If you would like to earn CPE credit for listening to the show, visit earmarkcpe.com.
Download the app, take a short quiz, and get your CPE certificate.
Finally, if you enjoy listening listening to fpna today please go to your podcast platform of choice click the subscribe button and leave a rating and review of the show and now on to the show From data rails,
Welcome to FP&A Today. I'm your host, Glenn Hopper.
Today's guest is Daniel Gardner, operational finance business partner at FirstBus, 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 DTC 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&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.
Hi, Glenn. Thank you very much for having me on.
It was a very kind introduction. Yeah, 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.
It sometimes feels like a billion pounds of weight, but there we go.
Yeah. Well, I did. I loved our, our. conversation we had before the show 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 eventually made your way to corporate FP&A.
Can you kind of walk us through that journey?
And maybe 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?
Yeah, sure. They definitely are. Um, I started relatively late.
I was in my late twenties before I even really began the study process.
I had to work without a mentor or support.
I had to borrow a bit of money to study, which is unusual for a lot of accountants in the UK normally go through a kind of a corporate route associated with an audit firm and they sponsor you effectively.
I didn't do any of that kind of stuff. A lot of the mass was quite new to me as well, especially around things like the options calculations, black skulls, formulas and things.
I had to build back up my understanding of algebra and calculus because I've been quite a number of years since I've done any math.
Right in the middle of it thinking, I'll get started.
No, no, I've got to 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 to do that process.
Frontline retail actually has always held me in good stead because as much as that's 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 exposing, which might have been a bit opaque just from the pure finance side.
But my journeys in environmental engineering as well, you know, that was focused around two particular industries, which was sewage and cremation.
So I have sit up to my thighs in human waste and been covered head to toe in ash from blowback from a chimney.
This was both before you got into FP&A, right?
Oh, yeah, yeah. Environmental engineering was where I started to study accounting.
I was helping out on site and doing a bunch of things as well as this. 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 strike me as dangerous as they might have.
I have done if I had just gone straight from, you know, I don't know, PwC or Ernst into corporate life.
So gave me a good grounding. Yeah. I love what you said about retail because It's so easy, especially people, if you just go the traditional route, whether in the US, you know, whether it's a you know, undergrad in accounting or whatever, and then, and or an MBA program, it's very easy to 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 And I think 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 in 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 sales, work with marketing, work with ops.
Yeah, 100%, 100% agree with you on that.
I mean, this is something that I've said to a number of people that, 99% of the people you're ever going to 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 a lot of finance business partners, particularly now that roles become so much more ubiquitous for us 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 unglue 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, it's a very large business, public 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 you have to really embed that to understand it.
Yeah. Another thing when we were talking before the show and you corrected me, and I love this because 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, I didn't study accounting or business at university, which is a clear distinction because I was picturing you just naturally just intuiting everything about finance and accounting.
But where I was getting with that, though, is your background, kind of like your work background, before coming into FP&A, it's not the traditional route into the job you studied international relations and philosophy 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 shaped your approach to finance and maybe especially in terms of decision-making or critical thinking or even long-term planning.
Yeah, 100% all of the above. 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 my 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 going to 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 I.
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 ask you to keep your thinking brief.
You should always be taking words out of the sentence that aren't required, which definitely sticks with me when I see PowerPoints with over 100 slides on the deck.
And truthfully, we talk now, I mean, a million years ago when I started out in FP&A, I didn't think it was called FP&A back then.
I was just the finance guy. But it was really about where you stood out was how great your models were 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 worthless.
Also, the inquisitive part of a philosophy major is interesting because That's what we do in FP&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 the right questions and keep digging out.
It seems like that would have to come through. yeah it really does i mean i've i like the commercial roles i i never was an auditor so You know, I didn't have the benefit of that experience.
So I want to acknowledge that very clearly that I'm not a tax person or the guy you ask about IFRS.
I am very good with systems thinking I can see the whole picture and laser in on one detail but I don't want to give people a false impression that I could do a lot of the very tight technical work that some of my colleagues are exceptional at.
And that's I think that's a key part of it is being able to know if you can emphasize your strengths, then you can really contribute to the story in the right way.
Yeah. And that's why there's an FP&A team and not just a single individual on there.
When you have the team with the different strengths and everything.
Yeah. I came up through FP&A too, and my first CFO role, if I didn't have a good controller, I'd have been sunk.
It was a mess for me. So I 100% agree with you there.
Well, digging into 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 current role?
Yeah, so I helped build up cloud forecasting systems for L'Oreal and Hunter.
We did Hunter system in like six weeks. The L'Oreal one was a kind of a bespoke one they developed themselves, but Hunter's was one I actually helped to get designed. right from the ground up, including all the calculations.
And that was some of the experience I was able to take across to FIRST.
Doing it in six weeks was not originally the plan, 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 It definitely wasn't me.
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 area I actually really found particularly enjoyable was kind of the challenging legal work I'm very much a dog with a bone, a bit belligerent.
So actually it helps greatly when I've got sort of high stakes conversations with people 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 branch and be able to say no that was not what was agreed at the point and Those kinds of things.
That kind of work 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 know already how to start the process to ensure we don't end up in that mess in the first place.
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.
We talked about at Hunter Boot, you led initiatives ranging from launching the new e-commerce channels to overhauling merchant fees and customer service.
That's broader than obviously your typical finance, but I'm wondering, From that experience, what does that teach you about making finance operational or what is something that you could share with our listeners about 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.
No, no, absolutely not. It's always got to be commercially and operationally leg, but it's, it's interesting that so often we put agility and control as some kind of two opposite ends of a spectrum.
And actually, I think there's an argument for saying that not only are they not mutually exclusive, if you're going to succeed, you have to be able to do both simultaneously.
I looked it up afterwards when I quoted it to you originally that it was actually the Prussian general Helmuth von Molke who said no plan survives contact with the enemy.
So we should expect every single forecast to be wrong the moment it's issued.
But it's not the fact that the forecast is wrong.
It's the degree of the wrongness that matters. you've got to 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 kind of control. i think where it boils down for me was too often we end up focusing on lagging indicators the P&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 than to argue about an invoice that's unexpected and takes you well over budget.
It's too late usually by that point. Yeah, it's so funny how that, I mean, that's just the default though, isn't it?
Just to go to the P&L and say, well, this is our trend and let's just...
And we've been creeping up 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. yeah 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 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 your way ahead of where you need to be in terms of preventing the ugly looking P&L coming out the other side of it.
Yeah, and we talked also about, and you sort of already answered this, but I'm gonna push a little deeper.
Forecasting, identifying those levers, looking at the leading indicators, that is the crystal ball of FP&A.
So I'm wondering if you have and you've probably pretty well hit on it there, but what your philosophy is on forecasting 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 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? there's the one billion dollar question the one billion pound question the big heavy question yeah and that definitely does weigh a billion pounds yeah 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 have to keep doing that Indeed, if you assume everybody heard something different and totally missed the point, you're probably in the right place about how people are receiving the information they've already got and are acting on it.
I have a forecast list that runs 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 ECOM about how they're really doing the digital marketing.
I also think there's something worth saying at this point, particularly in this modern era with us having a conversation over over the web that 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 a real challenge for a lot of managers, I think, to transition to being able to manage at a distance.
Cause 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 that means if you're going to retain it and look after it, you've got to make sure that people feel like they're included in things, even if they're typically on the outside of the end of a team's court. things companies do wrong I should be very clear here I am not talking about any specific company what I'm about to say because I'm going to give you some generic feedback from the last 20 years of me doing this because 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, particularly in previous companies, no joined up thinking where there'd be this We're going to have a sales increase of 20% and no attendant costs or requirements to scale customer service or the back office.
Okay. So how are you servicing this 20%?
You know, I'm thinking particularly of one organization here, allowing ego to dominate the decision-making and, If you want to work on gut feel, I don't think that's wrong, but you should at least have the courage to admit that.
Costs are hard and sales are soft. It's very easy to spend the money up front, and we typically do that in advance of making the sales in the first place.
You've got to 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 want to 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.
I have to say 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 some control issues that are going to start emerging, I think, that are going to really cause company some real headaches and they're not yet conscious of it because it's not yet hit them how real that problem will be. yeah and that's always a conundrum because you you put together your annual plan and the company has we know we've this is our Kiger 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 And then you have to reverse engineer to the goal, which is fine.
That's part of the job. I mean, is to take that strategy and figuring out what it costs and figuring out all the components of it.
But then you're already forced into 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.
In my experience, it always ends up 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 going to drive the operational headcount.
We're going to need to support the sales or whatever the case is, increased inventory, or it's just a lot can get missed when you have directionally competing drivers.
It is probably always gonna be that way that you get the meld, but I think that's the challenge is that there's a feeling that I've often come across that people, the senior team, The board wants the ownership and the accountability to be vested with the frontline staff because they want them to 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 takes far more of a There's an interesting sort of nature of belief you have to get into about how we can create some faith in the number and in the processes that you're asking people to follow.
I think that's the big challenge. Yeah, I 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 um when uh 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 going to 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 how, you know, it's like if you're running your budget based on, uh, you know, if your tax accountant is doing FP&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 that the companies I work with since then don't run that way.
A couple of items you hit on. There was a lot there, but two that stood out to me.
And the first one just... if you could dive a little deeper on it.
So you talked about your 500 item list 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 want to do.
And then it sort of gets stacked, ranked and prioritized based on what is feasible and you know you got your must-have nice to have whatever your stretch goal kind of things you know And then just one comment that we're removed from it, but I do want to mention this because you did bring up remote workers and This is a problem that I have to remind myself on remote workers, when 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, as you throw work, it's very easy to just throw that prompt across the board and let the person do it.
And I'm I know there's a mix between people going back to the office and being remote, but I guess this is my public service announcement Remember that person on the other side of the screen is a human being and not chat GPT that you're just lobbing prompts into.
Yeah, I mean, the funny thing is, 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, you can do coaching over teams.
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.
Most people I've found do not use things like the whiteboard function in 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, not in the kind of the fake icebreaker way in a meeting, but actually to genuinely engage and think about problems together.
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 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 around the remote workers that way, all getting to enjoy something like an office life, even though they're all actually totally physically separate.
On that note, coaching, again, 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 machine that is generating output and insights to remember the people who helped you get there along the way and being able to pay that back and bring your team up.
And listeners of the show are going out and hoping to get insights from all the guests we have on.
And so one of the questions I always like to ask, it's sort of a general broad coaching question out there, but if you could give just one piece of advice to finance team or it could be individual level or team level that are trying to level up their FP&A capability Wondering what that would be, and I like asking you of this because I know you're very skilled 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?
I think one of the, so if there's an area I'd say that perhaps FP&A could do with looking at more, it's systems thinking.
And I don't necessarily say that to just give an academic recommendation.
I mean that 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.
If you start thinking about it in those terms, because I think a lot of the time when I've seen FP&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 in 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 it doesn't even please the person doing it so 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, because 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 to 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 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 FP&A Today is brought to you by Data Rails, the world's number one FP&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 held 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 at datarails.com.
So for our listeners who aren't familiar with it, let's scroll down into systems thinking because whenever I hear that word, I picture either a Rube Goldberg machine or those overly complicated YouTube videos where somebody knocks over some dominoes and they run and 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 do you mean by that?
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 just trying to produce a map of the terrain.
So too often people mistake the map for the terrain.
You get those beautiful process diagrams, and you know this as working in America with the Sox.
You're going to have this beautiful, you know, 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 so 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 going to entertain certain options just because of the kind of person they are.
So if you start mapping things out that way and thinking about the different states that exist out there, and the ways that information and behavior transitions between all those things.
This is a far more fluid explanation of system dynamics than I think any act than any true system dynamicist would ever allow for.
There'll probably be people spitting out their cornflakes listening to this.
I'm trying to make it something practicable and workable 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 want to make it something that lives 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 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.
Yeah, and it's interesting that you say that because that's very topical for what I'm doing.
So I've pretty much in my day job when I'm not yammering into a microphone, but i've pretty much moved away from straight finance work now and i'm working on digital transformation so As you talk about the system and the processes, whenever I come in and we're looking at automations and how we can improve or layer in AI or whatever, analytics, whatever we're doing, I always start on two levels.
It's the process that the people are experiencing following?
What steps are they taking? And then how is the data flowing?
Then you identify where the speed bumps are or where data stops flowing and you've got these silos.
When you picture all that, that gives you the deeper view into the process.
But same kind of thing if you're doing investigation into understanding root cause, 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 what the system is that's driving that.
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 like a cloud network of how everything moves.
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 tiny little box and put a little note against it and say, this person does this at this point. you'll find that 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 because somewhere buried in the back of it, you've got a positive feedback loop that just keeps reinforcing that behavior over and over again.
Well, we talked also, we were doing a little comparison of books that inspired us and that we're reading.
And so I know you're on this continuous improvement and you're always learning new things, but wondering if you could walk us through, what are you working on improving or mastering right now?
I don't know if it could be something technical, strategic, or even on the personal side.
Uh, I mean, 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 the basic principles of it so I can get more familiar with that. but also trying to pick up more soft skills in relation to change management 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 yeah, that's where my main focus is. I do a lot of personal writing as well, but that's just more for my own amusement and to try and help me think through things.
You know, I find that writing is a form of thinking so.
Well, let me give you a warning. I am 100% there with you on the writing as a way to learn things.
And that's how I ended up accidentally writing two books.
It's just because suddenly you have you think, wow, I have 30,000 words right here.
If I keep going, I'm going to end up with a book.
So, yeah, but that's I've talked to a lot of people in FP&A.
The guest that was on just last week, as a matter of fact, the same.
She has a blog where she writes about FP&A and For her, it was the same deal.
We're just working through issues and somehow, and I love writing with obviously typing is much faster, but writing pen to paper, going through 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 there with you on that.
And you did – I know we talked about Kahneman and Nassim Taleb, stuff that we both – were inspired by, I love behavioral economics, so Kahneman is great there.
But I'm wondering, because of your shared love there and focus, Is anything from behavioral economics or statistical thinking of either of those influenced your approach to finance?
Yeah, indeed. I mean, I've got a ton of different things that came out of behavioral economics, particularly.
Most of the time, we don't really seek to be scientists in our work.
But actually, the innovations with AI should probably try and encourage that attitude in us.
And that probably sounds a bit pompous, but let me try and break it down.
So 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.
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 thought the future was going to 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 going to need a different discipline.
You've got to have to work on things like your cognitive nutrition.
You know, 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.
Nobody wants the kale that says growth is going to be 3% in line with inflation.
That's not an answer we want to eat. Now, AI is going to make that problem a lot worse because 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 think that's the danger with some of that stuff.
And I think that's what Kahneman and Taleb have given me is a vision of cognitive bias that enables me to see some of that.
There's a beautiful thing that Taleb says in his latest book, 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 principle which just my whole head sort of exploded as I read it, which was ensemble probability and time probability.
You've heard of these? No, no. Okay. So, so ensemble probability, a hundred men go to the casino and one of them ends up completely bankrupt.
So that's 1% chance of failure. Time probability.
A man goes to the casino 100 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 and seems reasonable, the one in 100 men.
If 100 men go to the center of one, that's 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 100% chance of failure. You're 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're working on.
It starts to change the picture a little bit, I think, about how people should think about the problems that come down the track at them.
But, yeah, I think that's a useful insight he gave.
Yeah, and I love where your head is around all this because talking about the two things you're looking at right now, 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 a machine learning engineer, but understanding how it works.
Guess what, by the way, since all these LLMs can write really good code right now, You don't have to become a Python developer.
If you know what you want the model to do, you just vibe coded it.
I know I see the face you're making. But that said, I wouldn't want my mother-in-law, who's jumping in and trying to vibe code a machine learning algorithm –
However, if you have the understanding of this is what machine learning is doing, it's classification or clustering or predicting whatever it's doing and you know the statistical rules around it that gets you from using a finance calculator or maybe are you know you never learned Python because you're 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 shaking up the magic eight ball.
Generative AI is going to open up possibility that if you understand the fundamentals of data science and you understand what's you're doing with machine learning, you may not know which algorithm out of the gates you want, but you could ask the AI.
Should I use whatever this forecasting method or profit or whatever, and then build this tool that the barrier to entry would have had to not knowing Python back then.
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 crippling most companies I talk to.
One, is they're getting pressure from the board, from investors, from their management to use AI.
So there's this sort of fear of missing out.
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 to adopt it, but then understanding what's happening over the hood lets you be an advocate for it.
I'm 100% there with you, and I think you're studying exactly the right things right now.
Thank you. Yeah, no, it's a very exciting time to be alive, I think.
That's for certain the... Was it the ancient curse that made you live in interesting times because you don't realize they're happening around you?
Yeah. And I want to get back to the books, but since we touched on AI and this is an area that I'm excited about, based on what you've seen and what you've talked about and just what what's out there and we see it built into our SaaS tools that we're using in our tech stack right now.
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 when can you trust it and all that.
I'm sure you've seen, like we all have, the potential there.
So what's your perspective on AI and FP&A?
I know a lot of companies are not quite there yet, but are you long-term optimistic, skeptical, or maybe somewhere in the middle?
I'd say that all technology is transformative.
It moves in both directions, depending on who you're talking to.
Some people are going to win big out of AI.
Others are going to feel like they lost, even if 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 want to be a real nerd, HAL 9000 from 2001.
Top three favorite movies of all time. Well, there you go.
But who thinks of Denis Hassabis? Who knows which Nobel Prize AI helped him to win in 2024?
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. 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.
When you think about that, the negative expectations on AI are founded on an abundance of science fiction that people have got out there, but they're not really paying attention to the positive side of it that's genuinely coming down the track at them, and it already exists.
So I think that's one of the issues I'd have with it around the skepticism that I see.
That said, for something on a more practical level for what people are doing in their workplace, 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 going to 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 can't have one system that says Chicago another one that says O'Hare, and another one that says somewhere else in Illinois.
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 a regulated way of managing that.
Now that... I would argue is why you're going to 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 going to create 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.
So there's 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 going to be preventing pointless jobs arising.
You know, don't let chat GPT lead to the circular creation of documents that nobody reads because that's the easiest thing that's going to happen with it.
You know, I think we should really 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'm optimistic on the long-term, but not in a... a kind of fantasist way where it's all just going to be a marvelous utopia.
There's an awful lot of horrendous risks buried in AI, particularly around 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.
I don't think you get rid of that by just saying, Oh, it's the terminator.
Therefore, everything's bad. Yeah. Yeah.
And it becomes like cybersecurity where it's a constant battle between the white hats and the black hats.
Exactly that. Yeah, exactly that. Yeah. I did remember a meme or a cartoon or whatever that I've seen a bunch when you were talking about ChatGPT generating documents that nobody reads.
On one side, it's split screen, but on one side it says, look, ChatGPT helped me turn these three bullet points into a full email.
Then on the other side, someone's saying, look, ChatGPT helped me turn this full email into only three bullet points.
So it's just we're generating all this AI slop and nobody's reading it.
Yeah, there's a definite danger of that.
You see it on social media and Substack and anywhere where people are creating content. tell people who were doing 140 character posts are now suddenly writing novels every day on their LinkedIn full of em dashes and all the other telltale signs of AI writing and it's, who's reading this?
Okay, I guess I'm going to step off my soapbox now, but I'm right there with you.
Gosh, I could go all day. I kind of went in a fugue state as we started talking about AI.
I could have a whole podcast just on that.
Yeah, there's a lot of Warrens to explore, so, you know.
I guess in the interest of time, I need to bring this home with the last two questions that we ask everyone.
So the first one is, 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?
I mean, I'd say I'm a bit of a whizzy wig.
You know, what you see is what you get. But I studied karate.
I've lived in 18 different addresses since I was 16.
I used to write poetry, but now I rather blog on Substack, although I haven't had the time to write on that recently, but I have been working on a book.
So I've got a variety of things, you know, sat in the back there.
Yeah, I suppose that would be the main things I'd be willing to admit to on a podcast.
Love it. Love it. And I love that you followed karate and living in so many places, because I immediately went to Kane from Kung Fu, just traveling the earth. say the American TV show or the David Carradine party greater service.
No, I was never wandering around saving the entire town from the bad guy who's, you know, running the local mine that was never you know i never had that kind of skill All right.
So finally, to bring it all home, what is your favorite Excel function and why?
That's going to be easy, Python. So it's funny, I talked to very few people who are actually using Python in Excel.
So you are using that today. What are some use cases? where you're using it.
Anything you've been doing with macros, replace it with Python.
If you're having to do stuff on Power Pivot, look at Python. there are a stack of use cases for things.
It's mostly about overcoming that initial barrier of feeling like an idiot, and embracing the fact that actually, well, I wouldn't use ChatGPT.
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 approaching it from the side of something like approach using AI like you're talking to a colleague.
So how do you want me to work with you is a great question to ask AI, because it will actually give you the steps to then use in the prompt 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 kind of stuff. 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.
You can use this as a translation tool to scale yourself up to the point that you feel far more confident about it.
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 have looked like complete wizardry.
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.
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 the third age of data is going to bring us when you've got the Internet of Things, and you're 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 you've got to get a bit more comfortable with that kind of stuff, I'm afraid.
Yeah. So I think we need to log this. You are actually the first guest on the podcast who has said Python even though I think a lot of people want to use it in dabbling, but there's some intimidation.
But I think, like you said, GenAI helping you write the code, I think we're going to start seeing it a lot more.
And I wonder if Microsoft isn't eventually going to get there where...
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 just... You don't know whether you're writing Python or a nested if statement or whatever, but you're telling clearly, this is what I want to do with this data. and it's doing it all under the hood.
Because there's tools out there right now that are add-ins to, you see more come every week that are either web-based or they're add-ins to Excel that are using LLMs to help you do things in Excel.
So I think that's where it's headed. But to your point right now, it's not there.
So being able to get that additional functionality of Excel is pretty huge.
Yeah, it's a funny thing, actually. 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.
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 of these other organizations.
I'm sitting there going, why aren't you just putting this proper thing to work?
I don't know what it is 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 Yeah.
Copilot seems to be very, so narrowly focused, so regulated that It actually isn't useful, which is weird.
But there we go. Yeah. All right. Before we let you go, you mentioned your sub stack.
How can our, and we'll put these in the show notes too, but how can our listeners follow you or stay connected?
Or if you want to mention your Substack, we can do that too. by all means, follow me on LinkedIn and at me.
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 hold of me 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 want to do with it it's been a pleasure glenn so thank you thank you for the invitation likewise daniel thanks for coming on thank you
