By Amy DeWolf
Coined “the next natural resource,” big data has endless possibilities to improve government operations, lower costs, reduce waste, fraud and abuse, increase efficiencies, and enable agencies to have a holistic view of citizens. From tax agencies to social programs, government organizations are benefiting from looking at their data, all of their data, in new ways. For example, according to an IBM white paper, “a regional social programs agency deployed big data analytics to help eliminate fraud and redundancy by identifying relationships between benefit recipients and programs.” The agency was able to almost immediately reduce waste by 11 millions dollars once the new system was set up.
According to the same IBM white paper, “Big data inverts the traditional analytics model.” Previously, organizations determined what questions they wanted to ask based on data they knew they had. The analytics system was structured around the data required to answer those specific questions. In contrast, big data analytics systems “are designed to enable creative discovery based on wide varieties and large volumes of data, some of which users may not anticipate in advance.” These systems allow for “exploratory analytics” and the opportunity to ask new questions.
So how can government agencies take advantage of big data? The white paper recommended 5 steps:
1. Define the use case that will be part of the initial big data deployment, key business objectives, and specific mission requirements and strategies. It’s important to get a quick win to build momentum and gain additional buy-in, so look for use cases with high ROI potential.
2. Assess your agency’s available data and technical capabilities against the data and determine what additional data or capabilities are needed.
3. Plan for deployment by determining what big data entry points the project will emphasize (volume, velocity or variety of data), what deployment pattern is best, which existing IT investments can be integrated, and how security and policy requirements will be fulfilled. The technical architecture should be designed to support the current project, but also capable of scaling.
4. Execute the planned project by deploying elements of a big data platform based on the recommended plan and architecture.
5. Review and report on roadmap steps throughout the process, and adjust plans based on findings and lessons learned. Assess cost and timeline results and measure return on investment to help plan future projects.
What else would you add to the list?