A 5-Step Process for Data Analysis
May 9, 2017 Testing PremiumContent
We all have data and we all know we need to use it more effectively. Here are five steps for getting started with data analysis, adapted from a talk by Karen Blonder of the Association for Unmanned Vehicle Systems International, Debbie King of Association Analytics, and Nikki Vann of the North Carolina Association of CPAs at the American Society of Association Executives (ASAE) Technology Conference and Expo this past December.
Step 1: Scope
This is the planning stage. You need to prioritize what you want to analyze and determine the questions you want to ask — often called “metrics.” It’s also important to think about what success looks like.
For example, when should you begin to see positive changes? What benchmarks should you set for your progress? What is the ultimate outcome and when do you expect to achieve it? At this stage you should also make sure you have the people and money you need to collect and analyze the data and make sure you spend time educating staff members across the organization on what you’re doing and their role so that you have the buy-in to keep it running smoothly.
Step 2: Collect
You probably have a lot of potential data sources. Inventory them and decide what to include. Make sure all of your business terms are defined and that staff members can easily consult those definitions to better understand what you mean. If possible, integrate your data sources so that it’s easy to pull the information you need and build reports that incorporate multiple sources.
Step 3: Clean
Your data is worthless if people don’t trust it. To clean your data, identify anomalies and correct duplicates, missing entries, or inconsistent data. Put in place standards that ensure data entry is consistent, but also expect that you’ll need to do regular maintenance over time.
Step 4: Analyze
The biggest reason why data is such a challenge for nonprofits is that it takes a lot of time and energy (steps 1-3) to get to analysis. But once you’re here, you open up a lot of new possibilities. Data visualizations, such as graphs in dashboards, can help you literally see trends, patterns, and outliers in your data. Once you have a solid data set, you can also explore it. Ask new questions and shuffle things around to see what you might uncover — don’t be afraid to just try something. You don’t have to be a data expert to gain useful insights from your existing data.
Step 5: Act
Once you have the data and understand it, what can you do with it? Visualizations of your data might make a compelling case to a funder. You and your colleagues might see opportunities to make programmatic improvements. You might have compelling evidence to end or overhaul an existing program or start up a new program to address an unmet need.
The answers are in the data. Just make sure that whatever actions you take can be measured so that you can understand how success your new ideas are. This might mean going back to steps one and two to revise your metrics and collect new data.