Justin [0:00:00]: If I'm a CFO, and I'm reviewing business cases or deliverables, I think we're now at the place where we can expect that AI is pervasive in basically anyone's analysis. Too accessible, too easy and too good at what it does to not assume that people are using AI to arrive at their assumptions. And if I'm a CFO, I'm now asking people in these different business cases: what's your intuition versus what's the AI saying? And I wanna see and understand where you as my trusted colleague in this organization are using your brain power to interpret what an AI synthesis would be and layering in your unique and softer nuanced synthesis on top. So I think that's gonna be increasingly included and expected as AI adoption reaches its apex. The value of human intuition incorporated within a recommendation is gonna grow in importance, and it's going to be isolated separately from the AI synthesis in those analyses.
Michael [0:01:11]: Welcome gentlemen. We have Russell Lester — I'm gonna put the spotlight on you for a little bit. Our CFO, you know, I'm sure at some point in your career, you've been called the CF-No. True or false.
Russell [0:01:24]: Guilty as charged. Absolutely.
Michael [0:01:26]: Okay. Cool. I'm not sure if you embrace that or not. I guess my question for you is, do you feel like in the past — I'm not talking about your current role because we're gonna dive into your current role — but do you feel like in the past, you've had to adopt that mentality or that title to some degree, just out of your volume of requests or because of the stupidity of some requests? I've never been a CFO, so I can only assume.
Justin [0:01:50]: It is interesting, and I think it differs based on how you ascend to the role of CFO. I came up through the ranks of analytics and FP&A. So we were naturally building partnerships and seats at the table with our functional counterparts. And so it was an interesting transition because I would go from being the analyst or leader that was advocating for that group to help them make the case for an investment to the CFO, to then being the one that was having to make the judgment call. And that transition — you have to be willing to push back more without feeling like you're breaking those bonds, breaking that partnership. You need to create that sense of respect that you're not just going to rubber stamp every single thing that comes across your desk. But it's not an instantaneous evolution. Maybe for some CFOs, they naturally are the type that say no to everything. But I've always... you know, it's funny growing up in leadership. I always found the people that thought that being the powerful naysayer on the call was the power move — that never impressed me because I thought it was the craziest and the easiest move in the room, because all you have to do is just disagree with what everybody's saying. And I saw a lot of CFOs do that, and it was rarely well informed. They were just like, oh, I am a robot, and my role is to say no and be grumpy about everything. So I've tried to finesse it — probably imperfect — and I've indexed on one side or the other at any given time. Sometimes I'm too nice as a CFO, other times maybe out of nowhere people feel like I'm being grumpy because I'm coming in hot off of some investor or board call. So I think it really varies.
Michael [0:03:46]: And that's where I wanna dig in today, because ideally as a CFO, you know when to push back and when you don't need to push back, so you don't need to either adopt the either/or of "hey, I'm having a good day and feeling charitable so I'm gonna approve" or "I'm having a bad day so I'm just gonna push back." I don't know exact stats, but I've heard numbers that for every dollar of budget that's approved, you have multiples of that requested. So to some degree you have to push back. Where I wanna tie this episode is — Justin and I were at a dinner last night with a bunch of heads of procurement. And one thing that we were talking about, and it's landing really well in the market right now, is that part of the reason Tropic is really helping us be so successful as procurement leaders is because it helps us know where to focus. We cannot anymore cover literally everything. Everything should flow through the process, but we can't spend our time on every deal. Some people default to "if it's over a certain dollar amount, that's where I'm gonna spend my time." And I can speak from experience knowing that sometimes I can spend a lot of time on a big dollar amount and not move the needle at all, and sometimes I can spend some amount of time on a smaller dollar amount and move the needle quite a bit — both from a savings perspective and a sourcing perspective. So knowing where to focus has been extremely valuable for a procurement person like me. And if we take a look at the opposite side of that same coin, from a CFO perspective, I'm guessing that because you work for Tropic, a lot of times a request comes through and it gives you data so you know when to push back and when not. Justin, I wanna go to you just for a second because you had a post — I can't remember when it was, a couple weeks ago, a month ago, it all blends together — but you had a big dollar amount come through, and I'm not gonna say you rubber stamped it, but you approved it pretty quickly. Maybe share that example with us.
Russell [0:05:58]: Yeah. So I mean, we drink our own champagne, and as part of that, when a renewal is coming up and I get a reminder alert, I get a digest that is a snapshot of data relative to a purchase that's coming through. So we had a fifty thousand dollar renewal coming up, and the data said that the pricing was really good, the terms that we had were really strong. And while it's relatively expensive for us, fifty thousand dollars, there was very little to actually do with this, and the recommendation was just to auto renew. Which was kind of a breath of fresh air, because usually when I get those reminders, it's time to gear up for battle and prepare for the time and energy to actually work through that renewal. But in this case, it was a relief because that was time I could spend elsewhere, and the data was telling me that it wasn't gonna be a fruitful effort to spend the next several weeks banging my head against the wall.
Michael [0:07:05]: I mean, Russell, I don't know how that resonates with you, but I feel like that's almost become a requirement in this day and age where clearly suppliers can pop up, AI tools can pop up left and right, you've got legacy tools, you've got new tools coming in. Knowing where to push back to make sure you're getting a better deal and stronger ROI, but also knowing "this has strong ROI, this is a good deal," and letting it flow through and impact your business positively as soon as possible — that can become a competitive advantage. How does that resonate with you?
Justin [0:07:40]: It resonates. And what comes to mind is: I'm not just a hammer and everything is a nail, thus I have to treat everything the exact same way. I also think about — ironically as you were talking — my marriage. I've been married twenty-six years, and I've learned that not every interaction and conversation needs the same response. I don't always need to be in fix-it mode. Oftentimes, we have a great relationship, and she will say, "I'm just sharing this with you because I need to explain where I'm coming from, how I'm feeling." And actually, I don't need to do anything. I just need to be a good listener and hear the process that she went through. And oftentimes this happens in business. I've messed up at times when I've read a signal to say I need to lean in and put my foot down and really rattle the cages and make sure my team is doing what they need to do and go beat up a supplier. And if I come in hot without context like that every single time, it oftentimes isn't pointed in the right direction. So to your point, knowing when you need to lean in and light the fires and escalate is just as powerful as knowing where everything is fine — where we're green across the board, the benchmark came back good, we're negotiating according to the playbook, the supplier is playing ball, the team feels good about our compliance standards, and all we need is your approval. Or, on the other hand: "Russell, we're very stuck. Here are all the areas of friction. We need you to lean in." It's hard to do that quickly and efficiently if every single time you go through that cycle, you're starting at square one thinking, "Am I here just to listen and nod? Or am I here to actually engage and be the hammer with the nail?"
Russell [0:09:43]: One dynamic I think is happening a lot more — as you said that — is that I feel like in years previous, it was incumbent on a supplier's account manager or sales rep to build the ROI case, and that served as the primary data points that would go through to the CFO to cover a business case. The buyer was using the supplier's ROI case to justify the purchase. And now what I'm seeing a lot more is that because this type of data is available around benchmarking and insights around how a purchase falls relative to a comparison set, the expectation of buyers is now to build business cases not just with supplier-driven ROI cases, but also with buyer-driven data points to say: is this a good deal for us or not? And that addition is giving CFOs a lot more conviction and confidence that we are spending our money effectively, and confidence that the business team is being good stewards of the business and fighting for the right price points and justifying an expense to say: we are getting a good deal here relative to what we could be getting for this set of products.
Michael [0:11:04]: Justin, I wanna dig into what you just talked about specifically — the data and what you're seeing. Because I'm guessing not everyone is leveraging the same data. Maybe key that up a little bit: what are those different levels of data maturity, and what are you seeing right now in terms of how people are operating?
Russell [0:11:31]: Yeah. I mean, we have the benefit of getting to work with CFOs at very different sized companies — from small SMB AI startups that are moving at the speed of light, to mid-sized PE-backed or growth equity-backed companies focused on efficient growth, all the way up to enterprises that are public companies with public markets to answer to when it comes to their costs. So the needs around data and the expectations are quite different across that spectrum. When I think about small companies, that group may only have a small handful of suppliers that really move the needle. From a time standpoint, there just really isn't the same time in the day to run a holistic sourcing event — it's really about what do I need right now, and how can I have some confidence that I'm getting a decent deal? For them, it's all about the time element and maybe a single benchmark data point to say, "Okay, we're not getting screwed, let's pass this through and go back to focusing on growth." Compare that to the enterprise, where you have dedicated technology sourcing teams and procurement functions whose entire remit is focused on making sure you're getting the best prices and working with the best partners. They're going to have a much deeper intent around the type and amount of data they need. But even at the enterprise, they think about strategic spend in such a way that even if a data point says you're getting a good deal, if it's above a certain threshold, you're going to spend time on it regardless because the dollars are meaningful enough. Where the real value comes is in everything else — the mid to long tail — where you don't have time, and data can really point you in the right direction to uncover savings opportunities that may not have surfaced otherwise because they were below the threshold that met the criteria for focus and attention from bigger technology sourcing teams.
Michael [0:14:01]: It makes me wonder — and I don't know if there's a contrasting point of view from you, Russell — but I'd love to hear: based on everything Justin has said, when a request comes across your plate, what are you looking for? And how has that evolved over the past couple of months or couple of years? What did you used to do and what are you doing today?
Justin [0:14:28]: Oh, I have so many things I could say about this particular topic. First of all, what does a CFO look for? We're looking for alignment to the company plan and how this particular investment will advance that — whether it's revenue growth, efficiency, risk mitigation, customer sat. Think about what are we solving for as a business this year, the critical few priorities. Sometimes you'll see them on a plan on a page. Does this investment support that? What's the through line that helps me know this is worth spending money on, because not everything is worth equal focus and equal effort. But we also want to know what alternatives were considered, what are the measures of success, how will we know if when we implement this it will work. Obviously, are we getting a good price — is on every CFO's mind — how do we know that, and where do we stand right now? What are the additional levers we can pull? Are we getting the best payment terms we could get? Do we want this to be multi-year or not, and if so, why? With all these new AI tools standing up, I've spoken in the past about fears around Frankenstack — where you have this proliferation happening and you're going to get all these tools. So should we sign single year until we see how all these work together and avoid putting ourselves on the hook for multiple years? But even pausing and stepping back from all that — just data in general. The data is now more accessible, but there could be more false positives and bad signals. Here's my fear: we were talking to a friend that does legal work, and we said, "Man, is this AI stuff really hurting your business?" And he said it's interesting because actually, now everybody has a legal adviser or a financial adviser or an analytics adviser — they can run things through AI and they come back knowledgeable and with a ton more questions. Well, this happens to CFOs as well, because people can go out and pull data from who knows what source with who knows what freshness, and they could give false signals that say "this meets our standards, we're getting a good price, this is the best alternative." But if you ask AI questions in a certain way, it will give you back exactly what you asked. Like, "I'm trying to get approval from the CFO on making this purchase — provide me evidence that I'm getting a good price and that I've negotiated well." And AI will say, "Okay, I'll give you that." But you asked the question incorrectly. So the problem is that for all the data we have flowing, CFOs have to know: are we reading the right data? Where did this come from? What is the source of truth? Is this reliable? Am I getting a false signal?
Michael [0:17:41]: How do you protect against that? Because I can see — if I put myself in your shoes — some sort of justification coming across that's very convincing but was likely generated by ChatGPT. And to be fair, I've heard of other procurement leaders using ChatGPT or Claude or some other AI as a benchmarking service, and that scares me — where is that data coming from? How do you protect against that?
Justin [0:18:14]: I think we have to ask more questions, and the questions we ask may have to change. So if they're getting benchmark information and negotiation strategies from Tropic, will we know that that's first-party data from live discussions and live negotiations — not some random generic AI casting a wide net into ChatGPT and praying something comes back that's legitimate, but is actually third- or fifth-hand and five years old and not even relevant? So it's: know your source. It could be software purchases, or it could be a marketing campaign or anything else — what is the source of their information and how did they validate the accuracy of it? CFOs have always needed to do that, though. When salespeople would come forward with funnel conversion estimates, we need to know: what's your source? That hasn't shifted for CFOs, but we need to not rest on our laurels and be lazy just because now we have all this data flowing.
Russell [0:19:26]: Authority and credibility play a much more important role in this age of AI — exactly to that point — because anyone can put together a really strongly worded and convincing argument with the power of ChatGPT. And I think it's tricky because as CFO, you want to trust your people to come up with convincing business cases and well-reasoned arguments, but knowing the tools that everyone has at their disposal now, there's a question that inevitably comes to light, which is: tell me your source. Where did this come from? How did you come up with this? And asking those questions casts a little bit of doubt on the credibility of whatever analysis has come together — which is not fun and not empowering for a stakeholder — but it's probably increasingly necessary to isolate whatever noise might be coming through the abundance of data from the real signals and reliable sources that matter for each of these situations.
Justin [0:20:34]: Yeah, definitely. I don't want good business leaders to short-circuit intuition, experience, and context in the process. Because in the past, if it was more manual, their own internal filter around those things would be higher. But now that it's easier and more accessible to spin this stuff up and crank it out in a beautiful presentation that's utterly convincing with charts and quotes — and then when you ask, "Did you make any of this up?" and it's like, "Well, yeah, I did, because I was trying to build a good business case for you." So I think it is all about the questions we ask, and really having our antenna up — not in a suspicious way, like people are trying to bamboozle us — but more in a way that understands: this is now where we're at, where we need to authenticate our source. They used to need the good housekeeping seal of approval, or like Consumer Reports — that's what we're talking about. How do we know: single source of truth, trusted, reliable. If only everything we got had some seal on it, like "this is trustworthy." We don't have that, so we just have to be in tune.
Russell [0:22:16]: Last thing I'll say on this: even three months ago, I would probably give extra kudos to some AI-generated analysis on a specific sticky problem. Like, "Oh great, we're using the power of AI to diagnose and look at this problem." And I can feel myself pulling back from that sentiment today, because I know the great parts and the faults of AI. I'm increasingly more supportive of the human intuition piece connected to AI. And I think there's definitely a tension that exists between using the power of AI to synthesize large amounts of data and come up with decisive claims, versus the human intuition side that's layering on business context and the softer parts of what good analysts have been able to do over time. But just to summarize: credibility, source, and authority are all elements of data that are going to have increasing relevance, and CFOs are going to be asking for the source to validate the assumptions or outcomes that are being determined.
Michael [0:23:43]: Yeah. I actually wanted to double-click on what you just said, because what I heard Russell say is that he's looking for: number one, how does this tie to very clear business objectives that we're trying to hit. Number two, how do we know this is a good deal — benchmarking and that sort of thing. Number three, how have we optimized for contract term length. Number four, how do we gauge this tool versus something else already in our tech stack. And number five, what's the source of the data for all of those things. Did I summarize that okay, Russell, or did I oversimplify?
Justin [0:24:38]: You made me sound so smart and knowledgeable — well done.
Russell [0:24:42]: That was the AI system running alongside this that just summarized that. That was good.
Michael [0:24:49]: Real time, real time. That's right. I wish I had that going. Justin was telling me about a tool he was using last night and I thought that was great. Look, I think that's very powerful information because I do feel like in this world we operate in, the way things used to be from a procurement angle — focusing on the biggest dollar things — is not sustainable. And also from a finance angle, speed matters sometimes, protecting against risk matters, but knowing when to optimize for each of those things in this scenario can be a pretty good competitive advantage. Justin, I wanna click on one more thing. We had an awesome conversation yesterday with some procurement leaders, and one thing they were concerned about is the evolution and recency of the data — how we can stay on top of that — because even benchmarking from potentially a year ago might become stale really quickly. There was a big discussion about a massive migration, especially when it comes to SaaS, from historical seat-based models to now usage-based or consumption-based pricing. How do you address those concerns?
Russell [0:26:15]: Well, I think the first piece is awareness — knowing that a change is coming is as helpful as having a bunch of benchmark data around that change when it does come. So for example, we can often detect when there's a pending pricing and packaging change or a new commercial model being rolled out by different suppliers. We have the technology and capabilities to identify changes in real time and broadcast them out to our network of customers. That allows them to be thinking proactively about these changes and not be caught flat-footed at renewal when supplier X comes with a thirty-seven and a half percent price increase due to deprecating one SKU and moving everyone over to another. Being aware and having those things on your radar can make all the difference in how you approach a sourcing event. And then in terms of the actual benchmarking data itself, that's where the engine and the machine become so critical. One thing I'm really proud of in what we've built at Tropic is the machine to capture these signals in real time and expand coverage on new pricing models and new packages in really rapid succession. I learned a new phrase last night: "human on the loop" instead of "human in the loop" — which means a human that's reviewing and on top of the loop to interject in the moment, versus a human in the loop, which is a stage gate that can slow down the whole system. Thanks to my friends at Meta — that's a new phrase we're going to be taking on. But yeah, building that human-on-the-loop mechanism allows us to rapidly grow and collect data points so that as those changes materialize, we're able to be first on the scene with what's possible and what's not possible in these new circumstances. But it's hard — there's a lot of change happening at every supplier right now as they rapidly evolve into AI-first with new models tied to outcomes, consumption, or credits. And usually when something new comes out, that inevitably means they haven't figured everything out, and there's a lot of flexibility that can exist. So what we always coach our procurement customers is that even in a less data-rich environment, you can be fairly confident that when something new is rolling out, there's a lot of flexibility in place as they look to understand and optimize pricing, packaging, and growth within those new models.
Michael [0:29:19]: I think that's a very astute takeaway — that the flexibility exists in the beginning and probably becomes more rigid over time. I don't know if that's universally true, but certainly a trend we're seeing.
Russell [0:29:29]: Yeah. Especially in these competitive dynamics where people are fighting tooth and nail to maintain market share and just land-grab from competitors. We know that commercial models and pricing is a tactic that can be used to get a deal over the line and expand adoption on a new initiative. So there are ways to deploy those tactics to drive the outcomes.
Michael [0:29:57]: Okay. Let's go around the horn as we wrap this up — one actionable takeaway that you feel like should be prioritized. What's one thing you're rolling out Monday? Russell, we'll start with you.
Justin [0:30:15]: I'll start with a small story if you'll permit. I had an opportunity to meet and hear Dabo Swinney speak last night. Do you know who he is?
Michael [0:30:28]: I've definitely heard the name.
Justin [0:30:29]: The head coach of Clemson for over twenty years — a very successful and still current head coach. I got to hear his story of how he came to be in his role, but also a lot of interesting points around just the role of the coach. There was a live voting thing where we were supposed to vote on what leads to success for football programs these days. Is it their NIL budget? Is it having a strong defense — of course a lot of the Georgia Bulldog fans were all for that. Is it recruiting the best quarterback talent from high schools, or is it the importance of the coach? And there was a lot of great discussion. Of course, you could make an argument for any of those things. And of course, the conversation was about the importance of a really strong coach. Coaches have years of experience in the trenches where they've done it, they understand it, and they've built up trust with the players so that when they grab a player by the face mask and say "listen to me, do this differently," they hear it differently than if they just Googled it or heard someone on ESPN say it — because they trust the coach. And so with all this data flying at us faster than ever, and not knowing who to trust or what the source of truth is, we've got to stay close to those that are experts in the swim they've grown up and grown accustomed to. If we're talking about buying things, we need to trust our procurement people who are skilled at doing this. We need to trust our negotiators who are hand-to-hand and current with what's happening, what's the latest trend with all of the suppliers. Short-circuit that, and suddenly you have AI tools and you can ask ChatGPT or Claude some string of questions and get back a response that makes you feel confident. That would be like Dabo Swinney's players ignoring what he says just because now they have AI. We'd agree that would be ridiculous. It is no less ridiculous in the world of buying things.
Michael [0:32:53]: Yeah. I like that you used the word trust, because one thing I'm taking away — and I'm already thinking about whether I should modify the workflow or build some sort of quick agent to help ensure that when a request a stakeholder submits comes to you, Justin, as a budgetary approver, or to you, Russell, the CFO — that it captures that checklist we talked about, but also challenges the sources to make sure I'm not violating that trust. Because the moment that trust goes away — the moment it's like "you leveraged AI or ChatGPT and your credibility is questioned" — that kills the trust. And then it's like, okay, everything looks good, but is it really good? So that source has to be there, as well as "here are the four or five things the CFO or budget owner is looking for." Russell, over to you.
Russell [0:34:00]: Yeah. If I'm a CFO and I'm reviewing business cases or deliverables, I think we're now at the place where we can expect that AI is pervasive in basically anyone's analysis — too accessible, too easy, and too good at what it does to not assume that people are using AI to arrive at their assumptions. And if I'm a CFO, I'm now asking people in these different business cases: obviously, the source, like you both mentioned. But I'm asking what's your intuition, versus what's the AI saying? I want to see and understand where you as my trusted colleague in this organization are using your brain power to interpret what an AI synthesis would be and layering in your unique and softer nuanced synthesis on top. So I think that's going to be increasingly included and expected as AI adoption reaches its apex. The value of human intuition incorporated within a recommendation is gonna grow in importance, and it's going to be isolated separately from the AI synthesis in those analyses.
Michael [0:35:18]: Nice. Alright, let's put a quick bow on this. If you are in procurement like me, leveraging a strong dataset is going to be a competitive advantage — it'll help you know where to focus and where you don't need to spend a lot of focus, regardless of dollar amount, because you're factoring in things like benchmarks, supplier data, stakeholder sentiment, sourcing data, and so on. And if you're in finance, a financial budgetary owner, or a CFO, you also need to make sure you're harnessing and leveraging that data — so you know when to push back and when you can accelerate and quickly approve something and get it into the hands of stakeholders who clearly need the product.
