When AI Changes the EPM Decision | with David Den Boer, CEO, Column5 Consulting

0:00Support for CFO Thought Leader comes from Salesforce. Unify selling and billing for a seamless customer experience. And OneStream, trusted data, faster decisions. Bonus episode. It's needed. And so you need this comprehensive understanding of what each technology component brings to the table in terms of potential. because I can tell you no individual component does it all and won't for any time soon. AI won't do it on its own. The LLMs, you can't say run my finance operations. And it does it without hallucination. That's not a thing. The data platforms can't do it. The BI tools don't do it. And the EPM tools come closest, but still don't do it.

0:55So a CFO has to be very educated and discerning on what these differences are. And that's part of what the event helps expose that conversation. Because everyone's doing it in a slightly different way. It's evolving in real time. And we have customers saying, I can vibe code this whole thing. We have other ones saying, I'm never going to change. So somewhere in the middle is the answer. And AI gives you a huge opportunity to experiment with these things and find your sweet spot and your formula that's right for your organization. Hi, it's Jack. On today's show, we speak with David Denboer, CEO of Column 5 Consulting and host of the upcoming EPM Summit in Las Vegas.

1:57Finance leaders are being told that AI is about to change everything. But the harder question may be, change what exactly? Because suddenly the lines between EPM, business intelligence, data platforms, and AI are getting harder to see. And CFOs are being asked to make technology decisions while the landscape itself is still moving. David Den Boer believes part of the answer is getting these competing ideas and competing vendors into the same room. That's the thinking behind the EPM summit he's bringing to the Bellagio in Las Vegas, November 16th through the 19th. And it raises an intriguing question.

2:41What happens when experimentation itself becomes part of the strategy? Our talk with David Den Boer begins after this.

3:03Hello, David Denboer is with us. David is the founder and chairman of Column 5 Consulting and Darwin Adelenics, and someone who has spent more than two decades working at the intersection of finance, technology, and enterprise performance management. When David joined us previously, we explored the thinking behind an EPM summit and his belief that finance leaders need better ways to evaluate technology before making critical decisions. Now, as the EPM summit takes shape and prepares to land November 16th through 19th in Las Vegas, we're back with David. David, today we'll be going deeper into AI, the changing EPM landscape, and what finance leaders need to be asking as they rethink their technology and processes.

3:53David, welcome back. Good to see you, Jack. Thanks for having me. So, David, what's changed? It's only been, you know, not even a year, but what's changed most in the EPM conversation since we last spoke? What's transpired? Everything's moving rather rapidly. Oh, very rapidly. I think since we last spoke, AI was something to be experimented with and looked at and considered. And now I think it's mandatory. Even more so, we're down in the weeds talking about incorporating agentic processes that not only help individuals work, but even replace tasks that were performed by individuals just really a few short months ago.

4:39And this is leading to unprecedented automation and improvements in cycle times for processes like the close, planning, analytics. Across the board, it's making real contributions. And every company really needs to understand how this can improve their operations. Now, you've warned against simply bolting AI onto existing finance processes. where are you seeing that happen really well i that's a great question i see uh everyone by now is probably familiar with claude chat gpt co-pilot and you know that you can have an existing finance process of which the output is maybe an excel report you can take that excel report you can hand it off to your llm of choice and you can ask what do you see and i'll come back with some surprisingly insightful information.

5:40It could, on the analytic level, it might find some dots to be connected in a way that you hadn't considered. From a technical standpoint, it might find structural improvements that could be made with the report to make it perform better, be more reliable or intuitive. And that to me is like just adding it, leaving your current process in situ and layering AI on top of it. Vendors are offering these hooks to make this even easier. They're called MCPs, model context protocols, and it allows for you to take an external LLM and connect it directly to your EPM database to make this even easier on the one hand.

6:25But it is this layering effect where you have your in-place EPM and some external bolted-on AI that augments individual processes. So one of the things we were talking about last time, everyone says that their EPM platform now has AI or is AI empowered. How should a CFO, though, separate real capability from marketing? Wow, these are loaded questions for sure. Well, I think, yes. What questions will expose the difference right away? What do you think? Well, having worked with some analysts recently on this topic, I think that there are two main thrusts that are happening right now. One is internal to the EPM tool.

7:28The vendors are actually quite innovative, building in tools that will help you close faster with higher quality, self-heal, making something like real-time close a real possibility so that anytime you run your report, you have closed financials available for you. Each CFO can decide what that value would be to their company to do away with the latency at the end of the month where you have to wait a week to get to those closed financials and then begin your analysis to explain those. What if you could do that incrementally throughout the month, catch small issues that may be mitigated and not have a surprise at the end of the month.

8:14Seems valuable, but all your processes that consume that data would have to catch up and adapt to data being available in real time. So what we're seeing is the tools that various vendors are embedding more and more function to do away with latency, to increase quality, to increase autonomy, and that's very valuable. However, the defect is that the EPM tool is only one part of your analytics landscape. You have operational planning tools. You have BI, an enterprise data warehouse. You have other systems in play that affect the outcome, external data as well. And not all of this flows through your EPM system.

9:04statistics say that despite the EPM systems being so great, over 50 % of the models remain managed by Excel. So if they were so great, they should have removed Excel. So there's something else is needed. And I think the main thing is orchestration, something that can apply AI to truly connect different systems. So while each vendor is working to make the area of management for their system run at unprecedented speed with much higher quality, all a huge benefit. What's needed is another layer on top of that for orchestration to tie those together, give the CFO a bird's eye view of what's happening in the organization and maybe why does one system think scenario A is like the outcome and a second system thinks something else is going to happen.

10:04It's those gaps that I need to reconcile and understand what are the assumptions driving this output and find what's missing, deal with that, and maybe uncover the real value. Now, you're fond of saying that technology selection can determine a project's outcome before even the implementation begins. So what are companies still getting wrong? Well, I think one key thing, and I don't think this is controversial, is that you have two main technology stacks that you have to make a decision for. Transactions, ERP, and companies don't all have all their transactions in one ERP. That's a fragmented reality.

10:55You have other transactions in operational systems as well, also fragmented. And then you're being asked to make a bet on analytics. So between transactions and analytics, these are key stacks that you have to choose. I don't think we've got there yet on the analytics side. It's still being looked at point by point by point. What do I want for a data warehouse? What do I want for my BI tool? What do I want for EPM? What do I want for operational planning? It's not viewed as a holistic ecosystem yet. And I think if I were to equip a CFO to ask better questions is how does the EPM tool play in a constructive way in a complex landscape where I have multiple systems?

11:46And their typical answer is move all your data into my platform. But in reality, no customers do that. So we need a better answer there. Orchestration, again, is the most likely answer. We've got the large language model companies coming forward and sharing a lot about finance and how they can serve finance professionals. You've got the data platform companies coming forward saying, no, no, we've got what you need. What's happening? How is someone to choose a partner in this kind of environment? that's it exactly the it's confusing and the answer is not going to be the same for every company uh i think like i mentioned there's these two stacks that you have to do the transactional one which has really been decided over the last 20 30 years you have to pick an erp but the analytics one is really starting to come into focus now where i think it's coming down to the CFO primarily as a driving force on what are the priorities, what are the requirements that we have, and maybe even what is the right stack.

12:56So data platforms in particular like Fabric, like Databricks, like Snowflake and others, and even SAP BDC are becoming more full-featured, where they're launching agentic initiatives that can clearly go into the office of finance. So that's one signal that you're getting. BI, Power BI, click other tools, allow right back. So that's uploading data. Well, what data might you upload? Plans, consolidations, maybe journal entries. So now that's another interesting angle that was previously exclusively EPM's domain. So the question is, what is uniquely EPM's domain that I need an EPM solution for? And how far can I get by somewhat cobbling together or use AI to vibe code or bridge the gap between the data platform and the front ends, including BI and EPM, Excel, and merge those so that I'm using these flexible tools to build a data set?

14:09My personal opinion is all of the above is needed. And so you need this comprehensive understanding of what each technology component brings to the table in terms of potential. Because I can tell you no individual component does it all and won't for any time soon. AI won't do it on its own. The LLMs, you can't say run my finance operations. And it does it without hallucination. that's not a thing. The data platforms can't do it. The BI tools don't do it. And the EPM tools come closest, but still don't do it. So a CFO has to be very educated and discerning on what these differences are. And that's part of what the event helps expose that conversation.

15:02Because everyone's doing it in a slightly different way. It's evolving in real time. and we have customers saying, I can vibe code this whole thing. We have other ones saying, I'm never going to change. So somewhere in the middle is the answer. And AI gives you a huge opportunity to experiment with these things and find your sweet spot and your formula that's right for your organization. One of the interesting aspects of the EPM Summit, as I understood the concept, you're going to be having competing vendors in the same room. So it's not like one vendor show. This is going to be, they're going to be side by side.

15:46What do you hope happens when they're sort of challenged side by side like that? Well, and there's definitely a lot of concern coming from the vendors. They're asking the same thing. Is it going to be a Jerry Springer episode or is it going to be a constructive conversation? We won't let it be Jerry Springer. But I think what's important for customers is just what you said. All the vendors sometimes say the same things, and the devil's in the details. What do you truly do that's different? If we have vendors in a panel and we're asking them live, side-by-side, questions, it wouldn't make much sense for them to say, yeah, we do the same thing, we do the same thing, we do the same thing.

16:29That's not interesting. they will automatically discuss their unique innovative ways of approaching these challenges. They'll focus on the differentiation, which I think is really hard for someone evaluating software to get to because they send out an RFP and the RFPs usually say, yes, yes, yes, yes, yes. We can do all of it. And how do you decide? Well, let's ask them live. Let's put them next to each other and ask them to differentiate their platform. Where are they more successful with customers? Where do they add unique value? And I think it'll be informative for the audience. For CFOs or FP &A leaders who are interested in attending the summit, what are some of the concerns that you expect to be top of mind there?

17:28I mean, and get addressed there in some way or stir conversation. What are they looking for answers? Well, I think step one, you know, we're all talking about AI these days. And one of the weaknesses of AI is bias. Meaning if you ask your question in a narrow way and only look in a certain place, It shapes the answer that you get. So if you're looking to evaluate vendors and you don't contact all the vendors, you automatically miss the opportunity to consider what they have as an offer. So I think anybody attending the event is going to look at our roster of exhibitors and make an effort to understand, are these worth considering?

18:21are these worth considering? Rather than relying on word of mouth or what a vendor tells them, which is obviously biased and agenda-driven, here, we don't have that agenda. We want to give them as much information as possible. On the vendor side, but even on the consultant side, the consultants that are there are leaders in North America. They have been implementing across multiple products, they'll have a point of view as well to say, hey, I've been successful with these three tools in this area. They may go to the next exhibit, talk to another consultant, and they may have a different recommendation.

19:03It might sound a little confusing, but opening the aperture of what you consider is one way to find the best fit for you. And I think customers are at an extreme disadvantage in this market. It's very difficult to get information. It's difficult to, you know, none of the vendors suggested, hey, Darwin, why don't you put together a summit and bring my competitors? That would be a great idea and be a great value to the customers. I know it will be. Customers, I hope, know it will be. Vendors took some convincing, to be honest, but they would prefer to have a show that only features their product.

19:47That's the ultimate form of bias. You know, what was your question? My solution is the answer. It's, we want to get past that and bring the power of the conversation, at least at an equitable level, maybe tilting towards the customer having the most leverage, more than they could get with an RFP, more than they can get at other events. And the output here is to convert attendees with lots of questions to very equipped buyers that will challenge the vendors and ask that depth of question that gets to the real value for them. These products, they have to live with for a decade or more. They're very sticky.

20:33They're tough to replace. They're expensive to switch. And they deliver a lot of value. Making a good decision here carries benefits every year, every month even. So it's important to get it right. We're frequently asking finance leaders, what roles or processes should we expect to see change most in the coming years? If AI really changes the economics of finance, what do you think? What does the finance organization look like? What processes perhaps are no longer mission critical? Well, I think that they all are mission critical, but the way they're conducted should change radically. We look at two types of innovation that AI enables.

21:23One we call operational agility, and that is automating what you're doing today. So there's a lot of very manually intensive processes still out in the world. We have customers using Excel that's almost always manually populated, refreshed with new actuals, adapted to additional accounts or customers or whatever, and making these lists and moving the data from one system to the next and from one workbook to the next is a tedious process that any finance person is well familiar with. But that should be done away with, that should be largely automated via AI. On the other side of agility, we have something we call innovation agility, and that is the ability to experiment.

22:15I mentioned how sticky these tools are, and when you adapt a new tool and then you find out that, you know, something new has been released and I want to try it, that was usually cost prohibitive. And you couldn't really experiment. You certainly couldn't roll back with ease, and you couldn't evaluate multiple solutions. I think AI is going to lower the cost of these solutions. It's certainly going to create more options where you can adapt more utilities and functions that will add value to finance. and AI will give you the freedom to try more of those and not make a decade-long commitment that, you know, this, yeah, the price takes$100 ,000, but the implementation costs and support are$10 million.

23:06So I need to proceed with caution. I think customers will be able to say, you know what, I'll take a shot. Let me try it for 90 days. And after that, if it works, I'll incorporate it in my standard process. And if it doesn't, I'll try something else. And that ability to experiment and to innovate is going to change where people spend their time. I think that instead of operating this manually intensive system, they're going to be out on the other end, experimenting, looking at new data sources, looking at new utilities, looking at new ways to wring more value and advantage out of your financial analytics.

23:50We've been speaking to David Denboer, an EPM thought leader and advisor, and he's bringing the EPM Summit to Las Vegas, November 16th through the 19th. David, when people leave the EPM Summit, what do you hope they think differently about?

24:13Well, I want them to be inspired. I want them, even if they've made their tool selection, but they're looking at some of these fantastic innovations and visions for the future that other vendors may have, that they're inspired to experiment with some of those ideas, or it elevates their expectations for where AI can go. I think the main reason companies are not trying AI and not experimenting with innovations is they just don't know what's available. We're asked often for a glimpse at the art of the possible. Well, just show me what's available, and I'll choose from that menu. You know, yes, this makes sense.

24:59This one doesn't resonate with me. These other two are good. And they want that broadened aperture, like I said, an understanding of where are these agentic processes being enabled? Where are they driving value? Who's using them successfully? What does it take? That understanding to accomplish that in four days is very compressed. And I feel like many companies are kind of frozen. They have their own job. They've got to do their day-to-day finance, often with manual and tedious systems. And they're also supposed to be evaluating these tools and, you know, understanding automatically everything AI can do.

25:39I wanted to create a resource for them where they can come in a compressed and efficient time frame, understand all of that, and come back as a tough negotiator for the vendors and their architects and internally to ask good questions of their IT. Well, have you heard about this? Can we get this going? challenge everybody in their own best interest to really get to a state-of-the-art capability. It's easier than ever. The bar has been lowered with AI that some amazing things are within their grasp. They just need to know what's available. David Denboer, thank you for joining us on CFO Thought Call.

26:23Thank you, Jack.

26:34Hello, everyone. Jack here. Hope you've been enjoying our recent episodes of CFO Thought Leader. Please know that I don't take for granted this opportunity to serve you. Your time is valuable. And of course, so is the time of our CFO guests. I'll admit to you on certain weeks, I will pinch myself and say, hey, I get to do this for a living. I speak to all these impressive leaders. Undoubtedly, there are more aggressive interviewers out there, some who leave no stone unturned. But our swim lane is for finance leaders only. And those finance leaders have a unique style of leadership, one that is rooted in the burden of trust.

27:21In light of this, finance leaders must take great care not to forfeit this trust as they step into the role of being a more expressive leader. And that is what CFO Thought Leader is really about. That's why I try to remind myself, Week in, week out, how fortunate I am. Still, I need your help. Tell us what we're doing right. Tell us what we're doing wrong. Email me your comments at jack at cfothoughtleader.com. And please visit cfothoughtleader.com. You'll find a listener survey right on the nav bar. As always, thank you for listening.