Unpacking the AI-Generated Donor Score and Ask Amount

CallTime.AI leverages artificial intelligence (AI) to help our users identify the right people to solicit, and then to make the right solicitation. In this article, we'll dig a bit deeper into the utility- and, naturally, the limitations- of the AI-generated Donor Scores and Ask Amounts.

What AI Is and Is Not

AI is a type of machine learning that takes into account as much relevant data as possible to discern patterns that can be used to predict certain outcomes. We program it by determining what data it should look at, and then telling it what outcome we are trying to predict.

Because computers can instantaneously look at and analyze hundreds of millions of data points all at once, its analysis can provide valuable assistance. But AI shouldn't be used as a replacement or substitute for human insight, expertise, or judgement. Because just as the AI will be taking into account and computing vastly more voluminous information than you will have access to, you will inevitably have context and information the AI lacks.

What kinds of things influence the Donor Score and Ask Amount?

The overwhelming factor that influences the Donor Score is that contact's giving history, including, among other things:

• The recency of those donations;
• The total amount of giving;
• The size of those donations;
• The partisanship of those donations;
• Whether multiple gifts tend to come from the same household;
• At what point in the cycle donations are made.

Beyond giving history, the AI also takes into account, among other things, the donor's:

• Demographic information
• Occupation and employer
• Geography

Finally, our data scientists are always testing new sources of data to determine the effectiveness of it to help make good predictions about donor behavior, including: social media data; philanthropy; issues trending in the news; and other proxies for wealth (i.e., stock/home/business ownership).

How is the Donor Score and Ask Amount tailored to my campaign?

The above information is focused on data that looks squarely at the donor. But the AI also takes into account data related to your campaign, and adjusts the Donor Scores and Ask Amounts, accordingly. Some of those factors include:

• The office pursued
• Contribution limits
• Candidate geography
• Issues and ideology associated with the candidate
• Type of committee (i.e., candidate, PAC, party, etc.)

How the AI Improves Over Time

As discussed above, we program the AI by telling it what data to look at, and what outcomes we're trying to predict. In our case, we're trying to predict what someone will donate when solicited (the Ask Amount) and the likelihood that they will max-out to your campaign over the course of the cycle (the Donor Score). AI can learn over time by looking at more relevant data, and also by assessing how well it performed in the past, and use that to improve future performance.

So in addition to constantly growing our dataset with new and updated sources, the AI is constantly learning from your performance as you use the platform, and uses that learnings to provide you with even more accurate feedback.