AI for Good: Preconfigured AI Predictions for Free with Einstein for Nonprofits

Artificial Intelligence algorithms of various kinds are embedded in almost all of our daily technology interactions. The shopping suggestions that you see on Facebook and Amazon are constantly learning about you and your preferences. That’s why the products you see listed are often uniquely interesting to you. The auto-complete suggestions in your texting app makes smart guesses based not only on its knowledge of language in general, but from what it has learned about you and how you write. The list goes on – I quickly generated a comprehensive list that included these examples by asking ChatGPT, “What are some examples of AI that everyone is already using?”

The nonprofit organizations that make the biggest impact in the future will be those that learn how to leverage AI tools today. If you already use Salesforce, there is a wide range of AI tools that you can start using to boost your org’s effectiveness.

Salesforce is staying abreast of the AI revolution by offering AI solutions that encompasses generative AI, predictive analytics, chatbots, and more. Furthermore, they are offering some free, pre-configured tools tailored for Nonprofit orgs to be able to get started with some of the newest AI tools, quickly and easily. 

First, a peek at some of the wider possibilities:

Good reports and charts are the best way to synthesize data into actionable insights. The hard parts are thinking of exactly what reports and charts you should have, and helping everyone to understand what they mean. Salesforce’s CRM Analytics tool can look at your data and actually suggest insightful charts. Not only that, but it will offer natural language explanations of what a chart means, and even suggest decisions or actions that you should take.

With another tool, Einstein Discovery, you define a goal, and AI will work on showing you how to get there. For example, tell Einstein Discovery that you would like to maximize the “Total Gifts” amount of every constituent in your org. AI will go to work analyzing trends and data points to understand what factors are the most strongly correlated with high Total Gifts. Based on results, you can adjust your fundraising to focus on the most impactful strategies.

We all know how to use “merge fields” to create lightly customized messages to our constituents, such as “Dear {First Name}.” Einstein GPT integrates your Salesforce with ChatGPT to take personalized messaging to the next level. To use it, you put merge fields into a generative AI “prompt” instead of into a cookie-cutter email template. An EinsteinGPT prompt might look something like this:

“We are running a year-end fundraiser with the goal of raising $100,000 for new programming at the community center next year. Write a letter to [First Name] to solicit a special donation for this project. The tone of the letter is casual and friendly, with a word-count under 200. For context: This contact works at [Primary Affiliation]. They have donated a total of [Total Number of Gifts] gifts in the past, totalling [Total Gifts], including [Total Gifts This Year] this year. The last Campaign they gave to was [Last Donation Campaign]. They live in [City], [State]. The email should be signed from [Relationship Manager]”

Try plugging this prompt into ChatGPT[chat.openai.com] with data from one of your constituents to see how it might work. The end result is that you can quickly generate highly personalized and varied messages to your constituents. Imagine if your relationship managers could double their personal donor outreach, in half the time!

These examples only scratch the surface. If you pop into your Salesforce Setup page and search for “Einstein,” you’ll see at least 10 different AI tools to explore. There are even more that can be enabled for you by your Salesforce AE, usually for an additional license cost. Chatbots, Conversation Insights, and Activity Capture are a few others that Nonprofits might want to consider. To learn more about all of the possibilities, this Trailhead Trail is a great place to start.

To help nonprofits start leveraging AI, Salesforce is offering a FREE Einstein for Nonprofits app that can be installed in your org. This package is tailored to help your org get started with the Einstein Prediction Builder tool. In general, this tool allows you to build predictive models around any question you might have. For example: 

• Is this contact ripe to become a first-time donor? 
• Might this donor become a TOP donor? 
• How likely are they to become a RECURRING donor? 

If you noticed that these example questions are the exact ones you wish you could answer, then you are in luck. The Einstein for Nonprofits package includes these three pre-built prediction models, plus 3 pre-set data models so that you can start seeing predictions more quickly. These models are configured to include a set of useful NPSP fields, and they can be easily cloned and customized to factor in additional data points that are unique to your org. 

Conceptually, these models are similar to other “scoring” formulas and algorithms you may have used. Leads or Donors with “High Scores” might merit more investment and personal attention. Many orgs also utilize “human intelligence” to rate donors as Hot or Cold, based on feelings and personals insights. These can be meaningful, but it isn’t scalable, and many potential donors will fall through the cracks.

What’s different with AI Predictions is that the formula is not static, and you don’t have to write it. Human intelligence can be factored in, but the AI can work through millions of records, outstripping the efforts of an army of human development directors. 

The AI model continually learns from your data and the outcomes you are interested in. It will become more accurate over time as it learns from what really happens in your org. In fact, the documentation recommends waiting ONE MONTH after turning on the Prediction Builder before you really start to use it. During that first month and beyond, the AI model is learning from your data and refining its predictions to be as useful as possible! 

In the near future, you could be making tailored outreach to donors who are rated most likely to become first time donors, top donors, or recurring donors, based on your org’s unique constituent data.