Exploring AI in Finance with Todd Spartz, CFO of LinkLive

[00:00:00] Joe Michalowski: Hello, and welcome to another episode of the Roll Forward podcast. My name is Joe Mic haowski, and this episode is brought to you by Mosaic, a strategic finance platform that transforms the way business gets done today. My guest is Todd Spartz,, c f o at Link Live. Todd, thank you so much for joining me.

[00:00:14] Todd Spartz: Thank you Joe. Uh, glad to be here.

[00:00:17] Joe Michalowski: Love it. So super excited to have you here. Uh, we have a, a really exciting topic today, but uh, before we get into it, do you mind just giving everyone the quick background about yourself, the work you’re doing at Link Live and all that?

Todd Spartz Introduction

[00:00:28] Todd Spartz: Yeah, absolutely. I’m a industry veteran. I’ve been in Silicon Valley for, gosh, 30 years now. Uh, I’m originally from the Midwest. Went to college out there, wound up going to business school out west here in, in the Bay Area. I. Met my wife, uh, and both of us wound up, uh, as in finance in tech. And, uh, we both kind of followed unique paths.

[00:00:50] Um, and mine’s kind of followed the valley as, uh, technology has changed in the valley over the decades. Uh, so is my kind of vertical focus. I started out in semiconductor equipment manufacturing, got into networking. Then, uh, wound up spending a little bit of time, uh, in kind of web analytics, uh, and, and advertising, web advertising, and then wound up, uh, really where I’ve spent the, the vast majority of my career, and certainly all of it as a C F O, which is enterprise SaaS.

[00:01:22] Joe Michalowski: Love it. Uh, first of all, power Couple good for you guys. Also, I too, started out in the semiconductor industry. I, uh, was editing technical manuals for a company called Analog. That was my

[00:01:34] Todd Spartz: Yeah. Very good. Yeah, I know

[00:01:35] Joe Michalowski: job outta college. Uh, 700 page manuals, uh, about semiconductors. Not a fun start to my career, but I hope it was better for you.

[00:01:44] Todd Spartz: definitely something I’d need a manual for.

[00:01:46] Joe Michalowski: It’s, uh, it was a lot. Uh, but we’re gonna talk about much more exciting things today. Uh, I. Was excited when, uh, our customer success team, our co-founder, everyone tagged me in a recording you had with a, uh, someone here Eman, um, talking about ai and we we’re trying to talk a lot about AI lately. I think everybody is.

[00:02:05] Um, and they told me they had a lot of great thoughts about it. So I thought we’d get you on the podcast and just talk about the potential for AI in finance. So as we, uh, get going here, do you mind just setting the stage? I, I would love your kind of like broad thoughts on. Sort of where we are now and the future potential of AI and finance.

[00:02:24] Would just love any sort of general ideas here.

General Thoughts on AI in Finance

[00:02:27] Todd Spartz: Yeah, no, absolutely. Well, you know, technology has certainly changed. Uh, in the, in the three decades I’ve been at it, um, I’d, I’d still though characterize us as being in the first inning. Um, you know, I think it all started with digitization where suddenly your financial systems, your E R P C R M, uh, H R I S, et cetera, started, uh, Digitizing data that you could then access within the system?

[00:02:57] No more. I, I recall time when I actually had to schlep in the rain from one building to another building to get to the filing cabinets where the invoices were, if you can believe that. Yeah. Um, we’ve come a long way, obviously. Those are all now stored in whatever platform you’re using. And then we’ve kind of migrated into automation where a lot of the things that we used to have to spend time doing, uh, are now provided for you.

[00:03:24] That can be everything from. Uh, downloads of data into an Excel file. So no longer do you have to like literally type what you need. Um, also, uh, a lot of business processes are automated approval workflows, for example, purchase requisitions, et cetera. Uh, you know, you don’t necessarily have to access. Uh, bank statements in a paper form anymore, right?

[00:03:46] Those are automated and to a certain extent, uh, reconciliations of cash now occur within the system. So those things, you know, have been available for a while now. Uh, where we’re kind of at is at this precipice of AI where people are beginning to, to use it in their everyday lives. Um, but I don’t think for the average C F O or his or her finance team, That they’re embracing that a tremendous amount today.

[00:04:14] I think pretty much every, uh, c f O out there, certainly the ones in technology are, are starting to kick the tires on things and trying to identify opportunities. Um, I know that, uh, our, our, uh, investor, uh, Invictus, uh, they built an AI platform. They have a, a head of ai. Uh, and that is now being used internally to help us generate qualified leads.

[00:04:41] So our, our BDRs take the leads that the AI says are the ones that are probably most appropriate for our solution. Um, within finance, now we’re beginning to look at some of the tools. We’re looking at the Amazon tool right now to help us, uh, forecast. So that we can try and understand, uh, with all these opportunities that are in our c r m, um, you know, certainly we have the Salesforce view, we have the finance view, uh, maybe marketing, uh, and, and others in leadership, uh, have a view.

[00:05:15] Uh, but we also want an AI view. And, uh, you know, maybe we can start, uh, allowing AI to kind of sniff into our networks and, and see if a rep is. Uh, committed a deal. You know, how much traffic is happening on that deal? Are keywords like budget, achie, or attained, um, uh, uh, lawyer engaged, you know, the, the key buy signals that those, those are happening.

[00:05:42] Uh, and then, you know, to be able to go back and look at that rep and say, Hey, Um, how has he or she performed on their forecast over history, uh, given this, uh, set of circumstances, how does that apply to the current deal and what, what is the a i C as as the opportunity? So, you know, I think some of those things we’re starting to look, look at, but uh, but certainly nowhere near where I expect to go.

[00:06:10] Um, I think, uh, In a relatively short period of time and I’m, no, I have no crystal ball. I wish I did, but I think we’re talking on the order of five to 10 years. Um, you’re gonna be seeing, and I think we’re gonna probably get into some specific questions on that, a great deal more.

[00:06:27] Joe Michalowski: I I love that. And I love the examples you gave too. ’cause I, where I wanted to go next was like where I, I think everyone, anyone listening to this, probably not a huge surprise to know like we’re. Like you said, uh, first inning, like, we’re not deep in this at all. So I, I was curious like what the current ways, like for me, every marketing tool, like I, I, I run content here, so every marketing tool under the sun is promising me some kind of AI powered something.

[00:06:53] So I just keep trying new tools and like I, for me, none of them have really made it into my like day-to-day. So I was curious if there’s anything that has, like if you’re using chat g p t for anything or any of like the more established tools, have you found anything that’s been useful right now today, uh, from the AI stack sort

[00:07:13] Todd Spartz: Yeah, well definitely generally J uh, uh, uh, AI chatGPT and. And, and its cousins Bard. Um, you know, we do, uh, I certainly do use it. Um, if I’m going to have to write policy, I’ll often have it propose a policy. I’ll give it the parameters of our situation and then see, ’cause often, you know, it’s not gonna provide me with the answer, but it will often.

[00:07:40] Uh, provide a, a potential answer that has elements of things that I hadn’t thought about. Um, and, and just most recently, you know, as part of our audit, we had to write a white paper on how we were gonna account for something. And I wrote it up and I’m like, gosh, I just feel like that’s not, not quite right.

[00:07:58] And I, I asked Chat g p t and after it, Basically said, you know, this is not something you should rely on. It did, it did provide a nugget of information on something that I hadn’t thought about, and then I was able to go to the literature myself, find that section and, and paragraph and, and quote it, chapter and verse and give it attribution and, uh, And, uh, so far so good with that audit question.

[00:08:21] So, um, there is some of that. Uh, we’re looking at tools right now. I wouldn’t say we’re, uh, using any right now. We’re looking at, uh, some C L M tools that have ai. So they will propose contract LA language, right? They’ll have our standard boilerplate agreement loaded. It’ll ask you, you know, is this third party paper?

[00:08:41] Is it your paper? And then as you’re negotiating certain red lines, it’ll propose language that. Uh, you should consider using, um, and, and, and ultimately try and reduce the amount of time that you’re spent negotiating and, you know, trying to use, uh, those expensive lawyers, uh, as, as, uh, efficiently as possible.

[00:09:01] Let’s put it that way. So, um, you know, we’re looking at. Some other, what we’d really like, uh, is to, to get a set of tools that will help us do some algorithms, right? So we’d like to not have to sit down and, and build a multivariate, linear regression model that you, you know, pretty much, if you’re me, the last time you did that was in business school.

[00:09:26] Uh, and you know, it just, Isn’t something that, you know, I’ve used, but I could absolutely would like, you know, a probability or Monte Carlo, you know, help some of these really difficult things, make them simpler. Or stock pace compensation, black Shoals model, you know, help, help make these input variables simpler so I can push those, uh, those decisions and some of those, uh, certainly the data collection, uh, further down into the organization, which also gives those folks more opportunity to do some really.

[00:09:57] You know, uh, meaty things. Um, so yeah, I, I think the promises still haven’t been realized. Um, I also, and this is one of the things that I mentioned to Eman, uh, that I would love, and I think there’s a question maybe further down about this. Um, you know, I would love to have an AI go through my dashboard.

[00:10:20] And, and there’s kind of like a public dashboard that I have and then an internal dashboard, right? My internal dashboard might be a a hundred metrics, way too many to really manage a company by really, you know, I think you can only remember seven things. Um, so, you know, I try and keep my public dashboards and that what I.

[00:10:39] What I focus the executive team on, uh, to a fairly limited set, but I would love to have it look at a broader set and look at our budget or our forecast, and then look at. What things are looking like, uh, based on results so far. And recent performance? Actual performance and alert me to things that I might miss or just haven’t had a chance to, you know?

[00:11:02] ’cause typically, right, a lot of us CFOs will wait till the board meeting. We get all this data from our teams, we’re pouring through all of it, trying to craft a story that we can deliver to the board that will give them a sense that, yeah, we’re, here’s where we’re at. You know, this is where we wanted to be.

[00:11:19] Uh, here’s where we’re gonna course correct and, and here’s where we’re gonna add rocket fuel to things that are going well. You know, it would be nice to be able to, ’cause I’m sure like any other, uh, C F O, we’re all wearing multiple hats. We’re all incredibly busy. Um, it would be great to have that assistant right that can, uh, comb through the data and potentially serve up some things for you to look at.

[00:11:43] Right, because it wouldn’t be that hard. Like you could create, you know, a, a flow chart of, you know, these are the 1000 things. Tech company CFOs look at and, and, and in this order, and it could follow through that in what, a second or less. And, uh, and serve up something that’s probably, uh, better than a lot of, uh, staff members, right?

[00:12:07] Because they’re obviously also working under deadlines. You know, the, the AI never sleeps as they say.

[00:12:14] Joe Michalowski: AI never sleeps. Should be the tagline for this one. I love all this. This is great. I mean, and we talked a little bit about this before we started recording, where it’s just like, you know, obviously first inning, again, it’s like there none of this is gonna be entirely like, Hey, go use AI to do this today.

[00:12:29] So I love that we’re kind of like throwing out ideas and uh, It just seems like you’re a, a wealth of knowledge on like ways that you want this to, to go. And we’re gonna get into some more like workflow specific ideas, but before we get there, uh, I want to talk about what you think are some of the, the challenges to making AI a bigger piece of the finance, uh, workflow.

[00:12:52] Like, this is not a function or the things you work on are not like, I dunno. I have like Grammarly and it’s like checking my grammar. Like for like articles I write, like that’s a like simple workflow with like very strict rules and like low leverage. Like there’s not a lot of like risk in this process.

[00:13:10] There’s a lot of leverage, a lot of risk in what you do every day as a cfo. So I’m curious, like are there just roadblocks you see, uh, AI companies or, or companies like Mosaic kind of running into as they try to roll this out so it’s useful to somebody like you?

Roadblocks to AI Adoption in Finance

[00:13:27] Todd Spartz: Yeah, I, again, I think we’re so early on that, that, you know, uh, the, the users of, of some of this technology haven’t really experienced a great deal of it. Um, but I, I do think, uh, you know what, what you’re. I think likely to run into and, and I was reading something, I think it was by Mark Andreessen, who has young children, and he was saying something to the effect of, you know, My children are growing up in the age of ai, they’re going to begin, you know, I don’t, I think his kids might be in elementary school, so by the time they get to middle school and high school, and certainly by the time they get to college, they’re gonna, they’re gonna have spent years and years working with ai.

[00:14:09] And I think, uh, and internally, uh, again, Invictus, they’ve kind of pointed out to us. Kind of the history of it and, and how we’ve kind of wound up where we’re at and, and I think one of the roadblocks internally is just fear of it. And lack of understanding how to engage with it. Because like any relationship, you need to work at it, right?

[00:14:31] It’s not the ai, at least the early, uh, development of it, it’s not, it’s gonna be kind of cookie cutter. It’s one size fits all. It’s not gonna be tailored. I. To me, Todd Sparks or, or you, Joe Michalowski right? It’s, it’s gonna be, uh, something that we’re gonna have to learn how to, how to handle, uh, you know, and, and learn how to quickly, uh, identify, uh, what’s accurate, what’s not.

[00:14:59] How to curate things that work and, and avoid or ignore things that don’t work. Like I think there’s a lot of that and you know, I think for those folks that are. Particularly earlier in their careers. I think they just really need to embrace it. Like you need to continually exercise that muscle because, um, you’re gonna be light years ahead of, of, of your counterparts that don’t, I think somebody said something, I’ll get this quote wrong.

[00:15:29] Uh, you know, you know what, what makes a great. Ai, AI lawyer or something. There are those lawyers that are good and there’s those lawyers that are good with ai. Meaning you can, you can still, you know, do your job, but you can probably do your job better if you know how to use it. And so I think, I think the biggest, uh, roadblocks internally within finance organizations are just.

[00:15:56] Learning how to use it. And then I think, uh, you know, we’re just kind of waiting patiently for, for the waves of greatness to roll apart, wash across us.

[00:16:06] Joe Michalowski: I, uh, I love that. And I think, I think it’s true probably of any organization, it’s, it’s true, at least in my circles as well. It’s just like, you know, every write like I’m a writer by trade. That’s how, that’s how I ended up, you know, running content. I do more than write, but for people who are like freelance writers, they’re like, well, like ai, is AI gonna replace me tomorrow?

[00:16:24] And, you know, there’s no sign that, you know, people will be completely eliminated. Not for quite a while, but. I think it’s a really great point. One one that I wanted to ask you about specifically. Um, and it’s just like data management, like from the outside looking in, like there’s a lot of private data that you’re working with as a C F O.

[00:16:43] Like, is, is this something that we just assume, like teams like, like we’re SOC two compliance, like I assume like anything, any company that is compliant with these kinds of things, like. We’ll handle that with AI as well. But as a C F O, are you concerned about data privacy or data compliance, uh, from the AI perspective at all?

[00:17:02] Todd Spartz: I, I, I definitely am. Um, you know, we, we all need to be aware and, and I wish I would’ve had this, uh, this next seminar. I think, uh, uh, Invictus is holding, it’s, it’s specifically on privacy and ai. because, uh, What you share with AI is, is public, right? It there, there’s no privacy promised or delivered. Uh, and so you’re, you’re absolutely right when you start opening up.

[00:17:29] Some of your most, uh, you know, proprietary data. Certainly customer lists have to be way up there. Maybe even customer prospects, even more so, like, I can’t think without, you know, except for maybe your IP on your product, your, your, your prospect list, maybe among the most. Personal, uh, sensitive data in a company and to just turn that over to a, to a bot that really promises nothing, uh, is, is something to be concerned about.

[00:18:00] You know, I, I wish I, I did have the answers, but, but I am and encourage everybody. Um, you know, I mentioned that seminar I’m gonna be attending, but I, I would definitely, ’cause I think, you know, we’ve, we’ve already seen where there’s some lawsuits now. Uh, in the AI community where these, these, uh, owners of content are suing, saying, Hey, you’ve, you’ve accessed personal data that wasn’t intended to be for your use to help you.

[00:18:30] Uh, you know, particularly in the generative ai, you know, build your large language models. And so, Uh, who knows how that’s gonna shake out. Um, you know, is it, is it just a, a roadblock or is this a, you know, full on wall that’s gonna be difficult to scale? Um, I, I, I think it remains to be seen, but I, I, I do think, um, it will be a while before I.

[00:18:55] We have definitive rules. Um, you know, the A I C P A has guidelines, and so, um, I would encourage, encourage us all to just be cautious and be aware of it. Um, and, you know, obviously try to mitigate to the extent you can any, uh, you know, uh, loss of of, of, uh, ip.

[00:19:16] Joe Michalowski: Yeah. I, I think it’s such a, a great point. I’m, I’m glad we, you know, talked about it a little bit because I think this, I mean, to me, this feels like maybe one of the first times that, uh, people are being forced into early adoption. Like it’s, it’s that. Sort of prevalent, like AI is sort of that powerful or prevalent in every industry that even people that might’ve kind of like lagged behind, uh, tech in the past, like everyone sort of has to get ahead of it.

[00:19:45] And so we’re all kind of just like feeling our way around together. Uh, and so I like what you said about, you know, being cautious, just kind of knowing what you’re going into and, and just trying to be smart. ’cause no one has. The exact answers, uh, but as long as you’re kind of mindful of it, you’re gonna be in a better spot than if you were just blindly trusting every AI tool that came by, you know?

[00:20:06] Todd Spartz: Yeah, I think too, um, certainly adoption curves have have accelerated as we’ve gone through time. Um, and, and that likely won’t change. And, and I think I. Um, you know, ai, you know, will be realized more quickly, uh, maybe than something, but I think I do think the market may be getting a little bit ahead of itself.

[00:20:29] Um, you know, certainly, you know, autonomous vehicles, you know, I don’t know how many years now Elon’s promised us that the cars are gonna drive themselves. Um, it’s gotta be at least five years now. Um, and, you know, I think back. To when I very first, you know, I’m, I’m that old. Very first saw, uh, a browser, you know, it was Netscape, Mark Andreessen’s, Netscape.

[00:20:53] And I’ll tell you, there wasn’t a whole lot of great content on the web at that time. There really wasn’t. And, uh, it took, you know, this was probably, uh, 1990. 2 93 maybe. Yeah, 93 I think. Um, and so, you know, fast forward, uh, you know, 30 years, um, it’s taken us that long to kind of get where we are today. So it didn’t happen overnight.

[00:21:22] Um, I never would’ve dreamt some of the things that we can do today on our, on our smartphones. Um, I think that’s probably true of ai, but this isn’t gonna happen overnight. It, it’s going to be, you know, a process. It may not take 30 years like the, like the web did. Um, but uh, you know, ’cause I think, are we saying we’re in Web 3.0 now or is it still

[00:21:45] Joe Michalowski: I lose track. I mean, knows? Who knows which data we’re in? It just I, it could be four. Who knows?

[00:21:51] Todd Spartz: Yeah. So, you know, things, things take longer, so I don’t think. You know, we need to rush out and, and, and make rash decisions. It’s, I always say, uh, it’s better to be right than quick, but, um, by the same token, um, it will come. So, you know, get ready.

[00:22:08] Joe Michalowski: I, uh, I love this. This is a great discussion. I want to, I want to get even deeper. You’ve mentioned a lot of great, uh, potential use cases, things you’re thinking about, but I want to kind of bucket it into what I think are sort of the three. I mean, this is a very much a simplification, but like kind of three buckets of kind of what finance and accounting are responsible for.

[00:22:30] And my thought is two of them are just like the analysis side and the planning side. But first, before we get there, I I, I want to hear what you think could realistically be like fully automated by ai. And so I’m thinking like, Some of those transactional workflows, accounting, compliance, things like that.

[00:22:47] Like what is that baseline that you would set where you’re like, I hope that AI completely automates this piece so that I can go and, you know, we’ll get to the analysis and planning side in a minute and sort of talk some of about those strategic use cases.

Hopes for the Future of AI

[00:23:00] Todd Spartz: Yeah, so transaction processing is definitely one I, I expect that to be achieved fairly soon. There’s already. When you look at some of the vendors out there, like Tipalti, um, AP is largely automated and they’re leveraging ai. They still have some humans involved, but you know, that’s basically a situation where you’re outsourcing AP even though it’s.

[00:23:24] Internal to you right there. The AI is booking, uh, the payables in your e r p for you on your behalf with some review and controls. Um, and certainly receivables and collections that can all be automated. Reconciliations, you know, basic consolidations of, of subsidiaries in books, that that can all be automated and probably will be, you know, I’d be surprised if.

[00:23:51] If there, if there isn’t the ability to do that in a, in, certainly in a simpler enterprise in the next five years. Um, you know, when you get into things like compliance and controls, that’s another thing that, uh, finance is responsible for. Um, you know, building a system of controls. And then testing and alerting to, you know, uh, to, to areas where you’re, where you’re not meeting those controls and you need to course correct.

[00:24:19] Like, those are things that AI can absolutely do. Um, and again, I’d be surprised if, if that largely isn’t in place for, again, smaller, simpler, uh, more agile organizations in the next five years. Um, so, you know, some of these. You know, going over some of those so controls, uh, you know, trying to figure out, you know, what the best ones are and then, you know, testing them and, and, and, and fixing ’em.

[00:24:47] That, that takes a lot of time and. Uh, that’s absolutely something that, uh, that an AI is built to do. Um, so, um, I, I absolutely think that’s the case. And then, yeah, you know, uh, so if you’ve got kind of the back office that’s focused on transaction processing and recording and, uh, ensuring that things are in a controlled environment and you’re, uh, compliant with all the rules and regulations of the industries that you’re, uh, competing in.

[00:25:17] Then you’ve got the front office, right? That’s really dealing with, um, uh, forecasts and budgets, right? So trying to figure out, we’ve got finite resources, um, we’ve got opportunities to, to invest those resources. Where should we do it? And then, uh, you know, making those commitments. And then, um, you know, I think most of that will still be, at least in the near term, uh, People doing that, right?

[00:25:47] Strategic finance, making the investment decisions, but then the budget versus actual highlighting the Deltas proposing solutions as to where those deltas are coming from. You know, gee, you know, your head counts five over plan. Uh, you know, we’ve looked at the open Rex. There’s, you know, people were hired without a rec, you know, um, it, it, it’s absolutely something that AI could do.

[00:26:12] And then, like we touched on it a little bit earlier. You know, algorithms that help you forecast volume and, and price. You know, when we negotiate with other competitors in the banking space, you know, other money center banks, what kind of discounts do we typically give? How long does that deal typically take to close, you know, based on the email traffic with their, uh, purchasing department, you know, how likely is that deal to close all that stuff?

[00:26:38] Uh, I think it could help, uh, assist in. And so, uh, you know, helping, uh, build some of the budgets or proposing, you know, so a, a budget planner, uh, you know, somebody in the marketing or sales or operations organization, rnd, whatever can say, you know, be presented with, you know, here’s the trend. This is where you bend.

[00:27:01] This is where we predict you’re going based on other things that we’re seeing. Um, Is there a reason that this isn’t right? Oh, yes. That, that’s absolutely incorrect. ’cause little did it know that we’re gonna make this huge investment in this great new thing. Right? Um, so again, that’s where, you know, people leveraging these, uh, assets, uh, will really, I think, be a, a true, uh, will really add the most value.

[00:27:30] Um, so, so yeah, that’s, I think, I think the, the, the budgeting, forecasting and planning, there’s the strategic element and there’s the more tactical and analytical, and I think the last two, it can probably, you know, that might be closer to 10 years down the road, but I, I fully think it’ll get there. But I, I don’t see it.

[00:27:49] I. In the 10 year horizon making investment decisions, right? Uh,

[00:27:55] Joe Michalowski: totally.

[00:27:56] Todd Spartz: we should buy this company to get us into this industry, right? We should, we should, you know, stop investing in our product and, and spend it all on customer support. Let’s just keep the customers we have and make ’em super happy and not get any new ones.

[00:28:11] Like I just, that’s not something an AI would do, I don’t think, and probably not too many finance individuals either, but, but that highlights where you could make a decision that maybe, uh, is right for you, uh, but isn’t right for a lot of other places.

[00:28:28] Joe Michalowski: Yeah, totally. Uh, love all that. Love the breakdown there. There’s two specifics. That I wanted to ask you about on the first, on the analysis side. Um, and sorry if I’m putting you on the spot, but I did see, I saw it in the call with Eman, you mentioned predictive churn as like a potential example. Like I remember you, uh, I listened to the little clip and were like excited about this potential use case.

[00:28:48] And so I would love to hear sort of your thoughts about what that would be. ’cause it, to me, it was just an interesting, uh, idea.

[00:28:55] Todd Spartz: Yeah. So, um, You know, when we’re building forecasts, we’re, we’re particularly for, for, uh, CFOs and SaaS finance, right? So we’re selling subscriptions to customers. It’s incredibly more difficult to get a new customer, uh, than to keep the one you have. Um, but keeping the ones you have is also difficult, right?

[00:29:18] There’s other competitors out in the marketplace. Um, there’s always pricing pressures. Uh, people move and they wanna work with tools that they’re familiar with, right? Um, so how do you, in the face of all that, how do you know what your, your churn is likely to be? And, uh, you know, I’ve spent time with customer success organizations that, you know, are on the phone with a customer every day.

[00:29:44] And that’s just one view, right? That customer could love you. You could absolutely be delivering everything they ever wanted and more. Um, but if there’s a mandate that they’re unaware of from their c f O to go with a competitive product because they’re getting a new investor that’s has a portfolio company that uses that product, like you can be blind.

[00:30:09] That person can be blindsided. So there’s, there’s obviously multiple. Uh, decision criteria that go into a decision every year or three years, whenever your renewal’s up five years. Um, To, to do we stay or do we go? And, you know, I think it’s, it’s not one or two, uh, variables, right? It’s, it’s a lot of variables.

[00:30:29] And, you know, to be able to go into your application as one example. And, and c, how often is that customer using your product? How a, how many, how much, uh, adoption is there? For your product. Uh, and, and if you’re seeing more and more users over time, and if those users are daily active users and their daily active usage is increasing, that’s a really good indicator.

[00:30:56] And that one would probably be one of, you know, my higher, more heavily weighted attributes. Right. Um, but there’s other attributes in there, right? That getting, uh, data from the users on, um, you know, some kind of analytic. Um, on how happy they are, would they recommend you or not? Right? Um, those are often really, really good indicators.

[00:31:19] Um, you know, payment history, if, if you’re always struggling to get them to pay for your product, they’re probably not too afraid of losing it, right? Um, I know, uh, uh, for a lot of customers that use, uh, a certain C r M vendor, they don’t wanna be behind ’cause they will get shut off and that’s where your pipeline is and you can’t have that happen.

[00:31:42] Right. So, You know, there’s probably 10 factors. Um, and it would be fi you know, how much investment have we made? How, how many new features, functionality have we, how do we fit competitively? You know, where you get those little pie charts and you look at all the competitors, like, how has that changed over time?

[00:31:58] If we’re seeing. More and more competition catch up to us, um, disrupt us, then, you know, maybe that’s a concern. And, uh, certainly if we’re priced more, uh, I think we should be very concerned. So anyway, yeah. I, I, I think that’s, again, that’s an algorithm that, that, you know, AI is built for far more than I am, let’s put it that way.

[00:32:21] Joe Michalowski: No, this, this is awesome. I, the, like I said, the, the ideas, the, it, it’s great to hear you talk about like where you think this technology could go again. Uh, we’re not there yet, but these are all like real realistic things that we can expect to see in the coming years. And so the other one that I wanted to talk about and mostly just because as we’re recording this, it’s August 3rd and like planning season is, If it’s not already upon you, as you listen to this, it’s about to be, uh, so I’m curious like how you think AI could change, like the annual planning process in the future, because we’re, we’re doing some content about it and we’re talking about the pretty structured timeline process, the traditional approach, but planning is just getting faster and faster.

[00:33:09] So I’m curious what you think AI could do to that kind of traditional structure.

[00:33:13] Todd Spartz: That’s a, that’s a great question and I’m, I’m thinking, you know, first I’ve heard it and, you know, look, I, I think, uh, I’m a big believer and, and again, I’ve been c f o of, of more entrepreneurs. Entrepreneurial companies, uh, not multi-billion dollar public corporations. Um, but what I’ve found is I always try and, uh, hitch my, uh, wagon to a horse that’s a, a best of breed that really is a leader because what they do is hopefully, uh, if not force strongly encourage you to adopt best practice.

[00:33:54] Right. And I think for a lot of us, um, you know, at least when I went to school, you didn’t study what the optimum planning calendar is. Right. I just kind of got out into the world and you know, my boss said, Hey, you know, you gotta develop a budget for this. And I kind of sat down and came up with a list and put some dates on it, you know, started with the day it was due and worked backwards.

[00:34:18] You know, I would think in time, uh, a company like Mosaic could absolutely, as part of their planning and budgeting application, have boilerplate planning calendars, right? So that you’re literally following step by step. What you should be doing in order to get your plan completed as quickly as possible.

[00:34:38] And because you, you can, uh, multi-thread, right? You can have parallel tracks. It becomes difficult. I think for me, I, I tend to not be someone that’s great at multitasking. Um, you know, I’m not one to be watching television while working on a spreadsheet and talking to my, uh, Parents. Like, that’s not something that I do.

[00:35:03] Um, so I would do those things in sequential order, which takes much longer. Hopefully I’m doing it better. But, um, AI could absolutely parallel those things, right? And it could make sure that things are following, you know, the critical path, right? Uh, Gantt charting, you know, that’s, again, I’ve never, I, I know they exist.

[00:35:24] I, I, I don’t Gantt chart my planning or my close calendars. Why couldn’t there be a, a, uh, a tool in a planning system that says, Hey, have you done this or this? Why not? You should. Right? And, and I think, and, and serving up all the right analytics, right? Like if I need to meet with r and d to go over their budget, it would be great if it can grab all the right data so that I show up to that meeting.

[00:35:54] Uh, ready. Right. And I’m not, you know, trying to carve out an hour or two before that meeting so that I can assimilate the data and, and, and, uh, distill it in kind of a, you know, understandable format. So, um, I, I think there’s a lot that can be done there, um, uh, around process flow, um, like. Sending emails out, Hey, you know, department manager, you’ve only got three days to get your budget in.

[00:36:23] You know, and, and we’ve looked at your calendar and we’ve looked at your financial analyst calendar and you both are free tomorrow at two o’clock. You know, like, why not? Right? So, uh, really simple administrative things like that are so easy, and I’ll be honest. I don’t hear a lot about those things. I’m sure we will in time.

[00:36:45] Um, but I, I think, you know, you start doing some work there and you start, you know, highlighting, oh, you have a major variance on this that you probably at a metric you probably don’t wanna look at or don’t look at, but it’s very predictive, right? So you should take a look at this. Like, those types of things I think will be super helpful.

[00:37:04] I, I wish I had them 10 years ago.

[00:37:06] Joe Michalowski: I, uh, I love this and, uh, you can be sure. I’m gonna grab this clip, I’m gonna send it over the product to you and they can get to work on building, uh, the Todd Spark specific AI tool. ’cause I, I feel the same way. I feel like when I read about what’s coming, like what, what’s on the horizon, it feels very, uh, high level to me.

[00:37:24] Like there’s a lot of big ideas about just like generally, like how it will transform everything. But what I really love is that you’re, you’re giving me a ton of just incredibly specific. Use cases, albeit like potential, like things we would like to see. But it’s, it’s really refreshing to hear that as opposed to just like kind of the, the normal discussion you hear about how AI is gonna transform everything.

[00:37:47] So I appreciate it. Um, however, I would like to trans transition to something that is important to me that I also don’t hear people talk about. And I, I want to know what you think ai. Will not touch. Like what? What are the things that we don’t think AI will transform? And so to put it another way, like by automating, by bringing all these use cases to light, what is that doing for you as the C F O?

[00:38:12] How is it elevating your strategic role so you can take on the human tasks that won’t be sort of automated or efficiency away, I guess.

Limitations of AI in Finance

[00:38:22] Todd Spartz: Yeah. You know, if we assume that, uh, the Ai, AI that we’re talking about is not sentient, right? It’s not, it’s not thinking. Um, then what it’s doing is it’s, it’s taking huge data sets and distilling, uh, actionable, uh, Uh, uh, processes, right? It, it’s taking all this information and highlighting, these are the things you should look at, or I’ve, you know, these are the steps you should take.

[00:38:48] Um, you know, like diagnosing a patient, you know, based on your blood results and your symptoms. We think you have this right, but it, it can’t, it can’t think. And so, um, I, I think we need to be aware of, uh, communication, leadership, mentorship. Training, you know, uh, even in this remote world where a lot of us are, are, are not face-to-face with our, with our coworkers, you know, there are things that we’re not picking up some non-verbal cues.

[00:39:22] Right? And, um, I know I’ve attended offsite meetings where I’ve had some aha moments, some true light bulb events where. I had never, ever thought about it. And whether it was something somebody said or did, or an event that we were at, or I made a connection with somebody that I had never met before and he or she caused me to think about something in a different way.

[00:39:46] Those, those aha moments or those light bulb moments, those are not ai. Uh, and those are gonna be incredibly important. And so I think we need to be, uh, very intentional. As, uh, leaders of companies. To making sure that as we adopt these tools and as they give us, uh, more time to be strategic, that we capitalize on it, right, and that we’re creating opportunities that en enhance or increase.

[00:40:20] The frequency with which these aha moments happen. Like I know there’s a lot of, I’m sure boardrooms of, of property management companies sitting around. How can we make offices re really relevant? Like how, how can they be constructed, organized, used in a way that will foster, um, Behaviors that are better than, you know, meetings over Zoom, right?

[00:40:46] Like, and, and so I, I think every organization there are gonna be things that AI cannot do, and we need to make sure that we optimize for, uh, those, those opportunities.

[00:41:02] Joe Michalowski: I love it. I, I appreciate you indulging the question. ’cause I, I think it’s as, as Mosaic as I, you know, work on content for this topic. I, it’s really important to me that there, it never feels like, you know, we’re replacing people and I think I. That’s like a line you have to balance. ’cause as you keep saying, oh, we’ll we’ll do all this automation, we’re gonna transform everything.

[00:41:23] It’s like you mentioned before, the fear of it. It’s like, okay, well if you automate all that, like what am I doing? And so I appreciate you, uh, kind of sharing that. ’cause I, I think everybody knows it in the back of their head, but to me it, it’s really important to just keep bringing it up because. I think, you know, we’re, we’re 40 plus minutes, 40 plus minutes of talking about all the amazing things AI’s gonna do.

[00:41:43] Uh, we should give a nod to the people now and then. So, uh, yeah, like I said, very much appreciated. Um, I, I wanna be mindful of time. I have two last questions and we can make them, uh, quicker. We can take as long as we want, honestly. But, uh, selfishly as Mosaic, like, you know, we have features coming out. We have features that are in pipeline, but.

Key Takeaways and Career Lessons

[00:42:05] I’m sending all this to all the product people at the company so that they can start, uh, getting more ideas about what we can do. But I would love to know as we wrap up, what you would love to see from a vendor in this space as like the top three priorities. So of all the things we’ve talked about, what are you hoping are, like the, the big meaty things that AI tackles sooner rather than later for finance.

[00:42:29] Todd Spartz: Yeah. You know, I think, um, you know, I think what we’ve enjoyed immensely about, uh, mosaic is not only is it a great platform that allows you to aggregate all your data sources and use prebuilt. Charts, graphs and LG and metrics to, to, to get real results more quickly. Um, you know, that’s been fantastic and, and probably your secret sauce is your team.

[00:42:55] Uh, just tremendous people that I’ve worked with there that have been very dedicated. Helpful. We’ve had some challenges. You know, we’ve done some things in our past that made things more difficult for us today and, uh, you know, we’ve had to do some workarounds and, uh, your team has, has really, uh, you know, measured up in, in every situation and, and has delivered.

[00:43:18] And so I, I definitely, definitely appreciate that. But I, you know, I’m never satisfied. Uh, and so, you know, again, it would be, it would be great to have kind of a, a checklist once you’ve built all these dashboards, um, and you’ve got budgets, plans and forecasts, it would be great for this, for this, uh, application to then start working through.

[00:43:42] Uh, a knowledge base of these are the things that you should, you know, kind of c f o 1 0 1, right? And, and maybe, maybe that’s, Hey, we’re gonna interview, you know, a hundred CFOs and get their opinions on various things. Right. Um, I think, uh, that could be super, super helpful, uh, for, for a lot of us. Um, and, you know, I, I think, um, you know, some of the other things that we talked about, like having, uh, uh, You know, kind of prebuilt, uh, calendars having, uh, you know, um, having some, uh, knowledge bases built around, um, what to do.

[00:44:26] Like, Hey, I, I noticed this. It’s clear like this is a problem. What should I do next? Or how do I, how do I solve this? Like, there could absolutely, and maybe the five suggestions don’t apply to you. I. But maybe one of them does. And you didn’t think about it because, um, I’m a huge, huge believer in, um, you know, uh, Multiple people are better than one, right?

[00:44:54] Like, I don’t know the answer to everything. That’s why I surround myself with, with talent, right? Nobody’s good at everything and we tend to do the things that we’re good at, and we tend to not exercise the things we’re bad at so much, which means we aren’t getting any better at those bad things. And so, you know, to, to have a crutch there that can, that can help you because one 800, uh, you know, analyst on call.

[00:45:18] doesn’t always work right. And so the more that you can kind of get some of that, or, or in a really small organization, a startup organization, you know, they don’t have the budget for that. It might be the CFO’s all they got, or their VP of finance or whatever that person’s called. So, um, you know, trying to to build some assistance, um, and, and some knowledge bases, I think would be, would be super helpful.

[00:45:42] Joe Michalowski: I love it. Uh, I think you’ll be excited about what’s coming in the next, uh, as we talk, you know, 45 plus days, but, uh, actually less than that. That was, that was a timeline I got a while ago. So stay tuned. Uh, ’cause I think, I think you’ll be excited about what the product team’s rolling out soon. Um. I do wanna, I wanna get to my last question and it has nothing to do with a, or I guess it could have something to do with AI if you would like it to.

[00:46:04] Um, but I ask everybody that comes on if you’re to listen and for anyone listening, probably sick of me, uh, presenting it that way. But I would love to know what is one thing you know now that you wish you knew when you started, uh, your finance career? I know it’s been a, you said you were a veteran. I would love to know, uh, the big lesson learned there.

[00:46:22] Todd Spartz: That’s a great question. I mean, um, You know, I think, uh, my finance education and my early career in larger finance organizations prepared me well. Um, and I don’t think there were too many surprises about what I was doing, right. Whether it was closing the books or building plans, budgets and forecasts.

[00:46:48] But I think the one thing that early in my career, I didn’t do, Anything on, uh, that I’ve now come to realize is super important is building networks and communicating with others, like being. Uh, identifying with the subject matter experts, uh, you know, belonging to a group, like making those investments, whether it’s, uh, you know, some LinkedIn group or it’s some trade show that you go to, or a panel that you sit on.

[00:47:22] Uh, or a panel that you watch, right? And then go up to the speaker afterwards and introduce yourself. Like nobody, you know, I kind of went into finance, I’ll be honest. ’cause I was kind of a shy freshman in college, you know? Uh, and I figured I’m not gonna be an engineer that, that’s just too much math. Uh, but hey, finance, debits, credits, uh, addition, subtraction, little bit of algebra.

[00:47:47] I can manage that. So, uh, I never, it never dawned on me that, um, at some point I would need to be making presentations and making introductions and, uh, solving problems that I didn’t have answers for and would need a group to go out and, and, and, uh, lean upon, right? Uh, for, for help. So, um, and, and, you know, those networks are, Hugely, hugely beneficial when it comes time to, to recruit.

[00:48:19] Um, you know, recruiters are fantastic. Um, but you know, The folks that I’ve sourced through my own network are almost always better. Um, so, uh, uh, I, you know, I can’t say enough about, uh, taking the time to, to physically, if you can, uh, go out and, and, and meet some other people in your, uh, industry and in finance, um, you won’t regret it.

[00:48:48] Joe Michalowski: I love that. Uh, I think it’s a great answer. It’s a, I love this question and I, I get, I always say this, it’s like you never get someone telling me, I never, I’ve never had someone be like, oh, I wish I. I was better with my Excel shortcuts in my younger years. It’s almost always like something to the effect of working with people better.

[00:49:08] Uh, Networking to your point, always like, you know, maybe it’s a specific function that they wanted to understand more. And I, I really love hearing about that because, you know, I mentioned wanting to talk about like what the people role is still in finance even with ai. And I, I think that’s it. Like without it, like AI is not, uh, replacing any of that for sure.

[00:49:28] So, uh, yeah, I really love the answer. Appreciate it. Um, we, that was the last question I had, Todd. I just, I just I the end and I was

[00:49:36] Todd Spartz: Joe, thank you. It was, uh, it was a privilege to, to spend 45 minutes with you.

[00:49:41] Joe Michalowski: No, this, this was great. I want, I wanna turn the floor over to you. Where, where can people go to connect with you to learn more about Link Live? Anything you would like to, to promote to the, the finance audience. Uh, the floor is yours, Todd.

[00:49:53] Todd Spartz: Oh, absolutely. Well, yeah, I’m always willing and able to, to take questions or, or connect with others that have interests, uh, that are aligned with whether it’s ai, finance, or uh, link live, which, which is, um, I. Obviously using AI and its solutions as well. So, um, yeah, you can reach me. My email is, uh, tSparks@linklive.ai.

[00:50:18] Joe Michalowski: Yeah. Love it. Well, thanks so much for being on the show, Todd. Uh, I hope we can do it again sometimes. This is really fun.

[00:50:23] Todd Spartz: Sounds great, Joe. Take care.

[00:50:25] Joe Michalowski: You too.