A better national government means finding improved ways to solve the nation’s challenges. Each agency, tasked with a different mission, must use its allocated budget in the most efficient way possible to have the biggest positive impact on our society. That could mean making more accurate predictions of disease pandemics at the CDC, improving the intelligence collection capabilities at the FBI or better analyzing the effects of global warming at the EPA.
To say that data is growing exponentially is stating the obvious. As data continues to saturate your agency’s operations, efficiency will become dependent on making smarter, more informed decisions. With over 44 zetabytes of data predicted to be produced by 2020, having the right business intelligence strategy and platform in place to analyze available data and access more intelligent insights is the first step to a more efficient decision-making process and achieving mission-driven success.
To achieve holistic, 360-degree analytical insights to make more informed decisions and drive your mission, start your data analytics strategy by answering the following questions:
1. What are your objectives, goals and KPIs?
Specifically, what operational goals will help you achieve your agency mission and which key performance indicators (KPIs) will help you identify successful performance? These three factors will form the base of your business intelligence (BI) operations to lead your agency towards more favorable outcomes.
2. What are the challenges standing in the way of reaching your goals and the answers to what questions could help you overcome those challenges?
Formulate questions that would require specific answers, which would ultimately drive you to solve your challenges and reach your goals. For example, if your goal is to make sure that agency programs remain operational for the following fiscal year, and your challenge is a limited budget and expanding programs, your question might be “what cost savings options are available that would result in the least interruptions to operations?”
3. What specific metrics define answers to your questions?
Identify which metrics you need to collect: real-life data figures or quantitative and qualitative manifestations of parts that, when put together, produce a complete answer to your question. For example, when tracking cost savings, your metrics might include: cost for accessing an item, item usage patterns, total item value, levels of access, effect on operations, etc.
4. What data sources are these metrics coming from?
The likelihood is you have a wide variety of data coming in from a wide variety of sources: website metrics, emails, recorded phone calls with feedback, requests and complaints, sleuth of historical data, data from edge capture devices (video, images, etc), other sensor data (outside thermometers and barometers if your mission is environment-related, for example), etc. However, only select data sets will contain useful information to address each question.
5. Where is your current data coming from and is it interoperable?
As you understand what data you need to be collecting and analyzing, you will see that it comes from many different sources, systems, sensors and “things”- many of which might not talk to each other. Since siloed data sets in and of themselves cannot produce insightful analytics or a holistic picture, look for a data integration solution that can blend structured and unstructured, and operational and IT systems data sets, plus the ability to scale as your data volumes increase.
6. What data visualization formats will help you understand your data?
When analyzing and extracting insights from your now-integrated data pool, dashboards, charts, graphs, geospatial maps and other visualization techniques will help you make the most sense of a large influx of randomized data points. Work with your Analytics provider to help you design visualizations around your KPIs. For example, when looking for operational cost savings, one way to interpret the data might be with a plot chart of items mapped according to their cost against their perceived value or access patterns.
Extra: Diagnostic, descriptive, predictive or prescriptive- Which analytics do you need?
The larger your federal agency, and the more data that you collect, the more you will benefit from all of the above. What is the difference?
Diagnostic analytics can help your agency understand its past performance and provide historical context to your current state of being. Descriptive analytics provide dashboards and insights into real-time information based on incoming data. Predictive analytics forecast the likelihood of future events. Prescriptive analytics use simulation algorithms to help you decide on a course of action. When selecting an analytics provider, ask them if they have experience with either of those data modeling techniques and how they can apply them to your agency’s mission.