Hi Jasmine,
The short answer is, AI and ML have piqued the curiosity of the education space but much of the work the institutions are doing is in areas such as predicting admissions etc. In the fundraising space, most major prospect research tools claim to use AI + ML to predict/score prospects, and some CRM platforms have canned algorithms for predictions but in this case, the end-user has little control over the algorithms and techniques used by the vendors. Besides, the accuracy of the predictions suffers because the approach here is to apply a one-size-fits-all type solution to everyone's data set.
I am working with some organizations where they feel limited by the AI offered by the tools and are venturing into performing their own analysis and machine modeling. Each has been a custom analysis project with its own dataset and involves applying suitable models to determine the prospect strength, understanding the donor base better, and predict donor behavior (to guide solicitation, donation amount etc.)
Creating a software tool to analyze contact reports and proposals to predict donor behavior is like creating one drug to cure all illnesses: essentially impossible. A skilled data analyst will take a look at the data and apply suitable cleaning and modeling techniques based on the data. Therefore this becomes a data-centric activity.
Regarding your question on what data will be required for successful AI implementation: in general, the more information relative to your prospects/donors and their activities you have, the better the accuracy of your predictions. There is definitely an optimal ratio of the number of records versus the variety of information on the donors, but a skilled machine learning analyst will be able to massage the available data to increase the accuracy of the predictions.
If you can elaborate on what your AI goals are and where you are in your journey, then I would be happy to further clarify.
P.S. Another way fundraisers use AI is to use natural language processing to interface with donors and prospects during events or on an ongoing basis. This is a different game from machine learning and predictions, but it's an area that is rapidly developing. Hope this helps.
Thank you,
Medha
------------------------------
Medha Nanal
CEO & Principal, Top Cloud Consulting
(Data, Systems, Analytics)
medhananal@topcloudconsult.com------------------------------
Original Message:
Sent: 01-23-2023 03:43 PM
From: Jasmine Dsouza
Subject: Artificial Intelligence/Machine Learning for fundraising
Has anyone tried to investigate and/or implement the use of Artificial Intelligence or Machine Learning in the fundraising space?
Do you use any software to analyze your contact reports and proposals to predict donor behavior?
We are trying to gather information on what data would be required for a successful AI implementation.
Thank you!
------------------------------
Jasmine Dsouza
Advancement Services
Temple University
jasmine.dsouza@temple.edu
------------------------------