Nonprofit Expert Episode 19 – Leveraging AI for Nonprofit Success

​​Darryl Moser Host00:14

Welcome to Nonprofit Expert presented by DonorPerfect.

Welcome to the Nonprofit Expert podcast presented by DonorPerfect. My name is Read More

​​Darryl Moser Host00:14

Welcome to Nonprofit Expert presented by DonorPerfect.

Welcome to the Nonprofit Expert podcast presented by DonorPerfect. My name is Darryl Moser, I am a DonorPerfect product manager and my passion is all things tech related to helping nonprofits succeed. Today, we will talk about AI, artificial intelligence and how it’s changing fundraising asks, and my passion is all things tech related to helping nonprofits succeed. Today, we will talk about AI, artificial intelligence and how it’s changing fundraising asks and cultivating a more meaningful relationship with donors, going a step beyond the chat GPT tools into data models that are fundraising specific. Today we have with me Sarah Teesco, executive Vice President and co-owner of DonorSearch. Sarah, welcome and tell us a little bit about DonorSearch.

Sarah TeDescoGuest00:47

Yeah, thanks so much, Daryl, and happy to be here today. So DonorSearch is an amazing company, if I do say so myself, and we work with thousands of nonprofits and have been partnering with them by providing them with best fundraising solutions and research solutions for over 16 years. And we work with nonprofits all over the country and are very proud partners of DonorPerfect, allowing us to give data insights and really great intel on your donors. So that way it’s right inside of DonorPerfect.

Darryl MoserHost 01:25

Great, great. Yes, our history together has been pretty incredible. We’ve seen your product help our clients be able to raise some major dollars for capital campaigns and major gift fundraising efforts. But something new that, of course, has come about and everybody’s hearing of it maybe even touching on fatigue a little bit is AI and artificial intelligence. And we’ve certainly heard about ChatGPT and the way that it’s impacting how people are authoring their letters and their emails and so forth, and that’s been making an impact, but in your situation it’s a little bit different than just text, right? So how did DonorSearch? How did it become clear that AI had a role in DonorSearch as business?

Sarah TeDescoGuest02:09

And it’s interesting, chatgpt has definitely helped. I think a lot of people cross, you know, this chasm, you know this technology chasm. It’s allowed for a bit more accessibility and people have become excited by it because, for you know, it’s just making it so much more tangible. So people are so much more aware of of the AI capabilities. But with DonorSearch, yeah, we are focused on predictive analytics and helping fundraisers and nonprofits really connect with the donors who are ready to make a gift. You know, this year so at DonorSearch, it became clear that AI had a role in, you know, becoming part of our solutions and part of, you know, the efforts for our nonprofits.

03:01

When 2020 started, you know, the pandemic had had some really detrimental effects on so many organizations and we saw so many nonprofits struggling with either maintaining their fundraising goals or even surpassing their goals.

03:21

We had made connections with Nathan Chappelle, our senior vice president of DSAI, and also Scott Rosenkranz, who is our AVP of DSAI, and we had instant synergy and they were already exploring, you know, these AI solutions and how they could help nonprofits.

03:43

So Bill Tedesco, our CEO, and I had met with Nathan several times and eventually we were able to make these great connections and onboard the DSAI team to DonorSearch. After a few years of courting, you know, they became the central department to DonorSearch and also, in addition to that, you know, at the same time, what we were seeing is on the charitable giving side was this decline in donors. So we saw not only a decline in donors but we saw many nonprofits unable to achieve or reach their goals, many nonprofits unable to achieve or reach their goals, and actually, Nathan and Brian Crimmins actually published a book called the Generosity Crisis, which really highlighted those challenges. So you know the timing worked out really well with bringing these AI solutions that were going to help nonprofits be more effective and more efficient at reaching their goals while at the same time, you know, trying to overcome these challenges that we were seeing with COVID and the pandemic.

Darryl MoserHost04:50

Makes complete sense, and something that you touched on, I mean DonorSearch, is business, probably since the beginning, has been about predictive analytics, and predictive analytics, for many of our customers, has been knowing who do I call to make that big donation that we really need right now to get us, you know, across the bridge to be able to make whatever kind of capital campaign we’re going to be doing.

05:12

And in the old days this would have been just sort of a raw data, somebody, seat of the pants maybe would have been estimating on who they thought was going to be giving the gift, based on some history. But now you’ve got this AI engine that you can actually start doing some number crunching. So, with AI and the other piece that comes along with it is machine learning and ML. From what I recall in my programming days, the idea here is that you had some serious computing power that was crunching large databases that are out there. Can you tell us a little bit about like what, what kind of power is necessary to try and achieve the results that DonorSearch is able to get?

Sarah TeDescoGuest05:56

Yeah, so since we brought in, you know, the machine learning technology and we brought in the DSAI team, we have been able to grow and enhance our technology stack. You know previously, when we were just doing screenings, or you know these sort of traditional methods of looking to see whether or not somebody was likely to make a gift, or you know what their wealth capacity is. We might screen, you know, or what their wealth capacity is, we might screen a million to five million records a month. Now, what we’re seeing with the AI solutions and working with our clients is that we may process a million records for one client in a month, and so what we need is we need this ability, and what we have now is we have this ability for you know an individual donor to have, you know possibly a thousand features that we’ve been able to review.

06:56

You think of the nonprofits big data set. You have to have this technology that’s able to do that, and not only do that, but need to do it well, and you need to be able to do it frequently, because that’s really how the models are going to continue to be trained. Right Is that you constantly are adding this new information, that you’re receiving these updates that we’re understanding the engagement of the donors and understand how, you know, that relationship might change over time. So it’s been pretty exciting and it’s, you know, it’s really cool to see those changes and of course, it’s also, you know, really amazing to see the results for those clients and how excited and happy they are with the accuracy of the data.

Darryl MoserHost07:43

For sure, for sure, and I think models is a key term that’s used in AI the data model. Some people call it a universe of information that you’ve got and how it ends up comparing an individual donor to what that universe looks like. I wanted to push into a little bit of the new DonorSearch AI Insight Scores, which I think is one of the key products that we see coming out of a lot of this number crunching Newly available, something that you can add to DonorPerfect today when a client gets a screening. But tell us a little bit about what these Insight scores can tell a nonprofit.

Sarah TeDescoGuest08:25

Yeah, and the Insight scores that are going to be delivered with DonorPerfect. We’re thrilled about that. It’s going to be really amazing and powerful for the clients. So DonorSearch Insights is actually you have a series of models that you or that a client is able to choose from, and what we’ve seen is it’s a great way for people to get started when it comes to AI. So DonorSearch insights allows for organizations to get a handle on some of those foundational strategies that they might have in place. For instance, we have what’s called a sustainer model, and the sustainer model predicts which donors are most likely to make at least four gifts in the next 12 months. Obviously, the benefit of this is that the score can help nonprofits build or grow a monthly giving program. It allows for sustainers or monthly donors typically to have. It allows for higher retention rates, and then these are also less costly to get started. The kinds of data that you would need for a sustainer program would be you know your constituent file. It would be the detailed giving transaction history from three to five years Detailed giving transaction history from three to five years and then optional data that might be able to enhance that for you would be detailed event and email history. So click through rates. You know that kind of thing as to looking to see what sort of engagement you’re having with your donors.

10:04

Some of the other insight models that we have is an acquisition score, so this would look to predict which non donors are most likely to make a first gift. You would use your client supply data more than 200 different data points from the DonorSearch and wealth philanthropic screening. It’s a little bit different, right? So this is a different type of model because these are non-donors. So you would look at the information, like your constituent file information and any sort of detailed event or email history, so you’re not going to have that sort of giving transaction. You also have a retention model, so this is going to predict which first-time donors are most likely to make a second gift. Using the client supply data again, and more than the 200 data points, we’re able to see which donors are likely to give a second gift.

10:59

When you think about, you know, retention as a whole in the nonprofit industry, you see so many organizations that are challenged with this where they have retention rates of 40% or less. So bringing in this retention model, you can really elevate the program. We also have an upgrade model which is going to predict which donors are most likely to make a gift that is at least 10% higher than the largest gift in the last three years. Again, the same sort of data information that you would want to pull. So your constituent file, detailed giving transactions, optional data would be that detailed event and email history.

11:38

A lifetime value score is also available with Insight. So this is predicting which donors are most likely to make a gift of at least $1,000 to your organization throughout their lifetime. So you would use similar information and you know the idea behind this, or the benefit, is that, on average, approximately 10% of donors will give $1,000 or more in their lifetime. So 90% of donors will give less than $1,000. So this score, this insight score, is going to allow you to see, you know, precisely where you should be spending your time, and the data required for that would be your constituent file information and then your detailed giving transaction. So lots of different ways that you can use it and for a bunch of different programs that you might have in, you know, your organization or development shop.

Darryl MoserHost12:27

So those scores are amazing, amazing, I mean. You’re you’re touching on all the different donor journeys, um that that can end up happening. The life cycle of the donor, everything from, you know, a donor acquisition all the way to their monthly giving, um, getting into some major gift, uh statistics as far as when they’re going to be giving more than a thousand dollars. It’s almost. If you had that information, you could manage every one of your individuals and just walk them through their life cycle, which is really what a fundraiser needs to do, I guess. So tell me about the output. In other words, if I, as a nonprofit, ask for one of these insight scores, what do I get back?

Sarah TeDescoGuest13:05

Yeah, and I just want to highlight too you’re exactly right, daryl that it’s all about the donor journey and making sure that you know the engagement that you are receiving, you know is going to have those positive results, that you’re connecting with the people that actually care and if there are opportunities to improve that donor, you know journey that cycle, then we’re able to identify that Right and you’re spending time with the people that will make a gift and those that are maybe not ready to make the gift right now or be a part of that program steward. And the question so your question about you know what sort of scores do you get back? So you know the way that we have, you know the way that we’ve delivered this for our clients is similar to any sort of model. So there’s a lot of you know, quote, unquote magic that happens behind the scenes. Right, we’re using all of this amazing technology, we’re bringing in all of these different data points and then we’re accurately, you know, predicting who you should be spending your time with.

14:11

The way that the scores are broken down is that you receive, you know, for instance, let’s say, an acquisition score. So this would be for your non donors. So you would receive a score between zero to 100. And those scores are relevant to your organization. It’s a national model but it’s relevant to your organization. That acquisition rank for the top third of the file are based on. You know your top priority. Then you would have you know those folks that would be your middle tier, that would be folks that you would want to, you know, begin to cultivate. And then you have your lower scores where maybe you need to spend a bit more time. So there’s no good or bad score. You basically use the results based on your specific use cases and objectives and again, it’s just all relative to you know your programs that you have in place and how you want to interact with those donors and we were to look at their acquisition score, it’s almost inconceivable to think that a development officer would be able to reach out by telephone to 10,000 people.

Darryl MoserHost15:32

I’ve not done the math, but I have a feeling it’s going to take a really long time. So the other temptation is well, why don’t we just blast all 10,000 with an email and say hey, you know, you must be a really likely donor. I don’t think that you’ve given yet. So how about making a donation?

15:47

But by using this score you can actually do what we often call multi-channel communication, where you can still use email and you can broadly communicate to individuals. But if you’ve got people that are showing up on the one end of this score, that are very likely to give a gift for the first time, their communication can actually be tailored to either include direct mail, which has cost to the nonprofit, or maybe even time, which is going to be the phone call to the donor, which both of them are very valuable commodities that the development director has to manage. And so getting this score really enables the nonprofit to budget where they’re going to be putting their resources to have the maximum effect. And that’s true for all the scores, right, it’s basically a 1 to 100 score and you can either divide that 50-50 or into thirds and you can segment your communication accordingly, based on whether you’re looking at retention, acquisition, sustainer, any of those models right.

Sarah TeDescoGuest16:51

Yeah, absolutely, and it’s a, you know, opportune time. Or, you know, a great example to bring up is the Smith Center, who we work with out in Las Vegas, and they felt many of the struggles during COVID. They had layoffs, they had, you know, I think they laid off 75 percent of their staff, they anticipated a loss of 40 million dollars, they had to cancel 450 performances, and so they were really struggling and you know they were doing their regular screenings. They were doing their fundraising efforts that you know we were all you know used to doing. But after meeting with them actually in person, we had some really great conversations, and Bill actually had a great conversation with them about why not trying some of these insight scores and bringing the AI into the mix.

17:42

So they processed actually 50,000 records and started a sustainer program, and so they were looking to see, you know, who’s most likely to give at least four gifts in the next 12 months, and so they received their scores back right, that zero to 100. And what they were able to do is they actually, you know they were a little skeptical at first too and Carol talks about this, there’s a great case study about it but you know, they just wanted to see, you know, is it going to work or is it not going to work? And so what they did is they emailed all of these folks and what they were able to see is, or find, was that they were able to double their monthly giving program. They were able to, you know, generate these great returns, and you know the benefits, obviously, of a sustainer program is that it’s kind of it’s recurring revenue, right Like you don’t have to keep going back and asking for that gift.

18:37

So they were over, they were able to overcome a lot of those financial challenges. They were actually able to start growing again, and that’s all because of, you know, trying something a little bit different bringing in some of that engagement data and segmenting, just like you’re talking about, on the folks, and prioritizing people that are in that top tier, that are going to be able to, you know, make that transformative gift or become a new type of donor for them. Make that transformative gift or become a new type of donor for them. So it’s pretty, it’s pretty exciting and it’s, of course, cost-effective, just like you were talking about.

Darryl MoserHost19:12

Yeah, yeah, amazing. Doubling your monthly gifts, that’s a. That’s a. That’s a great testament to to the kind of performance that you can expect. This with this tool, amazing, amazing. I guess you know we’ve known DonorSearch for a very long time as being a major gift fundraising tool. And amazing. I guess you know we’ve known DonorSearch for a very long time as being a major gift fundraising tool, and what we’ve talked about so far is something that I think of as sort of a transactional fundraising. Maybe the monthly gifts and the acquisition models are a little bit different than the way we look at major gift fundraising, where we’re also looking at capacity, which I think is key. So tell us how AI has impacted DonorSearches, major gift fundraising. Are there any results on predictability of gifts and how does that work?

Sarah TeDescoGuest20:09

Yeah, so it’s interesting. So the insight scores are definitely great for different programs that you have in place, right. So not necessarily focused on major gifts. Of course, you can do upgrade levels or you could do lifetime value, which have some of those transformative gifts inside of them about, you know, really enhancing a major gift program. Our custom DSAI solutions are what we would call the enterprise level, are also a great way to really give more power and momentum to what you’re currently doing.

20:39

So the idea with that is that you’re able to. You know it’s a dynamic model, right, and we’re looking at seeing you know who’s likely to make a gift this year, in the next 12 months, and then we’re also looking to see who will become a donor philanthropic capacity. In addition to you know who’s likely to make this gift right now. So in the past, you know it’s always been and it’s still powerful, right. Screening is still a wonderful way to do your research, but if you’re ready to do more, bringing in the AI and the machine learning is just going to give it all the momentum that you need.

21:24

But what this allows you to do is you’re taking out and you’re removing any sort of biasness, so you’re taking out any sort of things that might have biased your model to look at a donor, that capacity, that philanthropic information.

21:39

With the engagement data, use all of your different fields and criteria to figure out who you should be talking to and like, really segment it down, and it allows you to make those connections that maybe you wouldn’t have been able to make in the past. We also see that the DSAI component is really effective in helping manage the major gift portfolios, so you’re able to refresh those major gift portfolios much more quickly than you have in the past. We’ve also seen that you actually can increase the average gift size to find better donors In some cases it’s five to 10 times that size and then, of course, obviously, finding new donors that can fit that. So it’s really you know the the DSAI solutions and, of course, the insights to they are flexible and how you can focus on those people that are going to make those transformative gifts. It’s just a matter of like where your efforts are and what questions you’re you’re actually asking. Is that helpful?

Darryl MoserHost22:55

It is. It is very helpful and I think you know maybe just to reiterate a little bit so that that our nonprofits have an understanding of this is the the insight scores to me. They’re great for clients of any size, clients that are doing routine fundraising. When we start talking about major gift fundraising, oftentimes we’re talking about capital campaigns and large, significant gifts that are going to be coming through those channels, and so it’s just a bit of a unique way of doing fundraising, but two different tools that DonorSearch has that can apply to either end of the fundraising spectrum, which is great. Now, something that you had mentioned in that conversation about major gift fundraising is bias, and you were telling me a little bit before we went on the air about responsible AI for fundraising and tell us about what that is. What’s the concern?

Sarah TeDescoGuest23:46

Yeah, great question. So, first of all, there is a community it’s called the Fundraising AI Community, and they just had a summit where the conversation was all about, you know, responsible AI and being transparent. You know, when people think about AI, there’s a lot of hesitation, right, you know? Are we using this for good? Are we using this for evil? There’s a lot of hesitation, right, you know. Are we using this for good? Are we using this for evil? Which is a terrible way of saying it, but it’s true, right, you know there’s a lot of like. There’s a lot of questions associated with it and with responsible AI. You know, the goal there is that we want to make sure and ensure that we are delivering these solutions that are transparent, that are unbiased, that they’ll be used for good, and this fundraising AI community, you know, has actually created this framework that we are adhering to and we’re going to make sure that we adhere to, and, in fact, nathan Chappelle, our senior vice president of DSAI, he was the one who actually chaired it.

24:51

You know, another point, too, that I just want to throw in there is, you know, when we talk about this evolution of AI in 2019, you know, there was a study that was done. It was the state of artificial intelligence in the nonprofit sector, and that was done in 2019. In the nonprofit sector, and that was done in 2019. And it asked a series of questions and what they found was that 52% of people are afraid of AI. 75% of people believe that AI will make their life easier. 73% of nonprofits believe AI innovation aligns with their beliefs, but 83% believe that there needs to be an ethical framework surrounding it.

25:32

And when you, you know, fast forward to 2023, and we look at, you know, chat, gbt, and we look at, you know, some of these rollouts of AI and all of the different conversations surrounding it, I think those questions and thoughts, they are so much more relevant now and they’re very present. So people are starting to use it. But you know, now it’s. You know how do we harness this technology? How do we make sure that we are using it responsibly?

Darryl MoserHost26:02

Yeah, I think, maybe to touch on some examples for for our listeners listeners just a bit. I mean the sensitive areas here are when we start looking at geolocation, ethnicity, gender and so forth, and those are some of the attributes that I guess, before you let AI begin to process on the data, that they need to be removed from that so that it doesn’t end up finding patterns that can be false positives or assumptions that are just incorrectly based on some of this data. And I think from what you had shared is that after the calculation is done and the ranking happens, then that information can actually be overlaid back on top of it, along with some of their wealth, capacity and so forth, so that it doesn’t create these biases that are out there. So I mean, it sounds like it was handled very smartly and also probably just in time, as we ended up having these engines, you know, really push into some extreme data elements of information out there about our donors.

Sarah TeDescoGuest27:05

Absolutely, and you hit the nail on the head there, daryl again, is you know what we do at DonorSearch. We remove those you know specific or those demographic data points that would point us in the direction of bias. So that way, you know, you are focused on people that are engaged, that are focused on you know making that connection and are ready to make those gifts now, versus just you know what have we done in the past. So it’s an ability to actually uncover even more potential and to maybe even focus your efforts on some initiatives that allow you to engage new donors and a group of diverse donors too.

Darryl MoserHost27:52

Great, great. No, it’s excellent to hear that DonorSearch is not only participating but also chairing some of this responsible AI, and it’s going to be very important as we progress forward in adding more to these models and doing more with it. Well, this has been tremendously informative. Thank you so much, Sarah, for joining us today. We hope that our listeners have truly enjoyed hearing about how AI is impacting not just how they’re writing the emails, but also just how they’re approaching their donors for fundraising asks and just being smart and efficient fundraisers. So thank you so much for sharing about donororSearch today.

Sarah TeDescoGuest28:31

Absolutely. Thanks for having me, daryl, and look forward to talking more in the future. Sounds great, take care.

DonorPerfectAd28:38

Thank you for listening to Nonprofit Expert presented by DonorPerfect. For more information and a special offer, visit DonorPerfectcom/podcast.