Serving Clients Full Circle

podcast

Podcasts

Listen to the weekly podcast “Around with Randall” as he discusses, in just a few minutes, a topic surrounding non-profit philanthropy. Included each week are tactical suggestions listeners can use to immediately make their non-profit, and their job activities, more effective.

Find “Around with Randall” on Apple, Spotify, or wherever you listen to your podcasts.

Email Randall with a show topic: podcast@hallettphilanthropy.com

Email Randall with a thought regarding a specific show: reeks@hallettphilanthropy.com

Listen on Apple Podcasts
 
 
 

Episode 5: Artificial Intelligence - Analytics of Tomorrow


Thanks again for joining me this week. I want to spend just a few minutes talking about where I think the future is going to lie when it comes to prospect management.

And in particular, the idea of screening.  since the invention of the internet organizations and nonprofits have become incredibly dependent, and in some cases I might argue almost addicted, to the idea of wealth screening.  In every year in my 25+ - year career, I see more and more conversations, particularly in those that are maybe a little less experienced, to rely on those wealth screenings.

Scores of information are used to become determining factors in how or why a gift officer with an organization sets its priorities regarding prospects…which leads to the idea of prospect management. Over the last couple of years, probably for the last decade in the for-profit world, we've seen the idea of looking at prospects the same thing in the for-profit world, and potential sales opportunities differently.

And that's now beginning to finally matriculate into the nonprofit world. I like to think about it this way. It's not that we shouldn't wealth screen. We should… it's a good data point to better understand exactly what's going on. But something that we've lost in this entire equation, what we know traditionally, particularly those of us with enough gray hair….and before we actually had the kind of internet we do today, in terms of research, we relied in at least in the early parts of my career on 80 percent of the equation being about inclination.  Really the word that I want to use is likelihood…which people are most likely to be connected to our organizations. And the relationship piece is good equation for each one of those individuals to find out if that was actually true.  

And what I see today is gift officers many times saying, well, these people don't screen high enough, and thus, I really shouldn't bother to build a relationship or make that phone call or cultivate them. I was at a a major training at a major academic medical center, last year,  and I was doing the training for the entire advancement division for the entire university. And I asked the question, “which was more important inclination or capacity?” and most of the room said capacity.

And I said, “well, give me two minutes and we'll ask the question again.” And I did what I'm going to do here in a second. I had one or two people that said, “yeah, it's still capacity. They're not rich enough. I don't care.”  

 So where's all this headed. The last couple of years, we have seen more and more conversation about looking at using some type of I'll call it in a broad term, artificial intelligence.  It's really some kind of analytics to better understand who are the most likely, the most connected prospects instead of just using wealth. Now, the inputs for this can be various. They can be from a healthcare perspective, HIPAA compliant data…from an education perspective, it can be a FERPA. their privacy in the education world, HIPAA is the privacy for healthcare, but for compliant data about when they were student, it can be the related donor information.


It's just not about what's publicly out there in terms of wealth. And that data goes into a central place. And math. And the idea of artificial intelligence is defined at least in my world as continually learning that constantly you're putting more and more data in and the algorithm, the math is  constantly churning all of that data that you're putting in there, thousands and thousands and thousands and tens of thousands, of pieces of information.


And it's learning about what is the most likely combination. Weighting of those particular data points to get output, which is a listing of people who are most likely to give. So there's a ton of data coming in - more than the human brain can figure out. And the computer learning the analytics figures it out and then spits out names.


And instead of reading it by wealth, it rates it by likelihood and it can be done a lot of different ways. So likelihood could be like a correlation, one to zero with one being the most likely zero being none. It could be a hundred percent. Is there likely, and zero could be none, but at the end of the day, it's pushing out a listing, and then allowing you to look at who's the most likely.


Here is where I'm headed. I think still to my initial training in this industry and the 25 years since hasn't changed, that that likelihood is 80 percent of the equation on identifying the best major gift candidates. And frankly, it's probably 80 percent of identifying the people you should be building an annual giving program for as well.


And then 20 percent is wealth. So, what does that look like? Well, on a podcast, it's kind of hard. I do a presentation and talk about this in a lot more detail kind of show, some thought process of how this works. But if, if you have 12 people that you're looking at and you wealth screen them and you put them in order from “A” to “L” and you rate them “A” as the most wealthy by rating, “L” is the least. And then you do the same thing with those same 12 people, and you put a likelihood score on them. And what happens if your top two or three people's likelihood is five percent, but the next four are 80% and they all wealth screen above a certain dollar figure, your major gift level, like $10,000 or more in terms of opportunity.


I would argue that you should be spending a majority, if not close to all of your time on making sure you get to the ones that were the 80 percent or more, because they're going to be the ones that come to fruition. The concept of wealth might tell us how much people will give, and the likelihood tells us who will give in five years.


My prediction is that this kind of thought process will be the standard in the industry, education, healthcare, arts, and science culture, whatever that they're going to be, that we're going to learn from the for-profit world that has figured this out over a decade. And that this becomes the norm. And I think about a great company like donor search, who I have a great deal of respect for the owners, the Tedesco family. They’re moving into this. You know, they, they have this repository of amazing data. They're really beginning to push the envelope. Like, Hey, wait a minute. We need to look at this differently. 


So I mentioned the major gift, you know, kind of this rating of “A” through “L “on wealth. And then if you were to do it by likelihood and, you know, the top two or three or five percent, you want to kind of begin to move down to the ones that are more likely, as long as you're above a major gift threshold. This also applies the annual giving process. I don't want to forget that because we see mail returns of one and  two percent. 


And yet if we were to kind of do this same analysis, so let's use that the same analogy or example that “A” through “L”…well, if the bottom three people, prospects “J “, “K”, and “L” were gift thresholds of $5,000, $1,000, and $500, just for the sake of argument, they’re not in a major gift level, but their likelihoods are all above 70 percent. Well, that would tell you who you should be reaching out to from an annual giving perspective, whether that’s if you have enough staff and intermediate giving level for that $5,000 making a phone call.


Or a mailing program for those that are below, let's say $1,000. If you're an annual giving person, think about the dynamics of changing your return rate is if you could use likelihood and rate people and then not mail to those that are below a certain threshold. So you remove the major gift level first that are above a certainly likelihood and above a certain wealth screening, 80 percent likely and 20 percent wealth. And then you're left with everyone else. And anybody let's say that's below $10,000, but has a likelihood of above 70 percent, you mail to those and you don't even mail to the other.


We're going to have to learn to trust something that we probably don't understand entirely. And that's the math. The ingenuity comes from you to utilizing analytics or artificial intelligence it's foreign to us. We’re kind of afraid of the new, because it can redefine how you look at prospect management.


So what's the tactical piece, cause I never want to forget that the first thing is inform yourself. You probably need to know more about analytics and artificial intelligence, do some reading, and number two, be open to it. Be open to conversations internally about is there a better way for us to identify prospects?


Thirdly, listen to some of these experts. I think about the donor search team. When they do these webinars, get some outside perspective. If you do those three things, now you will be prepared when this becomes a reality, probably in 2021 for a lot of people by 2023 and 2024, it will be predominant. And by 2025 and 2026, if you're not doing it, you're going to be way, way, way behind the curve.


This is new. It's a challenge, but I'm telling everyone who's willing to listen. This is the way the future looks. The for-profit world has figured this out. The big companies do this to figure out who their best customers are and we should follow it because it's going to give us a better chance to serve our non-profits.