Predictive analytics is used to calculate your insurance rates. It even tells retailers how to design their stores. Whether you realize it or not, the uses for predictive analytics aren’t only expanding but becoming more and more ingrained in our daily lives. The world of grants management is no different.
What is Predictive Analytics?
The term “predictive analytics” means different things to different people depending on the context. So what does it mean for grantmakers?
Basically, predictive analytics takes historical information and combines it with data science and math to comment on future events. For granters, that makes it the ideal tool to leverage in order to increase the impact of funding and improve outcomes. How? By using prediction and segmentation.
Forecasting outcomes is an important part of what predictive analytics does. By looking at your past granting activities, it can help you make better-informed funding decisions in the future.
Let’s suppose you’re unsure that the $500,000 you allot in annual granting for animal rescue group support across your region is being used effectively. By creating a predictive analytics model, you can see how funds were used, and evaluate if you should commit more funding or less, or create different criteria for the grants you offer.
Simply put, segmentation is the ability to group people, places, or things that have common traits, like applicants, donors, or funding sources (government support, endowments, etc.). For example, applicants seeking grants to improve primary education could be one segment. Those seeking grants to improve middle school education would be another, and so on.
The Challenge of Extracting Data from Text
For most grantmakers, the biggest challenge with their data is that 80% or more of what’s collected from grant applications is in text format. That’s a problem for several reasons:
- It’s difficult to pull information from text and turn it into workable data because the text could be interpreted in many different ways.
- There’s no real way to evaluate your current applicants in an objective and standardized fashion.
- Combining the information with other (quantitative) data to create predictive models can be nearly impossible.
Take, for example, the sentence “Tim flew to Chicago to meat with the review panel.” Most would be able to understand it, but there are issues that would make analyzing the data difficult:
- Does Tim have the superpower of flight or did he take an airplane?
- Is Chicago the city or the band? We only know this from the context of the sentence.
- Is there really meat involved or is this a spelling mistake? Most would assume the latter or mentally substitute the correct word.
- Is the review panel a group of people or some intelligent wall covering?
The ability to untangle and get meaning from a simple sentence doesn’t always work, so we need another approach to help us understand and quantify the text. By using text analytics, you will have the tools you need to accomplish this and gain a deeper understanding of the information you gather on your grantees.
Where Do We Go From Here?
Using predictive analytics in modern grantmaking and philanthropy is more a question of how, not when. For most organizations, the journey starts with a particular objective or question, and an interest in using analytics to further their mission or objectives.
Getting Started With a Predictive Project in Grantmaking
To help get you started, SmartSimple and PEAK Grantmaking are hosting a webinar on April 25: The Future is Now: Incorporating Predictive Analytics into Outcomes Analysis.
This webinar will introduce predictive analytics and how it benefits grants management. We’ll also set the stage to take a deeper dive into different aspects of analytics. This deeper dive will come through three additional webinars that will be made available for members to view in the weeks following:
- Kicking Around the Crystal Ball – Predicting the Future: A look at what data we need from your granting activities, and how to collect it. Then we’ll explore available techniques to create accurate, actionable predictions using that data.
- Birds of a Feather – Grouping and Segmentation: We like to think we’re individuals, but, in reality, we can be grouped by things like size of our annual budget or communities we serve. This webinar will examine the use of segmentation, look at what’s required to create segments, and explore how to use this information to gain insight.
- Teaching Your Computer to Read – Text Analytics: From grant applications and reports to news feeds on Twitter, we live in a world of text. How do we analyze it and turn it into useable data about the grantees and nonprofits you fund? This webinar will show how you can extract greater meaning from your text.