Community IT Innovators Nonprofit Technology Topics

AI and Nonprofit Project Management with Alex Tuck

• Community IT Innovators • Season 7 • Episode 27

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 33:07

Carolyn Woodard explores the intersection of project management and artificial intelligence AI with Alex Tuck, founder of Tuck Consulting Group. As nonprofits face increasing pressure to do more with less, Alex shares how project managers are uniquely positioned to lead AI implementation and why human-in-the-loop systems are essential for maintaining accuracy and ethics.

The conversation dives into the practicalities of AI adoption, from the potential automation of administrative tasks to the specific challenges faced by healthcare nonprofits. Alex provides a clear framework for ethical AI use, focusing on well-being, transparency, and data security. You will also learn actionable tips for mastering the art of prompting and how to know when to iterate on an AI project versus when to walk away.

Alex and Carolyn discuss:

  • Why project managers are the natural leaders for AI initiatives in the nonprofit sector.
  • Four key pillars for an ethical AI framework: Human Well-being, Human in the Loop, Transparency, and Data Security.
  • Practical strategies for better prompting, including the use of context, constraints, and prompt libraries.
  • How to determine when to pivot or quit an AI project early to save organizational resources.

Resources Mentioned:

_______________________________
Start a conversation :)

Thanks for listening. 


Carolyn Woodard

I'm just gonna turn recording on because we say something like cool and then I'll forget and they'll be like, wait, can you say that exactly the same?

Carolyn Woodard

Welcome everyone to the community IT Innovators Technology Topics podcast. I'm Carolyn Woodard, your host, and today I'm really excited to welcome a friend to the podcast. So, Alex, would you like to introduce yourself?

Alex Tuck

Yeah, sure. Thanks for having me, Carolyn. Um, my name is Alex Tuck, and I am the founder and managing principal of Tuck Consulting Group, the most creative name on the planet. Um, we are a team of about 60 uh project managers, all US-based, um, mostly working with uh clients in North America, but we do work with folks all around the world. Um,

Alex Tuck

We specialize in project management in four main verticals, um, though sometimes we do stray outside of it. Um, healthcare, nonprofit, uh, professional services, and digital SaaS. Um, and that's really where we have like a deep uh bench of experts. Um, but when it comes to project management, it can be applied across all industries. And so um sometimes we do uh stray outside of that.

Alex Tuck

So really excited to be here with you, Carolyn, to chat a little bit.

Carolyn Woodard

Thank you for joining me. And I know uh we're gonna talk a little bit about AI, which is is there any other podcast topic right now? But uh but it is just something that um I was interested to have you on because I think you have some different things to say about AI, particularly with nonprofits. Um so

Carolyn Woodard

You gave up presentation recently, and I just wanted to kind of walk through that and ask you some questions about it and you know, get your perspective of what you're seeing from your nonprofit clients and kind of what's going on in the nonprofit and AI ecosystem right now.

Alex Tuck

Yeah, it's really interesting. Um, you know, because we we're project management shop, right? And so what the heck are we doing talking about AI, even right? And um, first of all, um, you can't go anywhere without talking about AI, right? But um, you know,

Alex Tuck

What we've found is that um project managers are sort of in this sweet spot where we can uh we're seeing, we're see, we're like leading AI projects like everywhere, not just in nonprofit, but everywhere. Um and we're seeing uh how folks are doing it right and and are doing it wrong.

Alex Tuck

And I swear to you, if anyone tells you that they are a straight up AI expert, they are lying because like every day it changes. And like now we're seeing that like AI might be like intentionally lying to us, and like there's all this stuff that's happening. Um and so uh

Alex Tuck

It's really about like how to try have the trial and error and fail as quickly as possible. Um, and so like just being in the PI PMI or project management space, we're seeing like where PMI is rolling out like the CPM AI framework of it, which is like very heavy, right? But still takes you through like the the AI framework for projects. And and then we're seeing these really lightweight implementations. So

Alex Tuck

It's really fun to just see how all of that's happening, and then especially for nonprofits, I I run a nonprofit too. I found it a nonprofit and run it. Like I know that um, you know, we're all so cash strapped and resource strapped that we have to figure out how to be lean and adopt AI and still do our like actual program work as well, right? So um, so that's why we came up with this content um that we did, this presentation that we did uh for some nonprofits.

Carolyn Woodard

So well, and I want to push back a little bit because it's interesting for you to say, oh, we do project management. Like, what would we know about AI?

Carolyn Woodard

Because that seems like of many things that people have started trying to use AI for, I think project management is one of the ones that's like the clear winner. There's clear tasks for AI to help you with in project management. So of course it would be impacting like how you're doing what you're doing, how your clients are doing what they're doing. So, you know, just interested to hear your insights into it.

Alex Tuck

Yeah, it's so interesting. Like, um, I don't know if you saw the Anthropic report about like the the impact on on employment. Um, but I mean, if you look at where I where I think about project management, I think it falls into two categories. So

Alex Tuck

For listeners out there, there's like a really awesome graph. Figure two is really, really interesting because it shows this interesting graph about how um there is the potential for overtaking jobs in a particular area, and then there's actual like overtaking of jobs in that particular area. And um, they overlay those two on the graph. And

Alex Tuck

There's two areas. Um, one is like office workslash administration, and one is management. And there are sort of like at different points on this graph. I don't even know what the graph type of graph is called, but um but uh when you look at it, like something like 90 or 95 percent of all office slash admin and management um has the potential to go away with AI theoretically. It's I they call it the theoretical potential. Um, and then

Alex Tuck

But then the actual penetration in those areas, like office and admin, is like 20, 10 so far. And then management's like 5%. You know, everyone knows why management has not been, you know, taken out yet. But um it you're exactly right, Carolyn. Right? Like I

Alex Tuck

In project management, we we estimate at our company, we estimate that approximately 80, 85% of all project management tasks that we did two years ago are gonna be completely replaced with uh AI. And then, you know, we look at ourselves as project managers. We see, I don't think the term project manager is ever gonna there is gonna exist exactly like that in the future, but there are still people that need to facilitate conversations and get work done. And um,

Alex Tuck

When you have cross-departmental um activities, you can't have a functional manager manage a project because like the CFO is not necessarily gonna agree with the CIO and or the executive director. And so you sort of need that person that's gonna play um the facilitator across those groups. And uh we think that we're gonna be there for a while.

Alex Tuck

And getting AI to give you an output that you expect every single time is also a huge shot, like an impossible challenge.

Carolyn Woodard

Right. We've been, you know, we've just been seeing that um with our clients as well. Uh, in that, well, it's some some of the things that you talked about, like having that human in the loop. And I think the more we do it, the more we're realizing that is just a non-negotiable. Like you cannot just take the output and like share it widely without having looked at it.

Alex Tuck

So yeah, it's so interesting. And then just especially, you know, this is this is in the nonprofit space, but it's really like healthcare nonprofits, especially, are really battling this, or healthcare industry in general is really battling this. Like, where do you put human in the loop with all these things? And

Alex Tuck

Especially because of the laws that we have in America, like physicians, even though you can like a physician can be part of an entire AI build-out. Like

Alex Tuck

There's this really uh interesting build-out, big hospital system in the Northeast. Um, they they built an oncology workflow where um average oncologists typically spends between approximately 30 minutes to two hours prepping for a call, like reviewing charts and doing all that work, which is insane. Like, because I mean most physicians spend zero. Like they literally like look at the chart as they're walking in the door. Um, but oncologists have to spend all that pre-work time doing it. And so

Alex Tuck

One of these um hospital systems, they spent the past two years building out an AI workflow to read all that data, give it a summary, link back to all the biomarkers, and then also sit in there and do the ambient listening and then do the output. It's more accurate than the humans, like literally more accurate by like two percent. Um, so it's like a 99.6% accuracy rate, but they still can't push the green button to deploy it because you know they they need it to be like closer to 100%. They need it to be dang near perfect. And so, you know,

Alex Tuck

That that brings up the point like when does the human need to be in the loop? And um, so it's really interesting. It's just a very fascinating challenge that we have right now to figure out like how accurate do we need to be?

Carolyn Woodard

Yeah, yeah. And that falls under like ethical can questions and considerations that you know, I think you've seen, you said in your health health work um activities. And are there other things that are coming up a lot? Or like we've talked about accuracy, we've talked about having the human in the loop.

Carolyn Woodard

Are there some other, I don't know, tips or trends that you're seeing around ethics uh adopting AI?

Alex Tuck

Yeah, I mean, I think, you know, there's there's like there's a bunch of different ways that you can think about AI ethics. Um so just like you know, just from like an environmental impact and like all these other things, right? But like

Alex Tuck

There's four sort of areas where um we think if if you if you have these four things um in mind as you build out your ethical framework, you know, it it will help you in whatever you roll out from an AI perspective across your organization. Um

Alex Tuck

One is that uh it should always be focused on helping enhance like human well-being, right, in some way. And so um, you know, it's not just nonprofits that have to think about that. Everyone should be thinking about that. It shouldn't be about like, okay, how can I eliminate jobs and make more money? It's like, how do we make our workforce or um our clients um or patients um have uh have better service um or have better well-being. Uh so that's one. Um,

Alex Tuck

There should always be human in the loop, right? We already mentioned that. Um I think it's real this like

Alex Tuck

The third one is really important. Uh it's it's about being transparent about how we're using AI and when we're using AI. I think if we did that more, then it would help people understand whether they can trust it. Like we see all these memes on the internet, and we're like, are these real? Are these not real? Like, it's so easy to get disinformation. Um, and so when we are able, when it when we make it a standard to be transparent about how we're using AI uh and when we're using AI, I think that's really important.

Alex Tuck

We're trying to figure out like where to place that, like on our website so that people know that, okay, we might be using AI to do research on certain pieces of of content that we're creating. Um, but just the more transparent we can be, the better.

Alex Tuck

And then one that I think the US hasn't really figured out yet from a legal perspective, right? And so in the absence of that, like how do we um how do we as a as a company or a nonprofit or whatever, how do we make sure that we're really making sure that, okay, when we're putting stuff into Chat GPT, are we putting it into a free version of Chat GPT and like now we have donor data that's feeding the model? Or um, you know, uh do we have things, do we need to have things air gapped or or protected in some way um so that client data is not being put out onto the internet and PHI and all that stuff?

Carolyn Woodard

Yeah, no, I've we've had a lot of clients that have been asking us about that, that that kind of leakiness, I guess I would say, of being um feeling that when they're using AI internally as well, like which vendors are they using, which tools are they using, what are those terms and agreements and conditions? And uh

Carolyn Woodard

For a lot of nonprofits that maybe have been kind of coasting along with kind of okay cybersecurity, I we are getting a lot of inquiries and anxiety around do we have enough cybersecurity to protect our client data, for example. Um because we're being careful of it. But if we are, you know, if there we're working on some kind of advocacy area, yeah, and there may be external actors who would really like to get our list, you know, and then you have AI that's made that list really easy to find what you need to find in it.

Carolyn Woodard

It's kind of prompting a lot of nonprofits to think more about keeping the whole thing secure, which is good. Probably should have thought about that, you know, years ago or whatever, because it's not the fact that you know, people aren't really interested in those lists of your clients or the work that you're doing, your connections, your networking, all of those pieces, but it is prompting people to think a lot more about that security, I think. Are you seeing that?

Alex Tuck

Oh my goodness, like that is such a good, I mean, such a good point.

Alex Tuck

And you know, people are super skeptical. And I love like I am married to a Luddite. She like absolutely hates technology, but she like brings up really great points. Like, you know, my kids just like they're they're four kids, four to four to ten years old, and they love to ask Google questions, right? And she's like, have you guys thought about like asking yourself what you think the answer is first before you ask the question? And I was like, Isabel, that's like such a smart thing to do, right? And so um, you know,

Alex Tuck

Having people that are distrusting and asking these questions is actually really helpful because I mean you saw the striker striker striker went down because their Microsoft environment got hacked. And it's like, wow, these like humongous companies that have almost everything in place. They got Intune, everything like Intune was their vulnerability, and it's like, oh my God, like what is going on? And so um, yeah, it's just it's it is, it's really interesting to see what's happening there.

Alex Tuck

I think that there are some, like I think government it all come just comes down to governance, right? And making sure that you have a set of individuals who is making sure that everyone is educated and also understands how AI is being used and technology across the entire organization is being used, and how you're protecting your internal data and your client data and all that stuff across the organization and approving each of the tools that is acceptable and what's not. And um

Alex Tuck

I think when you make that intentional step, you'll you'll everyone's gonna use like everyone's gonna be using AI. Like there, it's crazy, like because especially in nonprofits, it's like we expect people to do 60 hours worth of work in 40 hours, and there's a math problem with that. And so um the only way you can do that is through AI. So we have to give you know our folks the tools, uh, whether it's AI tools or technology or you know, whatever, better processes um to be able to achieve their work quicker. Um, so it's yeah, yeah, million-dollar question, though, right?

Carolyn Woodard

Yeah, no, I've I've said that a couple of times is that I feel having worked at for-profits and nonprofits, and um, you know, all of the concerns that are coming up around workforce, as you were talking about with the report um from Anthropic, the jobs that are gonna go away. And

Carolyn Woodard

I think, you know, a lot of for-profits will look at AI and think, you know, I can fire five people, like one person can do five people's jobs. And from the nonprofit side, it's like we already have one person doing five people's jobs. Like now, maybe they could have an AI assistant that helps them do three people's jobs better so they can actually spend their time on the last two jobs that they really need to do for their nonprofit to move forward. So I don't see it as much as being uh, you know, now we can fire people in the nonprofit sphere. Um, so

Carolyn Woodard

Hopefully we'll see some of that productivity gains to help people who are just stressed out and losing their minds from what they want to be doing with their nonprofit.

Alex Tuck

Yeah, and I mean, I know this is a hot take. Uh, you know, it's really interesting to see um with all the grant funding being taken away in the space. It was NTC was a very sad place to be last year. Everyone was bummed last year at NTC. And um, because all the funding was going away and there was so much uncertainty.

Alex Tuck

But what was interesting is that the nonprofit community did come together and they figured out that, like, with this less lesser funding, like how do we still serve our clients and serve our clients well? And AI just happened to roll out, you know, it was hitting peak, um, just as all this stuff was happening. And um, I don't know, I think the community really stepped up and used used AI in the right ways to um better serve their clients. And um, it was just really neat.

Alex Tuck

It's still not there's not enough money in the space, and it does not make up for everything that's happening, but it was sort of like one of those silver linings that uh came out of that grant funding um cut. So yeah, yeah.

Carolyn Woodard

Well, I wanted to um I wanted to ask you because be before we we got together and we were talking today, um, you had mentioned uh something about more about like the mechanics of using the AI and what your nonprofit clients are, you know, what kind of questions they're coming to you about.

Carolyn Woodard

And one of the things you mentioned was the prompting. And I heard um Johan on the pod uh just last week was saying it's almost like you know, command prompt, like you have to learn the language of what to put in the prompt to get better output and uh not to have to prompt and reprompt and reprompt and reprompt to try and get to something close to what you were looking for to tell you.

Carolyn Woodard

So can you just give me a little, give our audience a little bit more of you know what you advise and what are you finding around prompting and how can you learn to prompt better?

Alex Tuck

Yeah, it's really interesting because again, we are a project management shop, right? And so we we have the luxury of sitting sort of in like we we sort of sit in a co-passenger seat at places, right? Um, so it's like it's almost like we're like making sure that that we're following the instructions on the GPS, right? And we're helping, you know, guide based on what we're seeing on the screen. And so um, you know,

Alex Tuck

The decisions really come down to like how an organization, like what business challenge are they trying to solve? And then uh from there you figure out the tools and then you can figure out, okay, what does the prompting look like? And um, you know, it's really interesting. Like we're we're ClickUp shop and we've invested heavily in the ClickUps like a project management tool, but now it's so much more, and they have this built-in AI. And um so for us, prompting might look a little different than someone that's just using using um Chat GPT.

Alex Tuck

But at the end of the day, there are some like really good elements of of prompting that I think will give you a result that um you know that that you would really want from it. And you know, uh especially when you're building agents, uh having sort of the prompts this way or the instructions this way is important.

Alex Tuck

So like starting with just like a set of instructions, telling the context, um, you know, telling the the the um agent or um the GPT, like, okay, you are gonna act like X type of role, and um you are going to do this task, and here are the constraints around you know what you're going to do. And if you do have access to like, you know, things like standards, so like if you want them to do APA instead of MLA, like you know, you can tell you can give them links to that and um make sure that they that that this agent or um or GPT like has um the diff the different uh sources that you care about, right? So, you know, make sure that you link to your style guide if that's important. And um, you know, just figuring out like

Alex Tuck

What then after you have all that input, you like sort of look at the expected output, right? And so um you can give it like literally the output um structure. You can use code to show it even more. And you know,

Alex Tuck

What I love about it is that um you know it's constantly evolving with the new models, and so really figuring out um, you know, uh how which model works best for yours might not always be the same. Um and um, and then figuring out how to massage some sort of massage uh some of the prompting is good, and then um, you know, thinking about agents for a little bit. Uh, like

Alex Tuck

I have an agent right now that will tell me how my call, and this is this is applicable to nonprofits too, because you can do the same thing about a donor call. Um I feed my transcripts into an agent that tells me how well I did on a sales call. And um, it's so mean to me. Um, it's I'm just kidding, it's it's polite.

Carolyn Woodard

Um did you ask it to be mean to you? Or you know, because I would, you know, there's just the stereotype of they're always so helpful and willing to please. You have to tell them, like,

Alex Tuck

Yeah, it's so polite. No, it actually is very polite in how mean it is to me. Um, but it's like, you know, it's it's so

Alex Tuck

I have this one agent that will tell me, like, okay, what's my probability of you know closing this deal and like what what follow up activity should I do? And then give me feedback about like how I can improve on this call based on this. And so it tells me like sentiment and like all the stuff that's there. And it's really accurate. Like, I mean, really, really accurate with its prediction of the probability of closing. And I think it, you know,

Alex Tuck

That's just like a very nice applicable um use case for like donor, donor engagements as well.

Carolyn Woodard

Um, you've talked to your yeah.

Alex Tuck

But I mean, when you start to get into the agent space, like that's where you can put a lot more context into um sort of how the agent works. Um, you know,

Alex Tuck

With prompting, it's really interesting because like um if you're just going straight into Chat GPT, I actually like a really lightweight approach. So, you know, starting with like giving it context, task, and resources, like just starting with something simple and saying, like, hey, we're gonna do this thing together right now. Like, I need you to think like X, and then starting to work through it. Um, and I

Alex Tuck

I like to ask the agents like to ask me questions, right? Like, what are my gaps? Like, what am I trying to solve? Like, what am I missing with my prompt to you? And it's pretty I find that to be like really helpful as well.

Carolyn Woodard

Um, so yeah, I think that is um that's something I've learned and something I've heard also, is that because it was so easy to start, you know, AI tools, chat GPT, just this phenomenon of everyone trying it and it's so such an easy entry. You just ask it a question. Um, but to

Carolyn Woodard

Now that we've you know had it a couple of years and we're learning about this, the prompting that helps it do its job better and get you the better results, just asking it a question without any context, you're gonna get, you know, whatever it thinks you want from out there, you know, in internet landia of its uh model.

Carolyn Woodard

And so giving it all that context of what you, you know, who it's going to be to help you do the thing that you need it to do, that is just such a um, it's such a change. It's something you have to learn to do. It's not obvious from just, you know, having a little field and putting in your question, but it's something that makes that output so much better.

Alex Tuck

Yeah, it's really interesting. It I

Alex Tuck

I've heard people call the analogy of like, okay, it's just like what happened when like we had Altivista and Google and all this other stuff back in the day, right? Like we were all horrible at searching at first, right? And I still I still see people that are bad at searching today um on Google, right?

Alex Tuck

But you know, like once you figure out, like, oh, you can use Boolean and like all these other different techniques and like quotations and like all this stuff um to improve your Google search, it's a heck of a lot better. You get actually what you want. And

Alex Tuck

I think it's the same thing with you know using AI. It's just and and then going one step further and using agents and then chained agents where you have orchestration agents and like they can talk to each other and um trigger things, like that stuff. It it looks like magic to people. And honestly, like I work in the space and it still feels like magic to me sometimes, right? Um, but

Alex Tuck

It's the same concept, right? And it's the same thing with math and like everything else. Like you can learn how to do it, you can get it done, and then you can get it done really well.

Carolyn Woodard

I have one another question for you around, I said earlier, like you prompt and prompt and prompt and prompt, and it takes a lot of time.

Alex Tuck

Sure.

Carolyn Woodard

Sometimes that is good to iterate over and over, like help it, like it didn't quite give you what you were expecting, or it didn't quite give you something that was useful, and then you can ask it a further question or have it ask you a question of like why didn't that output work for you? That sort of thing.

Carolyn Woodard

Can you talk a little bit about that? Like, how can you add on to your prompt after you've gotten an answer back and it's not quite what you want?

Alex Tuck

Yeah, so a couple things, right? Um if this so I think I I

Alex Tuck

I like to bucket things into two different um types of prompts, right?

Alex Tuck

One is like a prompt that you're trying to solve this very unique thing that you're only gonna do once.

Alex Tuck

And then there's like this prompt that you may reuse again in the future um because it's part of your business workflow. And um, you know, when it

Alex Tuck

When we talk about that second bucket of like asking these questions um like better over time, like I I love prompt libraries for this exact reason. And so to your point, it's like, all right, um let's say that we want to give it context and but we want it to ask us questions and like you know, continue to iterate.

Alex Tuck

We might put that in the prompt library that's like, okay, step one, you know, uh do this prompt and give this context. And then step two, after that, like you want to start taking it this direction, step three, you go this direction. And you can continue to build some of your learnings into those prompts so that they're really nice and baked, so that you and other people can use that in the future with the prompts. Um,

Alex Tuck

And then there's like the um the first part, the first type of prompt, which is like, okay, I have this like research project that I'm trying to do on my own, right? Like I want to learn, you know, how are people using project management? Like, what do people think about project management right now? What are the different types of project management that people, what do people call it? And this is like a one-time exercise.

Alex Tuck

That one is really like that one's those are more fun projects, you know, like prompting projects to solve. Um, where you know, you might start down a path and you're like, and you start to have this feedback, and then you start to see hallucinations and like all this stuff, and you're like, all right, like let's wipe that away and restart. And so, and then um, so sometimes I'll just like go and delete that out and then like have it forget that it ever existed and then restart. So, you know,

Alex Tuck

I think that there's like a couple different ways that you can go based on the type of uh challenge that you're trying to solve with business.

Carolyn Woodard

Yeah, no, that makes sense. And I would say the other thing about getting crazy returns is sometimes people just give up. I know I have. Especially if it's like the core near the end of the day, and you're like, oh, I'm just gonna ask it this one thing and see if I and then you're like, that response didn't make any sense at all. Like

Alex Tuck

That's right.

Carolyn Woodard

I want to go make some tea and touch grass or something. Yeah, but then um,

Carolyn Woodard

I think it's important, you know, like we were talking about workforce earlier and nonprofit staff and um encouraging people to, you know, maybe go away and come back and tackle that problem a different way. Ask the AI in a different way or different question. Um,

Carolyn Woodard

And not just be like, well, the AI isn't useful. It just, you know, it couldn't give me the answer for this one prompt that I asked. And um just um sticking in there, maybe.

Alex Tuck

Yeah. And you know what we do a lot, like I love the culture that we have at my company because like we'll just we'll be like, oh, I'm trying to figure out this challenge. And it might be related to AI, it might just be like some other challenge that we're facing in a project management context.

Alex Tuck

But like just asking someone else, like you're like, I'm trying to do this thing, like check this out. Like, what do you what do you think? What am I doing wrong? Why am I not getting this context or the output that I expect? And um, and just having another eye on it like helps so much, or just like even talking about what your approach is. Like sometimes you say it out loud and you're like, oh wow, like that's that's my problem. I wonder it didn't know what I was asking it. Yeah. Um, it is really interesting. But

Alex Tuck

I do like about giving up, like, I know like as a project manager or like hearing that project management companies say this out loud is sort of scary, but like I love quitting projects, right? Um, but I love quitting projects really early in the process. And so um, especially with AI, things can accelerate so quickly. And so

Alex Tuck

We might try to build an agent or a prompt library, just or a library entry just around like some use case. And we try it seven different ways, and we might spend half a day on it, and then you know, we have to throw like we have to throw it away. And the I

Alex Tuck

The goal is to not spend a whole half day on it or a whole day on it, but maybe you spend half an hour or an hour because you have a you have a um hypothesis that it's something that AI could solve, and then um when it doesn't solve it, it's okay to just like let it go and move on because um, you know, it's just how much time are you gonna save uh at the end of the day, or how much, you know, how much higher is your quality gonna be if you do get this agent or prompt to work properly. So that's always I mean, it's tough to hear and to say, but like quitting, but

Alex Tuck

Quitting early um is really like an okay output, right?

Carolyn Woodard

Yeah. No, it's better to have spent half a day on that project than a month.

Alex Tuck

Yeah.

Carolyn Woodard

And it's not where you want it to be.

Alex Tuck

Yeah, yeah. The best time to quit a project is before it starts, right? But the second best time to quit it is really early in the process.

Carolyn Woodard

Yeah, yeah. Well, I mean, just in terms of not wanting to go down that path, it's not gonna be a fruitful path for sure. Yeah, that's right.

Carolyn Woodard

Alex, thank you so, so much for sharing all these insights with us. It was delightful to talk with you for the first time on the pod. I hope we can have you back sometime to talk more about all of the things that you do and that you're seeing at your nonprofit clients. And I just thank you so much for your for your time today.

Alex Tuck

Yeah, thanks so much for having me. I love your show and I'm excited to finally be a part of it. Thanks so much.