Daniel Gardner is operational finance business partner at FirstGroup Plc, a leading UK-based provider of public transport. Daniel brings a unique perspective shaped by a diverse finance career across iconic consumer brands like L’Oreal, the Body Shop and Hunter Boots, where he led major forecasting overhauls and drove commercial transformation. Now at First Group, he is leading the development of a cloud-based forecasting system for a billion pound division, working with more than 50 stakeholders to deliver scalable real-time insights.
In this episode he talks about the power of “systems thinking” in the CFO’s Office as a way of understanding how different parts of an organization (or any complex system) interconnect and influence each other.”If there’s an area, I’d say that FP&A could do with its systems thinking. Structure makes behavior and behavior makes structure. So it travels in a loop, which means that to make changes in an organization, you either have to hire people who do not exhibit the behaviors naturally or you’ve got to change the structure and alter the incentives that are producing the problems.”
In this episode:
- Bringing my philosophy-training to finance
- Six week finance transformation at Hunter Boots
- Getting from lagging to leading indicators
- 99% of People Problems Are Really System Problems
- Python+ Excel (practical examples)
Full Transcript
Glenn Hopper:
This is FP&A Today. I’m your host, Glenn Hopper. Today’s guest is Daniel Gardner, operational finance business partner at First Bus, which is part of first group PLC. Daniel brings a unique perspective shaped by a diverse finance career across iconic consumer brands like L’Oreal, the Body Shop and Hunter Boot, where he led major forecasting overhauls, launched new D two C channels and drove commercial transformation. Now at first group, he is leading the development of a cloud-based forecasting system for a billion pound division, working with more than 50 stakeholders to deliver scalable real-time insights.
Daniel combines deep technical finance skills with a practical systems thinking approach to fp and a, along with a philosophical curiosity about how we plan, predict, and make decisions. We’re excited to have him on the show. Daniel, welcome to the show.
Daniel Gardner:
Hi, Glenn. Thank you very much for having me on. It was a very kind deduction.
Glenn Hopper:
Yeah. Uh, <laugh>, I stumbled a minute because when I said a billion pound, of course, in the US when I say a billion pounds, I’m thinking the weight of your organization <laugh>. It sometimes feels like a billion pounds of weight, but there we go. You know, <laugh>. Yeah. I loved our, our conversation we had before the show and, uh, and your path into finance has been anything but linear, I guess is a simple way to put it. I mean, you started out in retail, environmental engineering, and then, you know, eventually made your way to, to corporate fp and a. Can you kind of walk us through that journey and maybe if there’s, I, it’ll, it’ll be interesting to hear, but if there’s any insights that you gained from those early roles that still maybe influence how you think about numbers, systems and people today.
Daniel Gardner:
Yeah, sure. Uh, there, there definitely are. I started relatively late. I had to work, I was in my late twenties before I even really began the study process. Uh, I had to work without a mentor or support. I had to borrow better money to study, um, which is unusual for a lot of accountants in the UK who normally go through a kind of a corporate route associated with an audit firm and they sponsor effectively. Uh, I, I didn’t do any of that kind of stuff. A lot of the math was quite new to me as well, especially around things like the options calculations, black skulls formulas and things. And, uh, I had to kind of build back up my understanding of algebra and, and calculus and ’cause I, you know, been quite a number of years since I’ve done any kind of math. So I let right in the middle of it thinking, oh, you know, I’ll get started.
No, no, I’ve gotta go all the way back to the beginning. But that actually was kind of good for making me confident in tackling new subjects. So that was helpful in a way, uh, to do that process. Frontline retail actually has always held me in good stead. ’cause as much as that’s a, a, a sort of a very standard job for a lot of young people, it helped me understand a lot of the detail in things like the body shop, L’Oreal and Hunter. You know, I could relate to a lot of the lower level staff. Some of the challenges they raised, the practical difficulties they were, were exposing, which might have been a bit opaque just from the pure finance side, but my, my journey’s in environmental engineering as well, where, you know, I, that was focused around two particular industries, which was sewage in cremation. So I have set up to my thighs, um, in human waste and been covered head to toe and ash from blow back from a chimney. So, you know, it’s,
Glenn Hopper:
Uh, this was both before you got into fp a right?
Daniel Gardner:
<laugh>? Oh, yeah, yeah, yeah, yeah. Totally. So, well, I mean, environmental engineering was where I kind of sta that was where I started to study, um, accounting. And so I was, I was helping out on site and doing a bunch of things as well as this. But coming into that kind of experience meant that, you know, a lot of the threats and challenges that corporate life throws up for people just didn’t, they didn’t strike me as, as, as dangerous as they might have done if I’d just gone straight from, you know, um, I dunno, PWC or Ernst into, into corporate life. So, gave me a good grounding.
Glenn Hopper:
Yeah. I love what you said about retail because it’s so easy, especially people if you just go this the traditional route, whether in, in the us you know, whether it’s, uh, you know, undergrad in, in accounting or whatever, and then, and or an MBA program, it’s very easy to think of finance and accounting as this ivory tower that is just, you’re doing the numbers and you’re not part of the broader business. And I think that’s, that, that mindset is going away really kind of across the board now. But when you’ve been in the other role and you understand that front end, what’s happening and that part of the business. And so even if somebody doesn’t have that background, I always encourage people starting out or wherever you are in your career, don’t just sit behind your computer and build your models. Get out there and see what the rest of the business is doing. Work with, work with sales, work with marketing, work with ops.
Daniel Gardner:
Yeah, a hundred percent, a hundred percent agree with you on that. I mean, this is something that, you know, I’ve, I’ve said to a number of people that 99% of the people you’re ever gonna come in contact with are never trying to do things badly. They’re just working with a broken system in the best way. They know how that’s nearly always the case. So it’s really the key for, for a lot of finance business partners, particularly now that role’s become so much more ubiquitous for us to, to get out there and talk to people and find out what are the real sticking points about why they keep putting things in the wrong box or filling out the form incorrectly or missing something. There will be a driving commercial reason or operational reason why that is. And if you can ue that, usually then the whole thing starts to flow so much more smoothly, and you get a much better appreciation of what the numbers really represent as well. In terms of a value of work, particularly in my current role, we get, it’s a very large business public, uh, transportation, but a private company operating that on behalf of either the government or local authorities or commercially. But you can get very, very small changes. 0.1% can represent enormous amounts of work on the front line. So it’s a, you have to really embed that to understand it.
Glenn Hopper:
Yeah. Another thing, when we were talking before the show, and, and you corrected me, and I love this because there’s, it’s, this is where clarity matters. We were talking and I said, you didn’t study accounting or business, and I started to go on and you said, <laugh>, I didn’t study accounting or business at university, which is a clear distinction because I just, I was picturing you just naturally just intuiting everything about finance and accounting <laugh>. But, uh, where, where I was getting with that though is, you know, your, your background, kind of like your, your work background before coming into fp and a. It’s not the, uh, traditional route into the job. You studied international relations and philosophy. Hmm. And I’d love to hear about that and sort of that background has to give you a different mindset. I’m wondering how that academic experience kind of shaped your approach to finance and, and maybe, maybe especially in terms of decision making or critical thinking or, or even long-term planning.
Daniel Gardner:
Yeah, a hundred percent. I originally studied politics and international relations at university purely because I was certain I could finish it. It was something that I’d seen with a number of, um, my, uh, compatriots that they had started university degrees, borrowed a lot of money, and then never actually managed to even finish the degree. So I thought, okay, look, let’s just do something. You’re actually gonna, you know, complete. Because 40% of the job adverts didn’t even specify the degree. They just said you had to have one. So I was like, okay, we’ll go down doing that. I actually took my masters in philosophy when I was already deep into fp and a because it was one of the things I’ve always felt was a benefit of philosophy is that it forces you to think through all the potential paths through the maze. You have to exclude as well as include to demonstrate proper rigor in your thinking. You have to demonstrate why the other options will not answer the question or the problem. So you have to actually be able to restate the problem and understand how your critics think. So really good quality philosophy will not only force you to communicate well in your writing, it’ll, it’ll ask you to keep your thinking brief. You know, you should always be taking words outta the sentence that aren’t required, which is definitely sticks with me when I see PowerPoints with over a hundred slides on the deck. <laugh>. Yeah.
Glenn Hopper:
A million years ago when I started out in fp NA, I don’t, I don’t even think it was called fp NA back then, I was just the, the finance guy. Um, but it was really about the, the where you stood out was how great your models were and, uh, and sort of the technical work you did. But over the years, it’s transitioned to now business partnering that we discussed earlier, that’s a big part of the role. And storytelling is more important, and I think it always has been, but we didn’t label it as such. So you could go through and build the greatest model in the world, but if you can’t spread the narrative, if you can’t tell that story, then it’s, it’s worthless. And the inquisitive part of a philosophy major is interesting because that’s what we do in fp and a. We’re investigators, we’re trying to figure out what’s going on underneath. So sort of that classical training on how to think and be curious and ask, ask the right questions. Yeah. And keep digging out. It seems like that would have to come through.
Daniel Gardner:
Yeah, it really does. I mean, I, I’ve, I like the commercial roles. I, I never was an auditor, so I didn’t have the benefit of that experience. So I, I I, I want to acknowledge that very clearly that I’m not a tax person or the guy you ask about IFRS. I’m very good with systems thinking. I can see the whole picture in laser in on one detail, but I don’t want to, um, give people a false impression that, um, you know, I could do a lot of the very tight technical work that some of my colleagues are, are, are exceptional at. And that’s, I think that’s a, a key part of it is being able to know, you know, if you can emphasize your strengths, then you can really contribute to the story in the right way.
Glenn Hopper:
Yeah. That’s why there’s an fp and a team and not just a single individual
Daniel Gardner:
<laugh> Yeah. On there.
Glenn Hopper:
Totally. Yeah. Yeah. The team with the different strengths and everything. Yeah, when I’d took, I came up through fp and a too and my first CFO role, I, if I didn’t have a good controller, I’d have beens. I didn’t, you know, I was, uh, it was, it was a mess for me. So I a hundred percent agree with you there. Well, digging into your, the specifics of your experience, I know working with those global consumer brands, what were some of the most formative projects you worked on there and was there anything that you took from that, that it expands over to your, your current role?
Daniel Gardner:
Yeah, so I helped build up cloud forecasting systems for L’Oreal and Hunter. Uh, we did Hunter’s System in like six weeks. Um, the L’Oreal one was a kind of a bespoke one. They developed themselves, but Hunter’s was one, I actually helped to get design right from the ground up, including all the calculations. And that was some of the experience I was able to take across. The first, doing it in six weeks was not originally the plan at Hunter. There was supposed to have been a six month project. Someone who shall remain nameless may have announced to the board a bit early that this great thing was coming. And then all of a sudden it was, you know, definitely wasn’t me, uh, <laugh>. But there were just lots and lots of major projects, product launches, new lines of business, creating companies from scratch. So there’s an awful lot of this kind of commercial activity and projects that I like to get involved in.
And I like the variety of it. One, one area actually I really found particularly enjoyable was kind of the challenging legal work. I’m very much a dog with a bone, uh, a bit belligerent. So, you know, actually it helps greatly when I’ve got sort of high stakes conversations with suppliers or customers who are challenging things on a legal basis to be able to nail on all the little details and produce those huge chains of, you know, timelines for emails and things. To be able to go through all of that, you know, tooth root and branch and be able to say no, that was not what was agreed at the point. And those kinds of things that, that kind of work has been, has served me very, very well in making sure that when I’m then building projects up for any future companies I work with, I I know already how to start the process to ensure we don’t end up in that mess in the first place.
Glenn Hopper:
Yeah, that’s an example. And I know we’ve got another one I want to hit on too, though, where again, it’s finance getting out of that ivory tower of just being the numbers and moving out into the other operations of a company. So we talked about at Hunter Boot, you led initiatives ranging from launching the new e-commerce channels to overhauling merchant fees and customer service. And I, that’s broader than obviously your, your typical finance. But I’m wondering from that experience, what does that teach you about making finance operational, or what, what is something that you could share with our listeners of, that may not sound like a finance role to begin with, but how you moved across it and maybe how you think about the balance between agility and control through something like that, where it’s not really like, finance doesn’t own that project, I wouldn’t think.
Daniel Gardner:
No, no, absolutely not. It’s always gotta be commercially and operationally led. But it’s, it’s interesting that so often we put agility and control of some kind of two opposite ends of a spectrum. And actually, I think there’s a, an argument for saying that not only are they, they not mutually exclusive, if you’re gonna succeed, you have to be able to do both simultaneously. I looked it up afterwards when I, when I quoted it to you originally, that it was actually the Prussian general Helmuth von Mulcher who said, no plan survives contact with the enemy. So we should kind of sur we should expect every single forecast to be wrong. The moment it’s issued, it’s not, it’s, but it’s not the, the fact that the forecast is wrong, it’s the degree of the wrongness that matters. You’ve gotta focus on the materiality of the change and the operational commercial driver that’s causing that move.
You don’t want to control everything so tightly that nobody’s got any ability to innovate or move or switch plans. But you equally need to be able to, you know, let things move around a bit to the extent that they’re not simply flying off the handle and spending money without any, um, any kind of control. I think what it boils down for me was too often we end up focusing on lagging indicators. The p and l is entirely a lagging indicator. If you switch to looking at leading indicators, you actually have much more effective control. Purchase requisitions is a great example of this. If you sum those up versus the budget, you can then see how much of the budget’s already been spent. And it’s much easier to cancel a PO that hasn’t been fulfilled and to argue about an invoice that’s unexpected and takes you well over budget. It’s too late usually by that point.
Glenn Hopper:
Yeah. It’s so funny how that, I mean, that’s just the default though, isn’t it? Yeah. Just to go to the p and l and say, oh, this is our trend and let’s just <laugh>. Yep. And it, we, we’ve been creeping up in, in cogs and creeping down in revenue, so let’s just keep pushing that out and not looking at actually the levers that actually drive it.
Daniel Gardner:
Yeah. If you go backwards in the process, if you get away from your desk, go to the frontline staff and understand the commercial operational drivers and get backwards through the, through the steps, you’ll usually find there’s something sat there as a little nugget that’s causing everything else to roll forward in a snowball. So if you start tracking that, if you find a way to measure that problem, all of a sudden you’ve got, you know, way ahead of where you need to be in terms of preventing the ugly looking p and l coming out. The other side of it,
Glenn Hopper:
Fp and a today is brought to you by Data rails. The world’s number one fp and a solution data rails is the artificial intelligence powered financial planning and analysis platform built for Excel users. That’s right. You can stay in Excel, but instead of facing hell for every budget month end close or forecast, you can enjoy a paradise of data consolidation, advanced visualization reporting and AI capabilities, plus game changing insights, giving you instant answers and your story created in seconds. Find out why more than a thousand finance teams use data rails to uncover their company’s real story. Don’t replace Excel, embrace Excel, learn more@datarails.com.
You might have sort of already answered this, but I’m gonna, I’m gonna push a little deeper forecasting, identifying those levers, looking at the leading indicators that is the crystal ball of fp and a. So I’m wondering if you have, and you, and you probably pretty well hit on it there, but what your philosophy is on forecasting and, and budgeting, especially as you have working with large distributed teams where you’re trying to coordinate everybody’s thoughts and getting them directionally right to, to be on the same page with the forecast. And in that, do you see, is there an area where a lot of companies get it wrong?
Daniel Gardner:
Ooh, there’s the $1 billion question.
Glenn Hopper:
<laugh>, the 1 billion pound question. <laugh> the big heavy question
Daniel Gardner:
<laugh>. Yeah. And that definitely does weigh a billion pounds. So in terms of working with a large distributed team, I’d say you cannot communicate enough. And alignment is an ongoing process that will never end. Just get over that fact that you can’t stop having meetings every single week about it. You just gonna have to keep doing that. You know, indeed, if you assume everybody heard something different and totally missed the point, you’re probably in the right place about how you, how people are receiving the information they’ve already got and are acting on it. Um, you know, I, I have a, a, a forecast list that rubs to over 500 steps. So I’ve got very, very detailed individual steps in the process. I actually spend most of my time marking down that things didn’t get done. But I’m okay with that because it means successively over time, I can start to show a map of these are all the areas that we just can’t get to in time, or these are all the areas that keep failing and falling over.
And by doing that, that’s how you then start churning the wheel to say, okay, now we can start pushing for that one baby step of let’s get commercial aligned, or let’s talk to the frontline retail staff, or, you know, let’s talk to e-comm about how they’re really doing the digital marketing. I also think there’s something worth saying at this point, particularly in the, in this modern era with us having a conversation over, over the web that, you know, we, we should never assume that just because somebody’s working from home that they’re not working. And that just because somebody is working from home, that they are totally with the plan and understand everything that’s going on. There’s, there’s a real challenge for a lot of managers, I think, to transition to being able to manage at a distance, because it’s not easy. It is quite intimidating to do, but I think it’s a critical one to get comfortable with because talent is now accessible anywhere.
But if, if that means, if you’re gonna retain it and look after it, you’ve gotta, you’ve gotta make sure that people feel like they’re included in things, even if they’re typically on the, on the outside of the end of the team’s call. I should be very clear here. I am not talking about any specific company, what I’m about to say. Fair. Yep. Some generic feedback from the last 20 years of me doing this because, you know, it would be easy to interpret that what I’m about to say applies to my current organization. And actually, ironically, one of the reasons I’m quite happy where I am is precisely because it doesn’t, but moving on, I don’t think people should assume the narrative is obvious. Why are we expecting X value of profit or sales? Quite often that’s communicated out, like the target is just a thing that everybody gets.
And I don’t think that’s the case. I’ve often seen, uh, particularly in previous companies no joined up thinking where, you know, there’d be this, you know, we’re gonna have a sales increase of 20% and no attendant costs or requirements to scale customer service or the back office. Uh, okay, so how are you servicing this 20%? Um, you know, I’m thinking particularly of one organization here allowing ego to dominate the decision making. You know, if you wanna work on gut feel, I don’t think that’s wrong, but you should at least have the courage to admit that, you know, costs are hard and sales are soft. It’s very easy to spend the money upfront. And we typically do that in advance of making the sales in the first place. You’ve gotta have the stock in the building in order to send it out, right? So it takes that degree of faith to buy all that stock, particularly if you’ve got many months of lead time on your supplies, but as soon as you do, we’re now committed to your 20% sales, increase that kind of cash flow cycle.
It tends to be something that only finance are the ones biting their nails about. So I just, you know, that’s, that’s often something I’ve found. I think one that’s pretty ubiquitous also is most organizations say they wanna forecast from the bottom up because then it’s owned by the frontline teams and there’s accountability, but then they want to issue top down targets. And I have to say, this is, this is actually a really big problem for the modern era with ai because it’s very easy for those circumstances for AI to simply hallucinate the gap and be overly optimistic. So there’s, there’s, there’s some control issues that are gonna start emerging, I think, that are gonna really cause companies some real headaches and they’re not yet conscious of it because it’s, it’s not yet hit them how real that problem will be.
Glenn Hopper:
Yeah. And that’s, that’s always a conundrum because you, you know, you put together your annual plan and the company has, we know we’ve, this is our keer for the last three years or whatever, and we putting more investment in or whatever. And so we need to increase that. So it starts with this goal, and then you have to reverse engineer to the goal, which is fine. That’s part of the job. Yeah. I mean, is to taking that strategy and, and figuring out what it costs and, and figuring out all the components of it, then you’re already forced into a, a top down. But then at the same time, you want the departments to say what they need to do their job and layer it on there. So in my experience, it always ends up sort of melding together in this weird place in the middle where there’s a whole lot of mistakes could happen because when you’re getting pressure from both sides, that’s where something’s gonna squeeze through the cracks and you’re, you’re gonna miss it and not put the right increase in headcount that’s gonna drive, you know, the operational headcount we’re gonna need to support the sales or whatever the case is, increased inventory or,
Daniel Gardner:
Yeah.
Glenn Hopper:
Um, it’s just a lot can get missed when you’re, uh, have directionally competing drivers.
Daniel Gardner:
It is probably always gonna be that way that you get the meld. But I think that’s the, the challenge is that there’s a, there’s a feeling that I’ve often come across that people, the senior team, the board want the ownership and the accountability to be vested with the frontline staff because they want them be bought into the mission to believe in what they’re doing and to go forward with it with gusto and to own it. But equally, if you push the top down target, it’s not theirs anymore. So it’s just that, it’s that subtle, subtle balance. I think it, it takes far more of a, um, there’s an interesting sort of nature of belief you have to get into about how, how, how you can create some faith in the number and in the processes that you’re asking people to follow. I think that’s the, uh, the big challenge. Yeah,
Glenn Hopper:
I, uh, just thought of something when thinking about those numbers meeting in the middle. My very first exposure to budgets was before I was even in finance, I was in the military and when, uh, <laugh> the, it would get to fourth quarter and you had budgeted certain items and you hadn’t spent them, the mandate was, go spend all this money right now because if you don’t spend it, that’s where your budget for next year is gonna start. So I guess I’ve come a long way from that era, but every, every year in budget season, I would think about just, you know, it’s like if you’re running your budget based on, you know, if your tax accountant is doing fp and a for you and they’re telling you spend all your cash before the end of the year, it was a very similar approach. But, uh, yeah, glad glad that the companies I worked with since then don’t run, run that way.
A couple of items you hit on, there was a lot there, but two that stood out to me and, and the first one just, um, if you could dive a little deeper on it. So you talked about your 500 item list for the, for the budgeting and I guess this is essentially an SOP and your whole team kind of goes from that list and you see what it is you wanna do and then it sort of gets stack ranked and prioritized based on what is feasible and you know, you got your must have nice to have <laugh> Yeah, absolutely. Whatever your stretch goal kind of things. Yeah. And then just one comment that it we’re removed from it, but I, I do wanna mention this ’cause you did bring up remote workers, and this is a problem that I, I have to remind myself on remote workers, when you only, you don’t have that water cooler thing, you don’t have the one-on-one, you know, walking into somebody’s office and doing the coaching and you think about the way that we use everything from AI to system automation.
If there’s just someone on the other end of your computer screen that works for you, it is very easy to forget. That is a human being on the other side of this thing. I can’t just prompt it like I do my ai and remember, you know, as you throw work, it’s very easy to just throw that prompt across the board and, and let the person do it. And I’m, and I know there’s a mix between people going back to the office and, and being remote, but that’s, I guess this is my public service announcement of remember that person on the other side of the screen. <laugh> is a human being and not, uh, chat GPT that you’re just lobbying prompts into
Daniel Gardner:
<laugh>. Yeah, I mean, the funny thing is, if, if you start getting more distributed teams, it actually opens up an opportunity to go visit other parts of the country and go and meet them face-to-face. It also is, is you can do coaching over teams and I have done it and I am doing it now. You can coach and train people remotely. It’s totally possible. Sometimes it can be even, you can make it more fun because there are so many more collaborative tools. You know, most people I found do not use things like the whiteboard functioning teams don’t use a load of the other transcription tools and other things you can get involved in doing. So there’s nothing wrong in starting to utilize some of those tools, um, not in the kind of the fake icebreaker way in a meeting, but actually to genuinely engage and think about problems together. I I used to have a colleague from a previous job who had a, because they were pretty much all remote, they used to take it in turns to all be on the company kind of, uh, intranet contact communication thing at the same time. And each person get a chance to play some different music over the speakers. So they just go round the remote workers that way. All getting to enjoy something like an office life, even though they’re all actually totally physically separate. Yeah.
Glenn Hopper:
On that note, you know, coaching, again, it’s, especially you get under stress and all that, it’s one of the hardest shifts to make as a manager to get from just producing output and being part of the, the machine that is generating output and insights to remember the people who helped you get there along the way and, uh, and and being able to, to pay that back and, and bring, bring your team up. And, um, listeners of this show are going out and, and hoping to get insights from, from all the guests we have on. And so one of the questions I always like to ask is sort of a, a general broad coaching question out there. If you could give just one piece of advice to a finance team that, or indi it could be individual level or team level that are trying to level up their fp and a capability, wondering what that would be. And I’m, I like asking you of this because I know you’re very skilled in the, in the forecasting and modeling and all that, but you also I think maybe bring a slightly different mindset to it. So I’m wondering is it a mindset shift that you would advise them or a process change or something else entirely?
Daniel Gardner:
So if there’s an area, I’d say that perhaps fp and a could do with looking at more it’s systems thinking. And I don’t necessarily say that to just give a kind of an academic recommendation. I mean that structure makes behavior and behavior makes structure. So it travels in a loop, which means that you e to make changes in an organization, you either have to hire people who do not exhibit the behaviors naturally or you’ve gotta change the structure and alter the incentives that are producing the problems. If you start thinking about it in those terms. ’cause I think a lot of the time when I’ve seen fp and a done badly, if I put it that way, it’s because they’ve made personality the issue that won’t do what I’m, you know, it becomes finger pointing and name games rather than looking at it in a more holistic perspective and saying, okay, why is it we keep getting this outcome, this thing coming out of the back end of the machine or the process or the organization or the person that satisfies none of anybody.
You know, that’s the thing. It doesn’t even please the person doing it <laugh>. So it’s often, it’s often that result that you need to get really underneath and that for that you have to kind of look at the, the structure of how you got there. And usually it comes to that really ugly answer. Like, well yes, ’cause we’ve got a 20-year-old IT system, we don’t have the money to replace it, but at least if you can get to that as a more concrete thing, you can start break everything down to baby steps. What’s the smallest step forward you could take that will start moving things in the right direction, even if it’s something as menial as getting one person to do all of the manual corrections on a particular process, because then at least you’ve centralized that in one place or distribute it out so that everybody’s having to share the work it, I’m not saying what the right answer is in that regard. It’s more, you know, thinking creatively in that sense about the structure.
Glenn Hopper:
So for our listeners who aren’t familiar with it, let’s let’s drill down into systems thinking, because whenever I hear that word, I picture, uh, either a Rube Goldberg machine or those overly complicated YouTube videos where somebody knocks over from Domino’s and they run in a series of events happens and it launches a hot dog into their mouth for their lunch or whatever. But break down systems thinking what you mean by that.
Daniel Gardner:
So what I’m talking about is the idea that you’re not seeking to say that the system is made out of steel and concrete and can’t move or change. You’re, you’re just trying to produce a map of the terrain. Too often people mistake the map for the terrain. They, you get those beautiful process diagrams that you, and you know, this is working in America with socks. You’re gonna have this beautiful po you know? Yeah. And what do we all know? We pull out the process map and we look at it and everyone goes, that’s not how that works. Now that what I’m saying about with systems thinking is that you start again and look at how does the system actually work and look at it from a position of all of the individual elements as particular stakeholders. You might have particular people with particular personalities that mean that they’re never gonna entertain certain options just ’cause of the kind of person they are.
So if you start mapping things out that way and thinking about the different states that, um, that exist out there and the ways that information and, and behavior transitions between all those things. This, this is a far more fluid explanation of system dynamics than I think any, than any true system dynamics would ever allow for. There’ll probably be people spitting out their conflicts listening to this, but it’s <laugh>. I I’m trying to make it something practicable and workable in a, in a corporate environment. And I think that’s the challenge is you’re not looking to just take an academic model and bolt it on. You wanna make it something that libs and breathes. So to some extent, it’s almost like having the process map open and it’s continually evolving as you start improving and changing things. You, you would often be very shocked, I think, how quickly you can get a ripple of change just by tweaking one part of the whole process right at the top miles away from where anybody’s even looking at it.
Glenn Hopper:
Yeah, and it’s interesting that you say that because that’s very topical for what I’m doing. So I’ve, I’ve pretty much in, in my day job when I’m not, uh, yammering into a microphone, but I’ve, I’ve pretty much moved away from straight finance work now and I’m working on digital transformation. So as, as you talk about the, uh, system and the processes, whenever I come in and we’re looking at automations and, and how we can, IM improve, uh, or layer in AI or whatever it analytics, whatever we’re doing, I always start on two levels. It’s the process that the people are following, what steps are they taking, and then how is the data flowing? And then, then you identify where the speed bumps are or where data stops flowing and you’ve got these silos and, and when you picture all that, that gives you the deeper view into the, the process. But same kind of thing if you’re doing investigation into understanding, you know, the, the root cause what whatever those primary levers are that are driving your financial, you have to go through and have that way of thinking to get down into that, into what the system is that that’s driving that.
Daniel Gardner:
Yeah, I think it’s, it’s the, the key difference for me, I think is that most people think of a process as A to B. It’s linear system is a matrix. It’s, it’s like a cloud network of, of how everything moves. And when you start properly mapping it out, you’ll discover that things are linked up that you never imagined they were, you can’t put that in swim lanes. You can’t just, you know, stick that in a timely little box and put a little note against it and say, this person does this at this point. You’ll find that, you know, things have feedback loops. Feedback loops are a crucial part of systems thinking and they are often the most confounding bit of the whole process when you’re trying to work out why things are going wrong. It’s ’cause somewhere buried in the back of it, you’ve got a positive feedback loop that just keeps reinforcing that behavior over and over again.
Glenn Hopper:
Yep, yep. We talked also, we were doing, we were doing a little comparison of books that inspired us and that we’re reading. And so I know you’re, you’re on this continuous improvement and you’re always, uh, uh, learning new things, but, uh, wondering if you could walk us through what are you working on improving or mastering right now? It’s, I don’t know if it’s could be something technical, strategic or even on the, on the personal side,
Daniel Gardner:
To be honest with you, right now I’m studying a bit of machine learning in my spare time I say a bit, I wouldn’t pretend to be anything like able to actually code up a machine learning tool, but trying to understand that, the basic principles of it so I can, I can get more familiar with that, but also trying to pick up more soft skills in relation to change management. Um, because there’s a, that’s 99% of what I’m trying to do in projects. And as much as I’ve tried to adapt my style, I need to probably still learn yet more to really get good at it, to the extent that I can get buy-in quite quickly, that things can roll forward in the right way. So I’m, yeah, that’s, that’s where my main focus is. I do a lot of personal writing as well, but that’s just more for my own, uh, amusement and to try and help me think through things. You know, I find that writing is a form of thinking. So,
Glenn Hopper:
Well, let me give you a warning. I I am a hundred percent there with you on the, uh, writing as a way to learn things. And that’s how I ended up accidentally writing two books, <laugh>. It’s just because you, you suddenly you have, you think, wow, I have 30,000 words right here. If I keep going, I’m gonna end up with a book. So, uh, yeah. But that’s, I, I’ve talked to a lot of people in fp and a, the guest that was on just last week, as a matter of fact, the same, she, she has a, a blog where she writes about fp and a and for her it was the same deal. We’re just work working through issues and somehow, and I love writing with, you know, obviously typing as much faster, but writing pen to paper going through. And a lot of times is the way I start just kind of free write mapping stuff out, drawing, you know, that way I’m drawing symbols and diagrams along the way with it. But, um, yeah, I’m right, I’m right there with you on that. I know we talked about Kahneman and, uh, NASEM Taleb is stuff that we both were inspired by. I love behavioral economics, <laugh>. Um, and then, uh, yeah, and then so Kahneman is is is great there, but I’m wondering because of your shared love there and focus has anything from behavioral economics or statistical thinking have either those influenced your approach to finance?
Daniel Gardner:
Yeah, indeed. I mean, I, I’ve got a a ton of different things that came out of, of, of behavioral economics, particularly. Most of the time we don’t really seek to be scientists in our work, but actually the, the innovations with AI should probably try and encourage that attitude in this, and that probably sounds a bit pompous, but lemme try and break it down. A scientist to me would be someone who objectively questions and tests, the hypothetical answers. Most companies I’ve worked in are more like cults than scientific labs. So, and, and operating a cult has some significant drawbacks. You know, you engage in groupthink naive realism. You think everybody thinks about the world the same way you do, you underestimate the chances of failure and overestimate the chances of success. But being a scientist means you don’t draw any conclusions that you’ve gathered some evidence.
It means applying some rigor to the tests and being honest about the limitations of your approach. So it requires quite a strong degree of humility. And, you know, we, we thought the future was gonna be flying cars and instead we got 140 characters. And this was ironically because we believed energy would be infinite. But it turns out information is well in an information rich environment, you’re gonna need a different discipline. You’ve gotta have to work on things like your cognitive nutrition. There’s nothing wrong in eating cheeseburgers, but they can’t form 90% of your diet. So in an information sense, what is your corporate diet made up of? I think for most corporations I’ve worked in, it’s been quite a highly processed ready meal. They want high salt, sugar and fat straight from the microwave. No, nobody wants the kale that says growth is gonna be 3% in line with inflation. You know, that’s, that’s not an answer we want to eat. Now AI is gonna make that problem a lot worse ’cause it can hallucinate the gap, so it will backfill all of our wants and desires with some kind of mental slop. And currently we have corporate cultures. I think we’ll probably encourage that. I mean, how many places have you worked where one individual could pretty much literally overwhelm the entire board by sheer force of personality alone? I’m hundred percent like every,
Glenn Hopper:
Yeah.
Daniel Gardner:
Yeah. I think that’s the danger with some of that stuff. And I think that’s what Carmen and EB have, have, have given me is a, a vision of cognitive bias that enables me to see some of that. There’s a beautiful thing that EB says in his latest book, skin Is Skin, uh, skin in the Game. The book itself is a bit verbose compared to his other ones. I wouldn’t necessarily recommend that specific book. It’s the last chapter in it. And he talks about a, a principle which just my whole head sort of exploded as a Reddit, which was ensemble probability and time probability. Have you ever heard of these?
Glenn Hopper:
No. No.
Daniel Gardner:
Okay. So Ensemble probability, a hundred men go to the casino and one of the ends up completely bankrupt. So it’s 1% chance of failure, time probability, a man goes to the casino a hundred times and eventually ends up bankrupt, 1% chance of failure. But straight away, you know, that those two 1% are not the same thing at all. One of them is a 1% that you totally understand seems reasonable. The one in a hundred men, if a hundred men get percent or one that fine. But if you just continually go to the casino until you end up bankrupt, that’s not a 1% chance of failure. That’s a hundred percent chance of failure. You are only gambling on when that will occur. Now you start looking at that in terms of the corporate system that you’re operating or the it that you are working on. It’s, it starts to change the picture a little bit, I think, about how people should think about the problems that are come down the track at them. But yeah, I, I I think that’s a useful insight ego.
Glenn Hopper:
Yeah. And I love where your head is around all this, because talking about the two things you’re looking at right now, uh, machine learning and change management, and one machine learning is the foundation understanding how machine learning, like you said, not that you’re trying to go become a, a developer or, or a mach, you know, a machine learning engineer, but understanding how it works and oh, guess what, by the way, since, uh, <laugh>, uh, all these, uh, LLMs can write really good code right now, you don’t have to become a Python developer. You can is if you know what you want the model to do, you just vibe code it and <laugh> I know, I know, I see the face you’re making. But, but that said, I, I wouldn’t want, uh, you know, my mother-in-law who <laugh> jumping in and trying to vibe code a, a, a machine learning algorithm.
However, if you have the understanding of, you know, this is what machine learning is doing, it’s it’s classification or clustering or, or predicting whatever it’s doing, and you know, the rule, the statistical rules around it, that gets you from using a, a finance calculator or maybe r you know, you never learned Python because you’re a a, a finance person, not a developer. But if you know the right guardrails to put on it and you know what you’re looking for, it’s not shaken up the magic eight ball. So generative AI is gonna open up possibility that if you understand the fundamentals of data science and you understand what’s you’re doing with machine learning, you know, you may not know which algorithm out of the gates you want, but you could ask the AI is this, should I use, you know, whatever this forecasting method or profit or whatever, and then build this tool that the barrier to entry would’ve had to been, uh, you know, not knowing Python back then, but, and then with change management, I deal with that every day.
Where right now around ai, there’s, there’s two competing fears and it’s, it’s crippling. Most companies I talk to, one is they’re getting pressure from the board, from investors, from their management, uh, to use ai. So there’s this sort of fear of missing out. Mm-hmm. And, and, and then the competing fear with that is, I don’t know when I can trust ai, I don’t understand what it’s doing. This seems like magic to me. It hallucinates. I can’t have it hallucinate numbers. So if you look at the two things you’re studying, I mean, that seems about as timely as it could be because change management, helping people get over that hurdle, <laugh>, and to, you know, to adopt it, but then understanding what’s happening over the hood lets you be an advocate for it. So I’m a hundred percent there with you, and I think you’re studying exactly the right things right
Daniel Gardner:
Now. <laugh>. Thank you. It’s a very exciting time to be alive. I think that’s, that’s, that’s for certainly, uh, you know, I think was it, was it the, the ancient curse that may you live in interesting times because you don’t realize they’re happening around you?
Glenn Hopper:
Yep. Yep. I wanna get back to the books, but since we touched on ai, and this is a area that I’m excited about, based on what you’ve seen and what you’ve, uh, talked about and just what’s out there, and we see it, it built into, uh, our SaaS tools that we’re using in our tech stack right now. Um, you know, AI is reshaping how we forecast, report, and analyze financial data. I know there’s a lot of confusion around what we can use and when we can use it and pulling data out of a system and, and when can you trust it and all that. But I’m sure you’ve seen, like, we all have the potential there. So what’s your perspective on AI and fp and a, I know a lot of companies are not quite there yet, but are you long-term optimistic, skeptical, or maybe some somewhere in the middle?
Daniel Gardner:
I’d say all technology is transformative. It moves in both directions depending on who you’re talking to. Some people are gonna win big out of ai, others are gonna feel like they lost even if their lo their lives actually end up improving in some way because of ai. The main problem I perceive right now is a kind of availability bias around the outcomes from the technology matched to this inflated set of expectations. So, you know, people can easily think of the Terminator or the Matrix, or, okay, if you wanna be a real nerd, how 9,000 from 2001
Glenn Hopper:
<laugh> top three favorite movies of all time,
Daniel Gardner:
Right? There you go. Yeah, <laugh>. But who thinks of de a sais who, who, who knows which Nobel Prize AI helped him to win in 2024? People, people don’t know this. And when I’ve said this to people before, like, AI won a Nobel Prize, I’m like, no, no, no, no. A human being used AI to win a Nobel Prize. And it was in chemistry and because it was called alpha fold. Look it up if you haven’t seen it. It’s incredible. You know, people used to spend their entire PhDs learning to break down the structure of One Protein. So we’d gone through things like 150,000 proteins, we worked out alpha fold, did something like 200 million in a couple of years. So we’ve now got the structures of nearly every protein in the universe, which means we can probably start developing cures for a whole range of diseases rapidly.
Now you think about that the, the negative expectations on AI are founded on an abundance of science fiction that, um, people have got out there, but <laugh> they’re not really paying attention to the positive side of it that’s genuinely coming down the track at them, and it’s already exists. So I think there’s, uh, I think there’s, I think that’s one of the, the issues I’d have with it around the, the skepticism that I see. That said, if you want, you know, for something on a more practical level for what people are doing in their workplace, you know, if the board’s pressuring you to put AI in, the first thing you need to say is, well, in order to automate, you have to have clean data and reliable processes that generate the same answer each time or otherwise. The generative AI is gonna just run off and create a whole bunch of noise and rubbish that actually makes all the problems worse.
You need to have a common semantic layer is what you technically kind of refer to it as. And it simply means that, you know, you can’t have one system that says Chicago, another one that says O’Hare, and another one that says, you know, somewhere else in Illinois, you can’t, you gotta have all three systems saying the same thing at once, using the same name. That’s the common semantic layer. And you have to have, have a regulated way of managing that. Now, that I would argue is why you’re gonna need data guardians and decision analysts. People forget that like 25 years ago, SEO engineers didn’t exist. Social media marketing experts did not exist. AI is gonna create human new, more human industries. So a data guardian is someone who’s watching over the process, looking at the guardrails, making sure that the AI is not, or the automation process is not generating a result.
We would not ordinarily expect anything that sits outside the boundaries of it. They’re not necessarily instantly scrubbing the result away, but they might investigate it. A decision analyst is taking a look at the process all the time, looking at that systemic effect of what’s going on. Copying from one spreadsheet to another is not a job for a human being. It’s a job for a machine. Talking to people and helping them understand is a job for a human. So the major challenge I would say about this new working environment is actually gonna be preventing pointless jobs arising. You know, don’t let chat GPT lead to the circular creation of documents that nobody reads. ’cause that’s the easiest thing that’s gonna happen with it. I think we should really let our embrace this as a means of freeing ourselves from our work. It’s the cumulative effects of this technology.
Imagine if you’ve got AI tied to a blockchain that proves the work was done. Now nobody needs to fill out the pointless forms. Would our lives not be better? I think they would. So I, I’m, I’m, I’m optimistic on the long term, but not in a, a kind of, um, you know, a a a fantasist way where it’s all just gonna be a marvelous utopia. There’s an awful lot of horrendous risks buried in ai, particularly around the proliferation of, of things that, you know, one person can create 10,000 bots very, very quickly. So there are significant risks and dangers that can’t just be regulated out of existence either. We’ll have to develop ever more powerful competing AI tools to deal with some of those things that impact us on a national security basis. But I don’t think you get rid of that by just, you know, saying, oh, it’s the Terminator, therefore everything’s bad. Yeah,
Glenn Hopper:
Yeah. And it becomes like cybersecurity where it’s a constant battle between the, the white hats and the black hats.
Daniel Gardner:
Yeah, exactly that. Yeah, exactly that. Yeah.
Glenn Hopper:
I did remember a meme or a cartoon or, or or whatever that I’ve, I’ve seen a bunch where, um, when you, when you were talking about, um, chat GPT generating documents that nobody reads. There’s the, on one side, it’s a split screen, but on one side it says, look, chat, GPT helped me turn these three bullet points into a, a, a full email. And then on the other side, someone’s saying, look, <laugh> chat, GPT helped me turn this full email into only three bullet points, <laugh> <laugh>. So it’s just, we’re generating all this AI swamp and, and nobody’s reading it. And you see it on, on social media and Substack, and anywhere where people are creating content, you can just tell people who, you know, were, were doing 140 character posts, are now suddenly writing novels every day on their LinkedIn full of m dashes and all the other, uh, <laugh> telltale signs of, of AI writing. And it’s, who’s reading this? <laugh>? Yeah. Okay. I I guess I’m gonna step off my soapbox now, but I’m, I’m right there with you, <laugh>. No fair place. Gosh, I could go all day. I didn’t, I kind of went in a fugue state as we started talking about ai. I could, I could have a whole podcast just on,
Daniel Gardner:
On that, but, uh, yes, it’s, it’s a deep rabbit hole. There’s a, there’s a lot of wars to explore. So, you know, <laugh>,
Glenn Hopper:
I guess in the interest of time, I need to bring this home with the last two questions that we, uh, that we ask everyone. Okay. So the first one, what is something that people might be surprised to learn about you? Something that they couldn’t get from your LinkedIn profile or your cv?
Daniel Gardner:
I’d say I’m a bit of a wizzywig. You know, what you see is what you get. But I studied karate. I’ve lived in 18 different addresses since I was 16. Um, I used to, I used to, um, write poetry, but now I read a blog on Substack, although I haven’t had the time to write on that recently. Uh, but I have been working on a book, um, <laugh>. So I’ve got a, I’ve got a variety of things. You know, sat in the back there. Um, yeah, I, I, I suppose that would be the, the main things I’d be willing to admit to on a podcast.
Glenn Hopper:
I <laugh>. Love it. Love it. I love that you followed, uh, karate and living in so many places. ’cause I immediately went to, uh, Kain from Kung fu just traveling the earth <laugh> the American TV show or the David Carradine with me. Do
Daniel Gardner:
You do me far too greater service? Uh, no. I was, I was, I was never wandering around saving the entire town from the bad guy who was, you know, running the local mine. That was never, no, I, I never had that kind of skill. <laugh>.
Glenn Hopper:
Alright, so finally to bring it all home, what is your favorite Excel function and why That’s gonna be easy? Python. So it’s funny, I talked to very few people who are actually using Python in, in Excel. So you are, uh, using that today? Yeah. What are, what are some use cases, areas where you’re using it?
Daniel Gardner:
Anything you’ve been doing with macros? Replace it with Python. If, if you are having to do stuff on power pivot, look at Python. There are a stack of use cases for things. It’s, it’s mostly about overcoming that initial barrier of feeling like an idiot and, and embracing the fact that actually, well if you go to like, I wouldn’t use chat GPT, I’d use Claude by Anthropic. It’s better at doing the coding side only marginally so, but enough to be worth doing. And it’s just, it’s approaching it from the side of something like, um, approach using AI like you’re talking to a colleague. So, you know, how do you want me to work with you? Is a great question to ask AI because it’ll actually give you the steps to then use in the prompts to try and ensure the prompts aren’t full of coding errors.
What sort of things should I look out for that you are likely to miscode this kinds of stuff? You know, you can, you can start asking it to break each step down. You can look at each individual piece of code and ask it to explain what that code means. So you can use this as a translation tool to scale yourself up to the point that you feel far more confident about it and you can just start banging the code into it. Given that you can then put that native into Excel, all of a sudden you can start doing all kinds of data transformation that previously would’ve looked like complete wizardry.
Glenn Hopper:
Yeah. <laugh>.
Daniel Gardner:
Yep. Excel isn’t really the right space to do that in, I would argue, but it’s a very effective tool in being so flexible and it’s obviously the one that everybody knows. Um, but I think it’s particularly key that people start to make that leap because at the end of the day, Excel’s only got a million rows in it, which sounds like a lot until you start to really understand what, you know, the third age of data is gonna bring us. When you’ve got the internet of things and you are getting a million rows of results every single day or every hour, when you start getting that kind of volume of information, you cannot do it through Excel. Some if it’s not going to work. So gotta get people more comfortable with that kind of stuff. I’m afraid
Glenn Hopper:
We need to log this. You are actually the first guest on the podcast who has said Python even though, uh, you know, I I think a lot of people are are wanna use it and dabbling, but there’s some intimidation. But I think, like you said, code, uh, being wr you know, gen AI helping you write the code, I think we’re gonna start seeing it a lot more. And I wonder if, if Microsoft isn’t eventually gonna get there where we, I keep saying this, but where copilot becomes Clippy 2.0 and you just, you know, you don’t know what you’re asking you, you just, you don’t know whether you’re writing Python or a nested if statement or whatever, but you’re telling clip, this is what I want to do with this data and it’s doing it all under the hood because there’s, there’s tools out there right now that are, you know, add-ins to, uh, um, you see more come every week that are either, you know, web-based or they’re add-ins to Excel that are using, um, LLMs to help you do things in Excel. So I think that’s where it’s headed, but to, to your point right now, it’s not there. So being able to get that additional functionality of Excel is pretty huge.
Daniel Gardner:
Yeah, it’s a funny thing actually. I I don’t really understand Microsoft’s direction on this because like copilot is just rubbish compared to any of the other ais I’ve used. And I, and I, it’s one of those things where this seems to be an open goal for Microsoft. They’ve got investments in open ai, they’ve got investments in a number 15 billion. Yeah. So I, I’m sitting there going, why aren’t you just putting this proper thing to work? I don’t, I dunno what it is. Maybe, maybe it’s the fact that they’ve got the investment, they want to keep it going and they don’t want to, you know, shoot the goose and the golden egg. But I, yeah, copilot seems to be very so narrowly focused, so regulated that it actually isn’t useful, which is weird, but
Glenn Hopper:
Yeah. Yeah. Alright, before we let you go, you mentioned your substack and we’ll we’ll put these in the show notes too, but how can our listeners, uh, follow you or stay connected or, or if you wanna mention your substack, uh, we could do that too. <laugh>, please,
Daniel Gardner:
By all means follow me on LinkedIn and, and at me. I, I don’t have any real social media. I’ve stayed away from all of that, so it’s just, um, Substack and LinkedIn if you want to get a hold of me. Um, I’m very interested and keen to talk to people. Happy to chat about a range of different things. So I’d be, I’d be happy to hear from any and all with some ideas about, you know, what’s coming next and what they wanna do with it. And it’s been a pleasure, Glen. So thank you. Thank you for the invitation. Likewise Daniel. Thanks for
Glenn Hopper:
Coming on. Thank you.
