"How are you?"
This question precedes the majority of our conversations. It is our way of figuring out how our conversational counterparts are feeling. If they are happy, sad, or neutral - we cultivate a discussion by matching or addressing their sentiment.
This is especially true in fundraising. We spend so much time trying to decode the emotions and motivations of our donors in order to make interactions and appeals more meaningful. Get this wrong, and you might put a sour taste in a donor's mouth and ruin a relationship. Doing this with people you don't already know is difficult. Doing it at scale is near impossible without the help of technology.
Within artificial intelligence and natural language processing, there is a branch of exploration called "sentiment analysis". This looks at a stream of data (in our case, social media posts and tweets) and tries to statistically figure out how someone is feeling and how they are motivated. We have implemented this important feature within our solutions at Gravyty:
Imagine the implications of doing this for nonprofit fundraising. If we know someone is motivated by hope, we might tailor our asks around something like: "We need $25,000 to continue to grow this amazing new initiative". If they are motivated by anxiety or fear, we might instead say: "We need $25,000 or the lights are going to go off." These are wildly different approaches, deduced through the science of statistics and AI, and offered to fundraisers to bring their artistry in building relationships.
At Gravyty, we took an anonymized look at sentiment analysis across 12,000 donors within our system to unearth some interesting findings:
- There is a pretty significant difference between the highest and lowest quartiles of lifetime giving. The sentiment within the highest quartile of lifetime giving is significantly more "positive" than the lowest quartile. This means that your highest lifetime giving donors are more likely to be motivated by hope and vision for the future in general, and should be cultivated accordingly.
- However, the same is not true of people who gave more recently. "Positive" sentiment people are associated with donors who give over a longer period of time (e.g. greater than 3 years). We posit that consistency/duration of donations represents a person that is hopeful about the future in general, and the fundraiser's job is then to talk about how the organization will continue to advance the future with their support.
- Based on Twitter data, the supporters of hospitals are significantly more motivated by "fear/anxiety" than other nonprofit organizations.
- In general, the distribution of sentiments of all donors we analyzed is uniform. Roughly one third of people are negative, one third are neutral, and one third are positive.
Some of these findings lead to fairly logical conclusions, and others have opened up further opportunities for research and exploration by our development team. Regardless, the power of sentiment analysis - to understand some facet of your donor before interacting with them - is incredibly powerful. It is another way that Gravyty is leveraging behind-the-scenes artificial intelligence to tee up insights for fundraisers and nonprofit organizations in immediately useful ways.