The Decision Litmus Test: Why CFOs Should Question Every AI Investment

Megan - 0:54: Today, I'm joined by Mark Gaudiosi, Chief Financial Officer at Burro. Mark brings more than twenty-five years of finance leadership experience across technology, healthcare, retail, and high-growth companies. Throughout his career, he has helped organizations scale through rapid growth, acquisitions, and transformation, including leading finance at GoPuff during a period of extraordinary expansion, and most recently guiding Medical Guardian through profitable growth and M&A. Today at Burro, he sits at the intersection of finance, robotics, automation, and AI, helping build the future of work in industrial and agricultural environments.

In this episode, we'll explore how finance leaders can use data and AI to make better decisions, build competitive advantage, and create measurable business value while separating real opportunities from the hype. Welcome to the show today, Mark, and thank you so much for taking the time to be here.

Mark - 2:05: Of course. Thanks for having me.

Megan - 2:07: I'm looking forward to this conversation. But to start, can you walk us through your career journey and how working across startups, large enterprises, healthcare, and now robotics has shaped your perspective on finance leadership?

Mark - 2:24: Absolutely. I've had quite an interesting and fun career at this point, I think. I got to start at an educational retail toy company called Zany Brainy, and worked there for about seven years and got to see the entire life cycle of a company. So from startup to IPO, ultimately to bankruptcy. And so everything I learned in school went out the window and everything I learned in those seven years taught me everything that made me who I am today. Taught me everything from how to make decisions, how to view things. And that's just a lesson that I've carried with me throughout my career.

And when I left there, I went to a company here in suburban Philadelphia called Wawa, and it couldn't be any more different than Zany Brainy, right? You're talking about a large iconic company here on the East Coast, very disciplined, very operational. And I took what I learned at Zany Brainy and built on top of that, how to build plans and processes. And I took that from there to Aramark, to eBay Enterprise, and on onto GoPuff.

And so that Radial, Aramark, eBay Enterprise, where I flipped over to the technology CFO role a little bit, but GoPuff is really the hyper-scaling part of my career. So got a chance to grow revenue at an incredible rate over, I think it was 1,200% over a five-year period, growing the finance function, growing accounting, growing valuation 25 times. And that pressure scenario pressure-tests everything I knew about what systems and process and finance and accounting were all about.

Then I went to the medical device company, did a $200,000,000 acquisition and now moved into robotics, which kind of brings everything together. I don't know that I really meant it that way, but when you think about autonomous robots and all of the operational data that we see today in the world that we live in in 2026, I think everything that I've done up until now kind of got me ready for the AI world that we're working in today.

Megan - 4:44: I have two questions, one about GoPuff and one about Burro. So first of all, at GoPuff with growth like that, how did you learn to stay ahead of it? How did everything not break?

Mark - 4:57: That's a phenomenal question. It's interesting. And you're, I always liken it to building an airplane while you're flying it, right? When you think about an airplane, you have to look at the altimeter, you need to know where you are at every second. And clearly we were building and flying at the same time.

And on top of that, we were managing capital raises with some serious investors like SoftBank, P1 Capital, Accel. And what I focused on in that time and what we shifted the organization into is the data. The data really gave us the confidence to understand, build and grow the business. And that confidence grounded in analysis allows you to make decisions more quickly as you're going through scenario building and what-if scenario analysis, as opposed to just forecasting the future.

And I think that's really what it was all about, is that it was about disciplined thinking more so than sophisticated modeling. And I think that's where the data is so impactful and really helps you and every company grow at that stage.

Megan - 6:09: And just real quick, Burro, what is it exactly that you guys do? Sounds like you're doing some really cool things with robotics.

Mark - 6:16: So we are an outdoor robotics company. So think of an automated vehicle, fully autonomous, fully electric, doing work outdoors. So the indoor warehouse space is very crowded. People do a really good job there, but nobody's really tackled outdoors from a physical AI perspective. And that's what we've done and built over the last nine years.

And so we've sold over 800 robots across the world, and we are tackling the $2,000,000,000,000 labor problem by putting an autonomous robot out there to supplement the workforce and make everything more efficient and easier to do, faster to market than solving problems for customers.

Megan - 6:56: And how long has this company been around?

Mark - 6:58: So company's been around for about nine years, so really six years of commercial success where we've deployed those 800-plus robots.

Megan - 7:07: And looking back over the last twenty-five years, what's been the biggest shift in how finance leaders use data to influence business decisions?

Mark - 7:16: I mean,

Megan - 7:16: it's pretty miraculous how far we've come in twenty-five

Mark - 7:20: years. As I think about my career, I don't think about myself being around in the workplace that long, but it's actually been longer than that. And what I've seen in my career change a couple of times, but really in the last ten years or so is the finance and the CFO role moving away from what I call a historian to really a copilot and a navigator, to go back to my airplane analogy again, has been the biggest shift in using data.

I think about some of my career when I was working at Wawa, managing a $5 or $6,000,000,000 budget. And that's a lot of discipline. That's a lot of work. That's a lot of backward looking. You close the books, you look at the report. And by the time you think about the answer that you have, you've already surpassed the decision.

And I think what I've learned from the beginning and what's transformed is learning to make better decisions with imperfect information and not having the right answer all the time. And so I think the shift to real-time data and utilizing that in the AI-driven world that we live in today and not just historical documentation is really the change.

I think I'll also mention a second shift that I've seen in my career is who owns the data, right? If you think about it, for most of those twenty-five years, it was really IT. Whether or not the CFO ran IT or had a hand in IT is a different story, but IT was the data owner and everything for data was IT's problem. Finance could have asked for something. We could have never got it.

And I think because now finance teams and CFOs in particular sit so close to the data, that allows us to not wait for things and actually have access to the real-time data immediately to run a business.

Megan - 9:22: And of course, AI is dominating almost every business conversation today. So from where you sit, what's genuinely different this time compared to previous waves of technology innovation? I mean, I can think back to the beginning of my career and that was like the start of the internet and seems like the trajectory of technology is just crazy right now.

Mark - 9:48: Just those two letters, AI, mean so many different things in the marketplace today and the world today. I think you mentioned internet. I think back to my own beginnings at Zany Brainy, and we were just trying to put a web store together when nobody even really knew what a web store was. Right?

When you think about internet, you think about ERPs, you think about big data, data analytics. All of those things are significant in the ecosystem of growth and all had the same type of hype in the time period that they existed that got well ahead of day-to-day reality of the situation.

Think about AI and really as a whole, but you think about Gen AI specifically. It's not just processing, but it's reasoning at the same time, right? If I go back and think about the couple of things I mentioned that you mentioned, right? Internet, big data, ERP, all of those things are about different ways to access and save and store information.

When you think about data analytics and some of the things that we were using with Looker back in the beginnings of my GoPuff days, it's really kind of visualization, but AI really helps. At least how we use it here, and I think most of the world is beginning to use it, and it helps with the interpretation.

It can see all of that and really for finance and accounting, it's a totally different leverage situation to take advantage of. That's one. I think the other thing I think about, just because I'm a finance person, is the cost.

If I think about I've gotten a chance to implement various ERP systems, implement data analytics when I was at GoPuff, you're talking about multimillion-dollar projects that span eighteen-plus months. If you think about AI today, I have two or three engineers that sit across from me here in the office, and in a couple of days, you can pilot something and put it into practice immediately at an enterprise level, which is just phenomenal to me today and how fast that moves.

Megan - 12:03: And it seems like there's enormous amount of pressure on executives to have an AI strategy. So do you think that pressure is causing companies to start with technology rather than what their actual business problem might be?

Mark - 12:18: I think that is likely the case. If you think about any of the AI-type companies that don't have real-world data in their toolbox, so to speak, I think that's one. I think companies want to jump on this AI wave sooner than they might be ready for and are not actually tackling a problem.

And so when you think about Burro, we're tackling a $2,000,000,000,000 labor problem that exists in the outdoor markets today, where labor is scarce and/or expensive in agricultural settings and industrial settings.

Sure. We can enable autonomous vehicle and the AI that it sees and feels in the world to enable that to solve that problem every day. But you have the problem to begin with that you're tackling.

I think everybody is focused on adopting whether it's Claude or OpenAI or ChatGPT, everybody's jumping to grab hold of something to say that they're utilizing it and not really doing it in a very effective way to solve a real-world problem that successful businesses, for that matter, need to be built on.

Megan - 13:29: And as we've mentioned at Burro, you're helping build autonomous robots that generate large amounts of operational data. How has working in AI-enabled business changed the questions that you're asking as a CFO?

Mark - 13:44: So for me, it's changed it, first of all, very quickly, but also a little dramatically, right? If I think about every other company I've worked for, whether it was Medical Guardian or Wawa or Zany Brainy, the financial questions are all the same. It's just as you could probably rattle them off. Revenue, margin, headcount, return on invested capital.

All that still matters for me here at Burro, but if I think about a robot running continuously in a real-life scenario every single day, generating terabytes and terabytes and terabytes of data, that operational data seen the same way as financial data is really the change for me.

And that leads me to very different directions in terms of questions and answers. And I think about something for me and us here as a company that's really impactful is fleet utilization. How often is the fleet being used? How many miles has it run? It's the input to the value prop for the customer.

And back to financials, right, revenue, cost, margin, unit economics. If the robots are running at 20 or 30% utilization as opposed to a 70 or 80% utilization, that has a huge implication on customer ROI and, really, our ability to demonstrate that we are the right solution for the business.

And so the shift for me is asking more and more questions about data. I think I've said to people, I am now an operationally minded CFO helping drive the business forward as opposed to a truly financially minded CFO that is just here to help make decisions. I think that really makes a difference.

The other thing is that I think about a lot is, as most CFOs do, is risk, but Burro's really changed the way I think about risk. If you think about Medical Guardian, right? Operational risk typically lags, right? You close the books, see the numbers, all of a sudden you're like, "Oh my gosh, how about that problem that we had thirty-five days ago?"

In real-time AI data in a physical AI robotics business, you're seeing it literally as it happens. So I can watch robots running in the world today, signaling issues, visualizing obstacles and build that into decisions immediately.

And so that's changed what I look at. It's also changed the question that I ask, which is what decision does this information change? It's less about what decision am I making and more about what decision would I've made that changes with the data and operational metrics that I see in real time today.

Megan - 16:39: And you mentioned being an operational CFO. For a CFO out there who's stepping maybe into a new industry, how do you go about understanding operations? What's one piece of advice you would give someone who's stepping into the role to really truly understand operations?

Mark - 17:00: So what I've seen in my career with my peers and what I've seen just generally in myself is that a curiosity about the business. Number one, you have to have genuine curiosity about the business you're going into.

Sure. The financial representation is great, but the best peers that I've worked with and mentors that I've worked with, don't just understand the numbers, right? You need to understand why customers buy, what makes operations tick, what the team is worried about that doesn't show itself on a spreadsheet or a dashboard.

At GoPuff, I went and spent time in the warehouse understanding the fulfillment operations. At Burro, it means sitting with a robot in a 101-degree field in Peru, understanding what people are going through and what that data is really telling us.

The curiosity is so important because it allows you to ask the right question. And when you think about asking the right question, that's how you get a useful answer for all the data that you can see.

So my advice would be take that genuine curiosity that exists and embed yourself in the business model from top to bottom. That will help you be a much better CFO, not only operationally, but financially.

Megan - 18:29: And along those same lines, you've led finance through periods of hyper-growth, acquisitions, and transformation. How has data helped you to make better decisions during times when there wasn't necessarily a clear playbook?

Mark - 18:42: So I'll take GoPuff as an example, and I'll go back to something I mentioned before, is the data. Utilizing the data to give you confidence when there's periods of uncertainty is what has helped me, and I think helps everybody make better decisions.

So growing at 25 times valuation, there's no playbook for that. There's very few companies that have been able to do that in a short period of time. And in that hyper-growth scenario, you need to spend more time making, I wanna call them gut decisions, but it is disciplined decisions with the data that exists, as opposed to relying on the past.

When you're looking at an acquisition and acquiring, doing a $200,000,000 healthcare tech acquisition, it's less about what worked in the past and more about what will work in the future.

And so to do that, my playbook had been not to rely on sophisticated modeling, really rely on modeling, but not sophisticated modeling, but really sit on the data and make those future scenarios credible rather than speculative.

And I think that's really what's helped me through those periods of transformation and acquisition, and clearly hyper-growth at GoPuff and what I'm using here at Burro to take us through the scaling phase.

Megan - 20:08: And as a CFO, when you're evaluating AI investments, how do you distinguish between projects that create real business value versus those that just simply sound cool or innovative?

Mark - 20:21: By the way, there's a lot of things that sound cool. I'm gonna go back to the statement I made before, and then this is really what I live by, is what decision does this change? And I put everything through that lens.

And if my teams can't answer that clearly, if they give me what I perceive to be a fluffy answer, like, "Well, it's going to make us more efficient," or "it alleviates the ability to hire today," or "we become more capable," I don't really do well with those types of answers and I keep pushing.

What I want somebody to say is today, we do this. It takes this long, and this is the decision we make. And tomorrow, with this particular investment, we will be able to do this, this, and this that allow us to make a different decision that is better for the company.

I think that is the most important thing and that's the lens that I force everybody here to go through. They don't like me at times, but they know that it's, we are all making the business decision that grows the company faster before we lose the ability to get through that window.

In conjunction with that, I think accountability is also something that I focus on. AI in particular tends to not have responsibility and accountability. I don't know why. I haven't put my finger on why yet other than it's cool, it's innovative. Everybody wants to be part of that initiative, but I don't think there's a lot of accountability and ownership responsibility there.

So for me, everything needs to have a named owner and a measurable target and a smart way to spend that money so that we can focus on achieving the goals that we set out to.

But those two things, I think, in conjunction with each other is really how I look at that, alleviating the fluff and the cool that comes out of evaluating AI investments.

Megan - 22:28: And one challenge that CFOs face is getting the organization to trust data-driven decisions. How do you build that culture without losing the value of experience and intuition?

Mark - 22:42: Yeah. I think that's an interesting question because I think that I use intuition a lot more than I ever thought I would in my roles that I've had.

I think intuition is all around compressing all the experience that I've had or that anybody's had and utilizing that to get something stronger than not utilizing the intuition.

So I'd say what I tell the organization is use that intuition. Like, don't discount it because when people say that something doesn't feel right, I actually think that there's, that's not just an irrational statement. It's really something that they've learned over years and years of experience.

So I think it needs to start with acknowledging that intuition needs to be used. So I always do that first.

After that, and when I think about trusting and allowing the information, the organization to trust data-driven decisions around a cultural aspect is talking to the organization about what we call intellectual honesty, because the data is either gonna confirm or deny the intuition.

And so you have to acknowledge the intuition and then utilize the data to prove it right or wrong. And I think what we've built here from a cultural aspect is that we don't use the data to justify conclusions that everybody wants, whether it's founder, whether it's me, whether it's our COO.

I think I've seen a lot of failures in my past life at using data to justify conclusions. I think it's really the ability to change decisions when the evidence shows it.

By the way, in the forefront at an all-hands meeting, recognizing that just as bad as everybody else is at making decisions. But if you do that over and over again, then people really buy into the data-driven decisions because you can acknowledge that their intuition is real, utilize the data to prove it or disprove it, and then show real-life examples of that working and not working.

That gets the full buy-in of the organization. That's what I've been successful with through my career.

Megan - 25:04: Yeah. That's great advice. So looking ahead, how do you see the role of the CFO evolving in the next three to five years and what new skills or mindsets will be critical for the CFO of tomorrow?

Mark - 25:16: So, you know, I mentioned earlier about not looking backwards and looking forwards and being a pilot, so to speak.

At the end of the day, the scorekeeper, if you wanna call it that, like, that's... That role's not going away. It's never gonna completely go away. Everybody's gonna need to close the books. Everybody's gonna need to run reports and read reports and talk to the board of directors.

But I think this, even more so in the last couple of years, this complete shift to forward-looking data-driven decision-making, even more so than when CFO roles changed to come from controllership to coming from FP&A. That was kind of the first wave of this.

And I think now, so if I think about where my time goes, it's still a lot of it goes to reporting and variance analysis and some compliance stuff, but that's clearly going to be automated. If not in the next five years, certainly before the next ten.

And so if I think about the next ten, it is that pilot or copilot next to the board and the CEO to think about making complex decisions for the business with AI-enabled data.

Somebody has to translate the complex and make it clear. And I think everything I've learned from Zany Brainy to eBay Enterprise, to GoPuff, to Burro has really translated, allowed me to make the translation of that to the organization.

And that's gonna be the norm in a go-forward five to ten years. The technology is the CFO's problem and issue today. And getting at the data with the AI that exists and will exist is really gonna give this role an advantage in the future.

Megan - 27:13: Mark, thank you so much for taking the time to be here today to share your experience and knowledge. This has been great.

Mark - 27:19: Thank you very much for having me. It was a pleasure. Really, really enjoyed it.

Megan - 27:22: And to all of our listeners, please tune in next week. And until then, take care.