Robert Decker, DMI Executive Director, Cloud Services

7 Do’s and Don’ts for a Successful Cloud Architecture

2019_Decker 7 Do’s and Don’ts for a Successful Cloud Architecture
Robert Decker, DMI Executive Director, Cloud Services

To empower businesses to be more agile and deliver services closer to the customer, the evolutionary shift toward the cloud shows no signs of slowing down. In fact, Gartner Research predicts that by 2025, 80 percent of companies will have opted to shutter their traditional data centers. The International Data Corporation (IDC) recently upped its prediction for cloud IT infrastructure spending to more than $57 billion, a 21 percent increase over 2017.

Here are seven dos and don’ts for Infrastructure and operations (I&O) leaders to help ensure a successful transition.


  1. Do: Ask questions. Identifying a company’s cloud needs begins with gathering insights around the business tasks (IT issues, process enhancements, product offerings, customer satisfaction, etc.) that the cloud solution is attempting to resolve. The cloud architect must understand the operational factors necessary to maintain the business. Pinpointing the answers to such questions as, “Does the business process create spikes or bursts of activity impacting available cloud resources?” and “Is customer impact a priority?” are critical.
  1. Do: Prioritize compliance. Compliance is never optional. It’s always better to understand compliance mandates before developing a cloud infrastructure than to go back to the drawing board after-the-fact. For example, are there specific industrial or governmental requirements that demand high-levels of data protection (such as in the EU) or long-term availability? 
  1. Do: Create a unique solution that fits your organization. With cloud service providers offering a growing portfolio of options and features, selecting the most effective and efficient solution can prove daunting. While a standard cloud solution might be convenient, a cloud architect should develop a tailored architecture to meet specific organizational needs. The cloud impacts an organization on all levels, which is why it’s important to gather support and feedback from all as many business stakeholders before digging in. For instance, a good cloud architecture will take into account departmental structures, as well as methodologies like Agile, DevOps or CI/CD. 
  1. Do: Keep the big picture in mind. The architecture of a cloud solution must align to or replace the current IT strategy, as it has the potential to substantially change strategic direction and bring a new era of innovation for the enterprise. Determining the cloud impact to existing IT infrastructure management and resources is important. Any cloud architect should ensure to incorporate both existing resources, as well as fully-outsourced Managed Services into plans for the architecture. 
  1. Don’t: Build artificial limitations: The old standard of building infrastructure with over-capacity or limiting development resources because of cost constraints must be eliminated. IT must embrace scalability and on-demand resources to allow and drive innovation for the business. 
  1. Don’t: Make assumptions. In light of the effort companies undertake when adopting a cloud strategy and architecture, it can be easy to assume a cloud architecture is meeting demand. But without accurate and timely reporting of the cloud, as well as routine interactions with the business and other IT organizations, IT can slip into a false sense of stability. That’s why it’s crucial to continue monitoring performance and gathering feedback. It’s impossible to know every detail and anticipate every problem that may arise. 
  1. Don’t: Remain stagnant. New cloud features and functionalities are released at a staggering rate (almost daily). IT organizations have to stay abreast of existing features while integrating new features into their business model. IT must continue to gain insights and identify opportunities to bring new solutions to the business to enhance cloud architecture. By leveraging strategic partners, such as cloud providers, organizations can get a glimpse into up-and-coming cloud features, such as AI. Once cloud is established, adoption is not complete. Daily management and routine optimization are required to ensure the cloud is meeting business and user demands while staying resource-efficient and cost-effective.