How federal agencies embrace flexibility and bring visibility to the cloud
This blog draws from an Oct. 13 webinar, the third in a series hosted by the Partnership for Public Service and the IBM Center for The Business of Government, that featured two federal IT leaders: Andrea Norris, then chief information officer of the National Institutes of Health and Brock Webb from the Office of the Chief Information Officer at the U.S. Census Bureau. Read our blog posts, “The 3 stages of cloud adoption” and “Providing a secure cloud for agency applications,” for recaps of the first two webinars in the series.
One commonly touted benefit of moving operations to the cloud is savings. A 2019 Government Accountability Office report found that 13 of 16 agencies reviewed had saved $291 million by transitioning to the cloud. However, GAO noted, and reiterated in its latest analysis, that inconsistent tracking of cloud spending and savings data likely leads to inaccurate figures. Comparing the costs of running on-premise data centers with the costs of cloud service subscriptions and software licenses oversimplifies the investments needed to hire and train a cloud-capable workforce and underestimates the savings potential and impact of cloud-enabled work.
With that said, here are some best practices to maintain financial accountability while exploring and expanding your agency’s cloud environment, particularly a multi-cloud one, where the benefits of diverse services and platforms come with management challenges and cloud sprawl.
- Build cloud financial operations, or FinOps, into the cloud strategy from the beginning. Cloud migration projects, like any complex IT initiatives, are susceptible to running over budget and behind schedule. A FinOps model encourages all functions to take ownership in tracking their cloud usage and spending so that cloud operating decisions are data-driven and value-based. U.S. Census Bureau Technology Strategist Brock Webb emphasized how incorporating FinOps early in the agency’s journey was key to fostering collaboration: “We could show [teams] how they were using [the cloud] … help them make adjustments and work together to improve the cost.” Given that the departments of Energy and Veterans Affairs, as well as other agencies, are piloting a FinOps playbook, it’s safe to say that FinOps is the future of the federal cloud.
- Consider flexible, scalable, long-term partnerships with mission-aligned cloud service providers. Pay-as-you-use cloud computing, while intuitive, runs counter to the annual and often multiyear federal budgeting cycles. Agencies can struggle deciding between a single cloud model with fixed costs and services that comes with the risk of vendor lock-in; or a multi-cloud one where the benefits of diverse services and platforms can create management challenges and cloud sprawl. Anticipating a ‘tsunami’ of genomic, clinical and health data from and for universities and biomedical research institutions across the country, the National Institutes of Health opted for an unconventional model. Three vendors were offered flexible partnership agreements to service NIH and more than 100 NIH-funded institutions and train them on how to use cloud tools to generate, analyze and share data. By using economies of scale, the STRIDES Initiative has saved tens of millions of dollars and trained over 4,000 researchers in three years. According to former Chief Information Officer and Director for the Center for Information Technology at NIH, Andrea Norris, “[having] new ways of thinking about how you’re delivering cloud capabilities and [having partners] that are trusted and focused on your mission” are two elements critical to the program’s success and savings.
- Account for value-add for employees and customers, even if it can’t be explicitly counted. It is possible to quantify the dollar value of migrating workloads to the cloud for agency leadership, Congress, Technology Modernization Fund project evaluators and webinar audiences. Webb described a real scenario where IT worked with business mission partners and demonstrated the reduction of a two-to-three-week statistical record matching process down to a matter of hours using cloud computing power. Rather than extrapolate out time and cost savings for existing workloads, Webb and Norris place stock in leveraging the cloud to address the unprecedented. With the onset of the pandemic, NIH was able to share the genetic details of COVID-19 with researchers around the world in a single day through the STRIDES program, while the Census Bureau used software run on the cloud to distribute the COVID Household Pulse Survey and check in with the public safely. The value of storing, sharing and studying biomedical and census data through the cloud will only continue to grow. As Webb noted, the cloud gives you “the power to do things that you haven’t done yet.”
We look forward to sharing more insights for federal cloud leaders and practitioners with the release of our culminative cloud computing brief on Feb. 23, 2023. View event details and register to attend here.
Daniel Chenok is executive director at the IBM Center for The Business of Government. Trista Colbert is vice president and senior partner of hybrid cloud management at IBM. Will Kimball is a former intern on the Partnership’s Research, Evaluation and Modernizing Government team.