Why AI Can Create a False Sense of Precision Without Human Review

[00:00:46] Megan: Let’s jump right in. Welcome back to CFO Weekly. Today I’m joined by James Vanreusel, founder and CEO of Vanreusel Ventures, where he and his team help emerging companies build the financial systems teams and decision making discipline they need to scale. James brings a career spanning Wall street, private equity, asset management, microfinance and CFO advisory, giving him a broad perspective on what actually creates durable business value. In this episode, we’ll explore where AI is genuinely improving forecasting, reporting and analysis, where the hype gets ahead of reality, and why fundamentals like pricing, margins, liquidity and unit economics still matter more than many founders want to admit. James, thank you so much for taking the time to be on the show with me today.

[00:01:38] James: Oh, it’s my pleasure.

[00:01:40] Megan: Yeah, I’m really looking forward to this topic. I think it’s super interesting. So as you look back on your early days on Wall street, was there a moment when you realized that sophisticated models and data can only take you so far if the underlying economics of the business are not strong?

[00:01:59] James: I think what I’ve learned over the, I guess it’s a few decades now is really understanding how business works. Back then in my early career, we learned a lot about investing, trading stocks, how IPOs work, how capital markets work. But we would rely relatively heavily on sell side research and analysts to really do the digging, listen to all the calls, et cetera. Now I’ve been on the other side of the table quite a bit and have conducted the calls that the analysts listen to. And so I have a much bigger, better understanding of how everything works and how I would say, you know, stock prices work when they go up, when they go down, why they go up, and really what the companies need to show from a fundamental standpoint that’s really going to move the valuation really of the stock and of the company.

[00:02:56] Megan: And before we go any further, just kind of walk us through your career where you started and where you are today.

[00:03:03] James: So originally I’m from Europe and I came over to the US to go to business school and started working on Wall street right out of business school. So I was doing institutional equity sales for about eight years and after that I really wanted to get into something that was more socially oriented, so social finance oriented. And I went into microfinance and started working with an individual who had just launched a few years prior a microfinance bank in Samo. And so for the next five years as that organization’s full time cfo, I was hoping to launch and operate banks in the South Pacific on different islands. Then after that I moved from New York to San Francisco and started what I’m Doing now, which is a corporate finance consultancy. We do a lot of fractional CFO work for both VC backed companies as well as international nonprofits, including some M and A work, transaction work and due diligence work as well. And so that’s been about 12 years now. The company has grown and as well as the services and the different industries and companies we work with.

[00:04:19] Megan: Wow, sounds like an amazing career. Congratulations.

[00:04:22] James: Thank you.

[00:04:23] Megan: And you’ve worked at both ends of the spectrum from sophisticated financial markets to early stage companies still building basic finance infrastructure. So what has that taught you about the difference between financial sophistication and financial discipline?

[00:04:40] James: I think when you look at early stage companies and I’ll just go the venture capital route, if you look at early stage startups, pre seed seed, they’re very unsophisticated, they’re really focused on product and sales. Develop the product, sell the product, get cash in to grow. And it’s really once you get to a Series A and you’re raising 10 million plus dollars that you’re being asked to present us GAAP financials and models to be able to raise that level of money. When you get to a Series B, you start developing a board, you might need to get an audit done as well. And so the sophistication increases as the, I would say sophistication and size of investors increases as the risk goes up and the governance that investors need to have also increases. And so if you don’t have that sophistication early on, it’ll come back to hurt you because if you invest early on to get the sophistication, it’ll be a lot cheaper than doing it after the fact. And so you can see as you go through to the IPO stage, which gets you to capital markets, that at that point you should be very Sophisticated, with some very smart financial people on your board. You’re probably, probably, you know, at the IPO level looking at getting audited by the big four and paying a lot of money for that because you want that brand name behind you when you get valued, that you go public.

[00:06:18] Megan: And AI has become almost impossible to avoid in finance conversations. So from what you’re seeing with founders and finance teams, where is it genuinely creating better outcomes?

[00:06:32] James: I would say for us, we are intently using it. We’re really focused on finding ways to do higher quality work faster. And I would say that the higher quality work is really the main focus. I would say you start off in different areas. I think modeling has become a lot faster. It might have taken you 15 to 20 hours to build a good sophisticated model. Now you could potentially do it in three to five hours. I think the accounting close as well can be automated. And instead of 10 day or 15 day close, you can get to a two to three day close. It does take quite a lot of work and organization and reviewing your step by step workflows, that gets you there. But AI, I think, is instrumental to everything you do. It doesn’t replace the people because you still need somebody who knows what’s going on, who can look at the output and say, yes, this is correct or no, there are some serious errors in here. And so you still need an expert, but it can go a lot faster, a lot more automated. And I find that the quality of the output can be a lot higher as well.

[00:07:54] Megan: And in my conversations these days, it seems like forecasting is one area where AI gets a lot of attention. Where have you seen it actually improve the quality or speed of forecasting? And where can it create a false sense of precision?

[00:08:10] James: With forecasting, I think having lots of data, so big data can be really helpful. It can run different time series analyses, just different ways of looking at the numbers. It can run lots of different scenarios as well, very quickly. And I think that’s very interesting. Where it can go wrong is that there is so much uncertainty out there in the world. And I think it takes somebody who has a lot of experience to understand what is important and what isn’t important. So, for example, you might have created projections and then you run a prompt saying, what are all the things that could go wrong or what are the risks? And it might give you 25 different ones, but probably 80% of them are just not relevant. And so you do need somebody with experience and a keen eye to be able to evaluate where to look and what specifically for the company you’re doing this for makes sense.

[00:09:14] Megan: And you’ve touched on reporting and analysis, which both be automated much faster. But does that actually lead to better decisions, or are companies sometimes just producing more information without improving judgment?

[00:09:28] James: The fact that we can close the books faster can lead to better decision making because you’re not waiting two weeks to get the numbers. So I think from that perspective, definitely helpful. One thing that I do see is that AI produces too much, and knowing how to cut it down is, I think, just as important as having it help you find answers or information. What I found is if I spend too much time using AI, I seem to get frustrated because you end up going in circles.

[00:10:05] Megan: And when a founder asks you whether they should invest in a new AI finance tool, how do you help them decide whether it solves a real business problem or it solves simply the latest shiny object?

[00:10:17] James: I would say there are some very good tools out there. I think that there’s a distinction to be made between a software that’s been around for, say, a decade and is implementing AI now that makes it a lot more powerful versus what’s called an AI native software, where you really start with the AI and then build the solution around it. The ones that we’re most used to is the older software that seems lots of improvements, and because they’ve already worked in the past, they’ll work even better now. The new ones I find need to be tested quite rigorously because sometimes the output just isn’t what you want it to be. It might be slightly more expensive as well. It still needs that person in there to look at the output and sometimes also, you know, get the right input in. It’s not quite there. It’s getting there. And I think that the biggest benefit really is just the ability to do so much more with the same number of people. That’s my biggest benefit that I see there.

[00:11:26] Megan: You ever see a day where Excel will be irrelevant, or do you think that will always be a tool in an accountant’s tool chest?

[00:11:35] James: I don’t know. I saw one AI native tool that wasn’t using Excel anymore, and it was very. The inputs and the outputs that were being produced looked very strange. So I think there’d have to be a re education around that. Everybody is so used to looking at Excel that it might take some time. But I would add one other thing as well that I’ve said many times, which is I have always found that the best entrepreneurs or the best founders who are successful typically are strong in finance. So they might already have a cpa, or they might already have something else. And I think that AI really benefits them as well because they can get the information and the data from AI a lot faster and in more detail and be able to figure out themselves what decisions they have to make. So AI is very helpful when you know and understand the output very well

[00:12:37] Megan: and you spend a lot of time helping companies to build strong financial foundations. Why do pricing, gross margin, cash conversion, and unit economics still matter more than many founders realize? Even in this new AI heavy environment,

[00:12:52] James: it really matters because it always comes down to the return on investment. So what do I get for every dollar invested? Whether you’re looking at buying stocks or advertising or anything else, it’s if I put $100 in, can I double it or triple it? And so a lot of times AI is a multiplier when it goes right, that’s really what you want to invest in.

[00:13:20] Megan: And can you share a situation where a company was focused on growth or new technology, but the real issue turned out to be something much more basic in the underlying economics?

[00:13:33] James: I’ve started working with companies that had some level of maturity. They’d been around for six, seven, eight years, their income statement become very bloated, meaning that they kept losing money and they kept having to raise money. And at some point, investors don’t want to invest money anymore in you if there’s no end in sight or the numbers don’t keep improving. Sometimes there are big opportunities to come in and really cut expenses by a significant amount without really hurting the organization. And it comes back to this ROI calculation of If I put $100 into you, what do I get back? And sometimes when you become slightly more mature, investors want to see some path to break even or profitability. And you can switch investors as well. So VCs are usually into hypergrowth, whereas PE firms are really looking at EBITDA growth. And EBITDA has to do with not just revenue, but expenses as well. And so you can get to a point where you say, okay, I need more money than a VC’s able to give me. Now I need to go the private equity route because I’m close to ipo, so what does that look like? And how do I switch over the type of investor that can give me enough money to get to ipo? Or you might be saying, I’ve grown a lot, now I need to do some kind of M and A transaction, sell the company. What do I need to start looking at? And then you start looking at EBITDA growth instead of revenue.

[00:15:13] Megan: Growth.

[00:15:14] James: So depending on the timing of the organization, that’ll switch as well. So it’s not only revenue growth. Sometimes once you get to the spending, you need to look at unit economics and benchmark it as well in the industry and say, okay, the cost to produce a widget, $5 doesn’t work, it needs to be $2. How do I get there? That’s another. These are all different challenges.

[00:15:41] Megan: So when you think about bloated income statements or a company that needs to cut expenses, what’s usually like some of the low hanging fruit? Where should first, a good way to

[00:15:53] James: look at it sometimes is just to say, okay, what’s the percent of revenue that I’m spending? Right. So on the SGA side you might just say I’m at 28% of revenues, which is high. The industry is at 22. I want to get to 20. How do I get there? You’re really setting targets and finding new ways to get there. A lot of it is innovation. And so when you do the annual budgeting and you’re looking at your department heads, you want to be able to say, here’s your budget and it might be 20% less than they had asked for. Okay, we’ll figure out how to do the same amount with 20% less. So you want to challenge people to innovate as well, to do more with less money. I think other areas to look at is really. You talked about gross margin before. So if you look at direct margin and gross margin, how do I maximize those? If I have gross margin at 33% and the industries at 38, how do I become more efficient in how I do things, in how I spend my money, in how I hire people? Do I hire more contractors versus W2s? Do I hire internationally? There’s lots of different ways. And figuring out how to be efficient in the money you spend is critical.

[00:17:21] Megan: And as AI spreads across finance and accounting, where do you think that human judgment becomes even More important? For CFOs and founders, it’s having that

[00:17:31] James: reality check and the critical thinking that’s needed and also the ability to look ahead. So what I think VCs have a big advantage of is they see so many different companies that are trying to get ahead and look ahead. And so they see several years into the future as to what people are trying to make and sell that’s just not available right now on Wall street as well, you get so much information that part of the skill is to be able to see what’s important and ignore the rest. You might be looking at 60 different industries, you see these trends appear. That’s a very important skill to have. You also just need to have the information flow to be able to get there.

[00:18:20] Megan: And speaking of looking ahead, but this is the last question, how do you see the role of the CFO evolving in the next three to five years and what new skills or mindsets are going to be critical for the CFO of tomorrow?

[00:18:34] James: The ability to use AI, I think it’s going to become an even more strategic role. You’re still going to be building teams and I think the more experience that you can get working with different companies I think is hugely valuable. I continue to see the CFO role as being very strategic. I like the fractional CFO model because I think smaller companies and startups only need a certain amount of time because it’s strategic and you really just need a controller or an FPA person to do the majority of the work and then you need the CFO to help you scale. But that’s really just a small percentage of the time that’s needed. And it’s only really as you get larger and larger that you’re doing more deals and transactions and negotiations and you’re really thinking alongside the CEO as to what’s the next step and which company do we need to buy and things like that. So I think it’s going to continue to be a very high level role and it needs to continue to be cross functional to be successful.

[00:19:47] Megan: James, thank you so much for taking the time to be here with us today and I really appreciate you sharing all of your experience and knowledge.

[00:19:54] James: It was a pleasure talking to you and I always like to share my

[00:19:59] Megan: latest thinking and to all of our listeners, please tune in next week and until then, take care.