Best Practices for Implementing Government Cloud

When Otto Doll used to discuss the implementation of new technology as the CIO for the city of Minneapolis, he’d talk about specific projects “the way the transportation department would talk about roads,” he says.

“It’s important to talk to stakeholders about the roadmap to modernizing their application, what role cloud-based solutions play and how the technology organization is prepared to implement it,” says Doll, now a Center for Digital Government (CDG) senior fellow.

When implementing cloud projects, government IT leaders must focus on enterprise needs, business processes, and the specific requirements undergirding each new application or use case. Among the strategies for success:

Consider cloud projects as part of a broader roadmap. It’s important for IT leaders to have an overarching cloud strategy and communicate it to business units. “A CIO has to think through their portfolio of all of the systems they have,” says CDG Vice President Teri Takai. Doing so provides the context for the more specific considerations involving each project. It also reinforces the importance of enterprise governance and management structures that address security and integration standards.

Be cloud smart, not cloud first. Determine whether solutions that meet department or agency business needs are available only through on-premises solutions (increasingly rare, but at times still the case), only through cloud solutions or a combination of the two. When both cloud- and non-cloud environments are feasible, it’s important to be responsive to customer needs. For example, some public-facing government departments may worry about the perceptions that may arise if sensitive citizen data is posted to the cloud, require additional levels of security or assurance about where data is stored, or have other requirements for which on-premises solutions remain a viable option.

Begin with business processes. Effective business processes drive effective cloud implementations. As a result, governments should assess the workflows being shifted to the cloud before they try to implement the technology. “Try to lead with business process reengineering types of questions,” advises CDG Senior Fellow Sergio Paneque, who has held senior procurement positions for the city and state of New York. “If we’re going to go down this road, what is the process we’re reengineering?

“I don’t think government is particularly good at this,“ Paneque adds, “but I think it will become more and more important” — particularly as demographic shifts and retirements leave some organizations lacking institutional knowledge of key workflows.

Develop an exit plan — right from the start. Both technology and the business needs of government agencies and the constituents they serve are evolving at a breakneck pace. That means a cloud solution that makes sense today may need to move in the future — to a different cloud environment, or even back to an on-premises data center. That’s why it’s important to develop “an exit strategy,” Doll says. “You don’t want to get locked into a given cloud provider. Position yourself so you can go from place to place, because those scenarios will come up from time to time.”

The time to do that is when initially evaluating cloud providers, according to CDG Senior Fellow Brenda Decker, who previously served as CIO for the state of Nebraska. “When I sat down and talked to cloud vendors, the partners I felt comfortable with were the ones who could tell me how I get out as easily as I could get in,” she says.

Don’t overlook key details. Governments must also evaluate each project to ensure a cloud environment can support key operational needs, including latency, integration, security and compliance considerations. It’s also important to ensure the business unit and IT staff have the skillsets and capacity to manage the new application and the cloud environment it will run in.

Focus on “quick wins.” That was a strategy many government IT leaders used to win initial support for cloud in the enterprise, but you can also use it to foster support and confidence in individual cloud projects. Identify opportunities to quickly launch components of a cloud application — one specific use case or a pilot involving a relatively small number of users, for example — before releasing the full version of the project. Doing so provides support for agile development methodologies in larger projects, and such projects could even potentially be hosted in a private cloud until demand meets the level where the resources of a larger, public cloud is needed.

Address “rogue IT.” In legacy environments, individual users or departments would sometimes purchase solutions without the support of enterprise technology staff. Now, the simplicity of purchasing cloud solutions can create similar challenges. When individual departments or agencies select their own cloud solutions, “You lose control of spending; you lose control of costs; and you can jeopardize your cybersecurity position,” Takai cautions. IT leaders also must ensure they have a role in not just evaluating cloud solutions for line-of-business needs, but also in implementing them so they can understand how they connect with other enterprise solutions.

Remember that everything’s connected. Existing datasets may draw information from multiple sources, and applications may provide information to other solutions. It’s important to understand and address these connections as part of the cloud migration plan.

“A lot of times the thing governments don’t think about is what pieces of information are tied to what,” says Decker. “That’s incumbent on anyone who goes to the cloud.”