity and cost to a minimum requires planning and strategy.

Best Practice #2: Learn from the mistakes of those who came before you

Early adopters took a standalone approach to cloud computing. The services were readily available, easy to consume, and economical. Implementation challenges were few. However, for enterprises where traditional IT infrastructure often serves core business operations, the “detached” cloud might deliver only short-term value and potentially require future re-implementation or migration. Although a standalone approach risks creating silo-ed applications, an integrated cloud strategy will deliver long-term results.

Best Practice #3: Avoid upfront costs

One reason businesses turn to cloud computing and software-as-a-service (SaaS) is to mitigate risk. With a pay-as-you-go model, you can simply turn the system off if it is not working for you, and you only pay for what you use. Be sure not to jeopardize return on investment with expensive software licenses -- your integration solution, like every other part of your cloud computing investment, should use a flexible subscription model, too.

Best Practice #4: Get serious about autonomy

Integration of on-premise applications has traditionally required a team of IT specialists who have a deep understanding of underlying application frameworks and processes. SaaS applications are designed to be managed by business users -- non-domain experts who will need to quickly and easily connect data with other enterprise systems. Cloud integration should complement the model by minimizing development, implementation, and maintenance resources, allowing users to focus on their core business.

Best Practice #5: Address security concerns

According to analysts, nearly 75 percent of CIO and IT executives cite security as their number one concern when it comes to cloud computing. Because integrated cloud computing involves moving sensitive data between the cloud and on-premise networks, guaranteeing security is vital. When vetting an integration solution, determine which standards are supported for securing the data in transit. Keep in mind that as enterprises move more processes to the cloud, the volume of sensitive data flowing to and from the cloud increases.

Best Practice #6: Ensure performance and availability of the data

Today, popular cloud computing services provide levels of availability and performance that outperform internal infrastructure, in many cases 99.9 percent uptime or better. When designing a cloud strategy, identify integration requirements for each system (real-time, near real-time, batch), and determine the number of simultaneous requests to be handled, and specify all special architecture requirements. Success depends on ensuring information will not be lost if the cloud or on-premise source goes down. (Yes, even the cloud needs a backup plan.)

Best Practice #7: Maximize connectivity options

Cloud computing has become a loose definition for services on the Web: everything from SaaS and platform-as-a-service (PaaS), to Web-based utility and storage solutions, and emerging Web 2.0 properties such as Google Docs, LinkedIn, and Twitter. According to research1 from Saugatuck Technology, by the end of 2010, one-quarter of business process improvement initiatives will include integration of information from enterprise social computing solutions. Connectivity requirements will continue to evolve beyond standard enterprise applications, legacy systems, and databases, to modern Web service and Web 2.0 APIs.

Solving the Integration Puzzle

Current economic conditions are driving businesses toward cloud computing, and with good reason. No one can argue that the cloud doesn’t deliver value. It is already proven to be a disruptive force. However, the cloud’s “Achilles heel” is a lack of integration with the rest of the enterprise. Realizing its full potential relies, for the foreseeable future, on integrating data in the cloud with on-premise applications and databases.

Today’s enterprise cloud initiatives require decoupled data systems working together -- without the need for personnel and other resources to set up and maintain them -- making integration key to success. Most companies cannot and will not abandon their previous IT investments to make the leap to the cloud all at once. Instead, there is more likely to be a gradual shift in business processes to the cloud over time, similar by nature to a perpetual proof of concept. As the cloud delivers on its promise, more processes will be shifted to this computing model. Complexity and diminished ROI will be the consequence when long-term strategy and goals are not implemented in advance. Put simply: integration needs to be a forethought, not an afterthought.

Sharam Sasson is an entrepreneur and experienced software executive. He has co-founded and led three successful technology companies: Scopus Technology, Extensity and, currently, Jitterbit, a data and application integration company that offers open source software for cloud, SaaS, application, and data and business process integration. You can contact the author at

Must Read Articles