Skip to content

Live Events Industry - Cash Forecasting Discovery - 2025-04-01

Metadata

  • Date: 2025-04-01
  • Company: Live Events/Festivals (PE-backed group)
  • External Participants: Dom Dubois (Treasury)
  • Palm Participants: Emma Sjöström
  • Type: Expert Interview
  • Domain Areas: Cash forecasting, Cash visibility, Connectivity, IC activity
  • Recording: None

Summary

Context

Expert interview with Dom Dubois from a PE-backed live events and festivals group. Discussion focused on the unique challenges of cash forecasting in an industry with highly unpredictable revenue streams (ticket sales dependent on lineup quality, weather, etc.). Met at the ACT conference.

Key Discussion Points

  • Cash forecasting is extremely difficult in live events due to unpredictable ticket sales
  • Using external signals (Instagram engagement, post-event surveys) to inform forecasts
  • 23 operating businesses submit Excel templates for indirect forecasting
  • Currently implementing TIS Payments (payment hub) + Integrity TMS
  • Building internal cash pool to avoid expensive external debt facilities
  • Bank rationalization journey: from 40+ banks down to ~20, aiming for global bank consolidation

Pain Points

  • Historical data not useful for forecasting - "the nature of live events and festivals industry, it is really hard to predict how well tickets are going to sell"
  • No cash visibility across entities - "we had no way of seeing live balances on accounts"
  • Relying on opco-reported balances with no verification - "theoretically, at the moment, they could say they've got one million because they don't want us to take"
  • Fragmented data across ERPs - "many different teams on different ERPs, there's no unified ERP"
  • Half day spent collating forecasts - "We've got 23 different operating businesses and they send back templates... you spend half a day collating all the data"
  • Excel template errors - "they make certain changes thinking they're being helpful and then all your lookups go wrong"
  • Expensive debt facilities for short-term funding - Using central debt facility to fund opcos during pre-festival cash shortfalls

Feature Requests & Needs

  • Automated variance analysis - "automated variance analysis would be really useful"
  • Dynamic forecasting that adjusts with ticket sales - "if ticket sales go better, it can automatically move the dial"
  • Proactive alerts on variances - "it can give us like, think where we need to look at... here's major differences, these are the things you need to investigate"
  • Historical balance data for trend analysis
  • Ability to export to Excel alongside tool capabilities

Jobs & Desired Outcomes

Job: Forecast cash across decentralized festival group with unpredictable revenue

Desired Outcomes: - Minimize the time spent collating forecasts from multiple operating businesses - Increase the ability to incorporate external signals (surveys, social engagement) into forecasts - Reduce reliance on fragmented, manually reported balance data

Job: Optimize working capital across festival group with seasonal cash patterns

Desired Outcomes: - Minimize the use of expensive external debt facilities for short-term funding - Increase the efficiency of internal cash pooling across entities - Reduce the time between identifying surplus cash and deploying it to entities in deficit

Job: Gain visibility into actual cash positions across 20+ banks

Desired Outcomes: - Minimize reliance on opco self-reported balances - Increase the frequency of balance visibility (daily vs monthly) - Reduce the risk of opcos misrepresenting their cash position

Domain Insights

Festival Cash Cycle: - Positive working capital cycle: cash in (ticket sales) months before cash out (festival costs) - Large cash buffers sit on operational accounts for months - Sharp cash dip just before festival (site fees, artist payments, production) - Short-term cash deficit during/after festival until food & beverage revenue comes in - Then surplus again until next cycle

External Signals for Forecasting: - Post-event surveys asking "who would you want to see next year" - Instagram engagement on lineup announcements - Historical ticket sales timing and volume - Weather forecasts (affects on-site revenue like beer sales)

TIS Payments + Integrity TMS Architecture: - TIS Payments = payment hub for bank connectivity (not TMS) - Integrity TMS = logic layer for target balancing and cash pooling - Connected via SFTP - Target balance sweeping: pull excess cash centrally, push down for payment runs - Internal transfer pricing for intercompany loans

Why Not Traditional Cash Pool: - Many opcos have partnership requirements (e.g., festival sponsor is a bank) - Local banking needed for cash deposits (physical cash from on-site sales) - KYC processes too slow for full bank migration - TMS-based internal pooling as interim solution

Action Items

  • [ ] Emma to connect Dom with Palm founder re: internal cash pools
  • [ ] Follow up end of year to hear how TIS + Integrity implementation went
  • [ ] Share relevant insights from other treasurer conversations

Notable Quotes

"The nature of live events and the festivals industry, it is really hard to predict how well tickets are going to sell... sometimes one festival that goes mad one year and then just no foot for the next year." - Dom Dubois

"At the moment, we have really fragmented data... we had no way of seeing live balances on accounts. So previously, we just relying on these forecasts and them giving us the correct balance, we have nothing to tie it back to." - Dom Dubois

"We were relying on a central debt facility to fund each opco when they had those short term needs. So what we're trying to do is by connecting all the bank accounts... create a sort of in-house cash pool." - Dom Dubois

"You spend half a day collating all the data to get to something that should be done as soon as you cook a submit button." - Dom Dubois

"The revenue side of it, whilst it can be sort of spread out of the machine learning, it really needs that human touch to really get the vibe of like how things are looking." - Dom Dubois

Full Transcript

Introduction

Date: 01/04/2025, 14:05

00:00 Emma Sjöström: Oh, I bet.

00:02 Dom Dubois: Yeah. Yeah.

00:03 Emma Sjöström: I, So yeah, fun. Let's talk some cash forecasting. Then from one thing to another

00:10 Dom Dubois: Okay.

00:12 Emma Sjöström: Um I have like quite a structure format so I'll be asking you questions and they might be a bit repetitive but it's just to really try and make sure that I understand correctly or even giving you a chance to maybe. Hey, Think about it deeper or pivot or so, I'm just gonna start asking questions and Feel free to ask Clarifications or anything like that.

00:00 :

00:38 Dom Dubois: Okay.

00:40 Emma Sjöström: Cool. And my first question a bit broad would be just like Would you mind walking me through a bit? How you currently approach cash, forecasting?

00:49 Dom Dubois: Yeah, yeah sure. So I think I mentioned to you at the act, it's quite difficult for us to cash, forecast, using historical data.

01:00 Emma Sjöström: Yeah.

01:00 Dom Dubois: So, we have tried that previously. We had a whole group position and we had the balances which were reported by entities and you can get a rough trend, but the nature of live events and and the festivals industry, it is really hard to predict how well tickets are going to sell.

01:21 Dom Dubois: And unless you have an event like a Glastonbury, or one of our ones, which is whack in, which is just has that loyal fan base where pretty much sells out, every time you have sometimes, one

01:30 Emma Sjöström: Right.

01:34 Dom Dubois: festival that goes. Mad one year and then just no foot for the next year. So it's super difficult for us to to really forecast, how we see the cash inflows outflows tend to be relatively straightforward because you've got the same. Same sort of thing. Every, every year you can kind of forecast how much your ass is going to cost.

01:59 Dom Dubois: You can forecast your General sort of maintenance cost you your production costs, etc. But then when it comes to actual On the day performances as well. We also find that quite difficult to to really forecast because sometimes you have events where It's been a scorcherer by the weather. So beer sales, just go through the roof.

02:23 Dom Dubois: And then sometimes it's just he just it just doesn't hit how you expect it to do. So what we do at the moment for forecasting is we send out a Template to each of the different operating businesses. So the operating businesses might have multiple festivals underneath them.

02:44 Emma Sjöström: All right.

02:45 Dom Dubois: We fill out based off their balance sheet positions. So it's an indirect forecast and they use the budget to essentially map out where they've recognized revenue. So for ticket sales there, we'll see have deferred revenues and then they recognize the revenue on the event, but we'll see have the cash flow coming in earlier hence why it's a crude is deferred revenue and it's at the moment.

03:14 Dom Dubois: It is a bit of a, an art finger in the air exercise. So it's a lot of assumptions being made. so, we use sort of based off last year, When the ticket sold, how well they sold, but then adding some own assumptions to it, just to see how well it's being received in in sort of, until the market.

03:41 Dom Dubois: So you look at your Instagram post, etc and see what the responses are. For example, one of the festivals Field Day the lineups hit quite well. So there were forecasting that it's going to be a sellout. Whereas last year, we struggled to To sell out and we didn't even sell out to even on the event.

04:05 Dom Dubois: So quite a few tickets available so it's it is very much A. Guest guesstimation, I would say, it's it's using the budgeted numbers, which

04:05 Emma Sjöström: I think that's super interesting. Yeah, I haven't talked to that many treasures

04:09 Dom Dubois: are stretch budget in themselves. So we really try and stretch bobcos and just really allocating where we think for the revenue at least where we think that's going to hit hit the bank.

04:25 Emma Sjöström: who use these kind of like external signals, signals to that extent. Right, I remember you told me about like, getting the lineup, right?

04:31 Dom Dubois: Yeah. Mmm.

04:35 Emma Sjöström: So is it correct? Like I think you told me like, you collect the feedback after the

04:42 Dom Dubois: Yeah.

04:42 Emma Sjöström: capital you ask people already then about like, what would be a

04:43 Dom Dubois: decideline next year, like

04:46 Emma Sjöström: Yeah, that's that's what we've started doing. Which I think is really interesting things to do. So the team in the in the Data and Partnerships department of started, sending out in At the end of each festival, they'll send out an email saying Do this do this survey and you'll have a chance to get a free ticket to next next festival.

04:47 Emma Sjöström: So it gets quite a few, especially if you have lots of people that keep coming back against quite a few people doing the survey. And as part of that one, you fill out the survey you you have, you're able to type in. Who would you want to see next year and it filters and we can.

04:47 Emma Sjöström: So there's no errors in data select site who they are and then we can use that to really focus the artist booking team on who who should be playing at which festival and then also on the other side of it, there's also what, what didn't go? Well, so if there were any Sort of downsides of the festival sit toilets or whatever.

04:47 Emma Sjöström: Went up to up to specification or a bit mucky or something. Then we can look at spending or changing provider to improve that or long cues in in certain areas. We can look at like changing over around the site a bit. So we use a lot of surveys to try and improve processes where possible in the festival and then we try and use that external data.

04:48 Emma Sjöström: Well the opcodes we try to encourage them to use that external data to try and position there forecast accordingly.

04:49 Dom Dubois: Yeah, that's that's what we've started doing. Which I think is really interesting things to do. So the team in the in the Data and Partnerships department of started, sending out in At the end of each festival, they'll send out an email saying Do this do this survey and you'll have a chance to get a free ticket to next next festival.

05:09 Dom Dubois: So it gets quite a few, especially if you have lots of people that keep coming back against quite a few people doing the survey. And as part of that one, you fill out the survey you you have, you're able to type in. Who would you want to see next year and it filters and we can.

05:28 Dom Dubois: So there's no errors in data select site who they are and then we can use that to really focus the artist booking team on who who should be playing at which festival and then also on the other side of it, there's also what, what didn't go? Well, so if there were any Sort of downsides of the festival sit toilets or whatever.

05:49 Dom Dubois: Went up to up to specification or a bit mucky or something. Then we can look at spending or changing provider to improve that or long cues in in certain areas. We can look at like changing over around the site a bit. So we use a lot of surveys to try and improve processes where possible in the festival and then we try and use that external data.

06:11 Dom Dubois: Well the opcodes we try to encourage them to use that external data to try and position there forecast accordingly.

06:19 Emma Sjöström: That's really cool. I think.

06:21 Dom Dubois: Yeah, I mean it's not quite there yet but that's the idea.

06:22 Emma Sjöström: To. Yeah. No that's that I get that because that's the thing. Like how do you deal with all that uncertainty and to meet it sounds? Like one way that you're now starting to deal with that massive and certainty.

06:35 Dom Dubois: Yeah. Yeah exactly exactly. Yeah because we just really can't predict about having a bit of control so that kind of gives us a bit of control. So if we do pick like I said that survey we pick an artist that we know is gonna Resonate, well, with the audience.

06:52 Dom Dubois: So we can be more We can kind of make a more reliable assumption on on ticket sales and if we kind of go in blind. So it's really just knowing your audience and making improvements that they would want. and then you can have more reliable guesstimations, I say,

07:11 Emma Sjöström: How do you cut like this sounds like a lot of interesting context and knowledge right. How do you quickly go about like saving that somewhere or making it or

07:16 Dom Dubois: Yeah.

07:19 Emma Sjöström: even if you don't have anything implemented yet like what are your general thoughts around? Oh man, this festival the feedback last year and the years before, I'm assuming

07:31 Dom Dubois: Yeah.

07:31 Emma Sjöström: ideal state, you'd want to compare that your assumptions in the forecast and see how well they correlated. Stuff like that. If that's something you guys are thinking about already or

07:42 Dom Dubois: It's something that I'm not involved with personally. It's the Fp&a team are

07:43 Emma Sjöström: definitely using that.

07:47 Dom Dubois: Right.

07:48 Emma Sjöström: To incorporate into their into their planning. So it's something that that you we have a expanding data team that are creating models etc to work with the FNA teams to help improve that. But it is all still quite manual on Excel. So it's like it's not. It sounds oh, very forward thinking but it is, it is still very much in its infancy.

00:00 :

07:49 Dom Dubois: Yeah, now sure but it still super inspiring. Do you? So do you typically do a lot of or try to reconciliate between the PNA forecast in your forecast?

08:17 Emma Sjöström: So, the

08:33 Dom Dubois: And how yeah, how, how does process look like for you?

08:34 Emma Sjöström: so that Yeah, so the forecast is managed by by myself and the Fp&a. So each fp&a member basically has an a region which they manage. So we'll have someone that manages

08:35 Dom Dubois: so that Yeah, so the forecast is managed by by myself and the Fp&a. So each fp&a member basically has an a region which they manage. So we'll have someone that manages

08:49 Emma Sjöström: Right.

08:51 Dom Dubois: the Australia region. Someone that manages, Germany, etc. And they know the business is better than me. So my job is more sort of the high level. Once it's come in to see, Where the whole whole business moves, but they would work with the FPNA team to see if those assumptions are correct based off the sort of data that they have.

09:15 Dom Dubois: and in terms of how we sort of map it to our balances, it's something that we're still working on because I think I mentioned to you At the act. We're currently implementing a TMS and payment provider. TS payments. So, At the moment, we have really fragmented data. So whilst we're getting better data on in terms of like what people want to see next year and The actual financial data that we have is, is super fragmented.

09:44 Dom Dubois: We have many different people, many different teams on different ERPs, there's no unified ERP but equally we had no way of seeing live balances on accounts. So, previously, we just relying on these forecasts and them giving us the correct balance, we have nothing to tie it back to so as part of the forecasting, Process.

10:05 Dom Dubois: We ask them to one compare it to the previous forecast to see where it was different. So it is of course a quarterly basis and then compare it to the current balance. What they have in the bank now and see how far off it is and try and explain before we come back with the questions, what that might be.

10:28 Dom Dubois: But once we have the live balances connected via the tis payments and the TMS we would be able to do that comparison ourselves, but very much rely on the Opcose and the honesty of them to say that this is their balance because of, theoretically, at the moment, they could say they've got one million because they don't want us to take one two million or whatever up to us.

10:55 Dom Dubois: So it. Yeah.

10:53 Emma Sjöström: Oh yeah.

10:54 Dom Dubois: Can you see? So we don't have access to all that, all their banking banking. Portal. So this this implementation will actually be able to give us that cash visibility at the moment. We don't have cash visibility and we want to get that cash visibility and then would also have a better after we have that a better historical view of cash that we can also hopefully blend into this forward thinking forecast as well, so we can kind of mix it to a historical and a forward thinking forecast.

00:00 :

11:25 Emma Sjöström: Yes, I would love to there's so many questions. I think it's interesting. So so you're saying like essentially one like the cash visibility and when it comes to variances right between like what was predicted what was the outcome and actual balances sitting around there so that you can maybe is it because you want to drive a more efficient use of that cash or like I'm asking lately just to really like, get to the get

11:51 Dom Dubois: So yeah. So the question you say like what we're trying to see what balances are to use the cash more efficient in the business, is that they say. Yeah, so it's

11:59 Emma Sjöström: Yeah.

12:00 Dom Dubois: essentially, that's exactly hit the nail on the head there. So, like the way, Our businesses work is they have a very, Positive Network and capital cycle. So you get cash in before you have the festival. So you buy a ticket months in advance the cash, it's in our account and it doesn't get paid out until we pay all the site fees.

12:23 Dom Dubois: We obviously pay half the artist fee but four and then also yeah, all the setup, all the, all the admin or everything goes just before the festival and then you see a dip in cash, so there's often quite a lot of cash which is sat on operational banking counts for many, many months up to the festival and then they run out of cash and then they need funding just before And because we didn't have that cash visibility, the big pain point was that we were relying on a central debt, facility to fund each opco when they had those

12:56 Emma Sjöström: Okay.

12:58 Dom Dubois: shorts short term needs. So what we're trying to do is by connecting all the bank accounts, by this tis payment system, we are hoping to create a sort of in-house cash pull, so we'll be able to use the excess cash on certain people's

13:14 Emma Sjöström: All right.

13:17 Dom Dubois: businesses. Draw it up, centrally, and then loan it to another business when they need it just before their festival because it only really need the cash for a very short period of time, a couple of months while they have that that deficit just before the festival. And whilst they're waiting for food and beveraging come to come in.

13:38 Dom Dubois: And then they have the Surplus again, so we'd found we'd be drawing on expensive, debt facilities to fund for short terms and to pay off. Whereas, if we pulled all the cash together, we could utilize that better in the business. And the reason why we decided to go down this route rather than move everyone over to a single bank.

13:59 Dom Dubois: I was a couple of reasons. Well, one of the reasons is that quite a lot of them have partnerships involved with the festivals. So, It's not so easy, easy to move them over. So for example, bring the noise at

14:06 Emma Sjöström: but,

14:11 Dom Dubois: their biggest partners ABN bankers are not avian. So that's a Spanish bank, which provides all the payment systems in the festival. So they need that as an operational bank, so we couldn't go to a HSBC and say, Can we

14:21 Emma Sjöström: but, Can we?

14:26 Dom Dubois: Set up payment pool and transfer. Everyone's banks over because they need that. Local bank service. Whilst we're looking to improve it as well. Quite a lot of the festival business is done in cash. So we also have a lot of cash on deposit, so we need that local bank service to deposit the cash.

14:48 Dom Dubois: And withdraw the cash. Etc. So we needed to keep that local banking relationships at the. Only way that we could set up a cash pool would be to utilize a Treasury management system to target balance. And then what we're trying to do on the side of that is, Slowly transition them over to a bigger bank, like a.

15:11 Dom Dubois: Maybe I hate this bc JP Morgan. Chase Barclays to

15:15 Emma Sjöström: Yeah, you've gone from like 40 to like 20 banks or something already.

15:21 Dom Dubois: Yeah yeah exactly. So we're looking to rationalize them down and then where possible move them over to a big global bank where they can actually get the cash management. Processes in place to to do cash pool, sweeping or virtual pooling etc. But for now, this is the easiest or best option.

15:40 Dom Dubois: We could find to pull the cash in

15:44 Emma Sjöström: Yeah. But you just explain to me how so tis. Are you setting up like automated payments based on target balances? Like How is the setup or like what is the

15:55 Dom Dubois: So I'm tis is it doesn't have the capabilities of a Treasury management system so it's just a payment hub. So what can do is what is sort of expertise is doing is connecting to the banks and creating a payment connection very fast. So we've managed to set up. I think four accounts and it's what we're in April.

16:15 Dom Dubois: So about about bank a month, which is not too bad to connect to the system and then on the system, the opcodes were able to start creating their payments through there. So, we have a centralized payment system, but on the back of that we can

16:28 Emma Sjöström: Right.

16:31 Dom Dubois: attach a Treasury management system to it and they can talk to each other via an sftp connection. And that's where the logic would sit on. Figuring out what the balances are on each bank and where the surpluses are, and it can put in rules to turn pool the cash up.

16:50 Dom Dubois: So, send a request to pull the cash up to our central header, but on when we push catch down, Because it's not all within the same bank and you can't have zero balancing and them talking to each other like that. We have to set up a process where we every opco has a weekly or bi-weekly payment run.

17:13 Dom Dubois: so, In the tis tool, they have access to templates. So they would log on to this isn't actually implemented yet. This is all theory. They would they would log on to the tis portal and in there they can say they put in a request for this payment run. So my payment runs every Thursday I want to put in a request to fund that account, five million.

00:00 :

17:39 Emma Sjöström: Okay.

17:40 Dom Dubois: That would then shoot down the five million for them to do their payment run and then the logic behind it would leave that account for a couple of days to clear. And then, once the accounts cleared it will then check it again and it will scoop up any cash that's left over and you'll keep track of that in the TMS.

18:00 Dom Dubois: So it's it's we've set up our own transfer pricing which will keep track of it and

18:05 Emma Sjöström: Okay.

18:06 Dom Dubois: we'll price the the loan positions between all the entities

18:11 Emma Sjöström: But how what teams are you looking to add on top or like?

18:14 Dom Dubois: so, the TMS we've selected is integrity,

18:17 Emma Sjöström: Integrity. Okay.

18:18 Dom Dubois: So, I think it was the only TMS provider. That wasn't at. That, that wasn't at the HD. Yeah, yeah. So it's um, Yeah, I mean it's a pretty established tms

18:26 Emma Sjöström: I have.

18:32 Dom Dubois: When and the reason we went for them was weird. There was super impressed with the understanding of what we needed. So we basically pitched them. This is what we want to do and they and they came up with this process with us. And it seemed like a very feasible process to implement especially on the short term just because of our needs.

18:53 Dom Dubois: Like most most people would have just gone for a banking RFP and just try to convert everyone to the same bank. But as

19:00 Emma Sjöström: Yeah.

19:02 Dom Dubois: We have such an array of banks and quite a lot of them needed this. This local touch. We thought this would be a better better way to do it, and get that instant impact straight away. Whereas a bank change can take very, very long with kycs etc. Whereas we're what? Four months in and we've managed to connect most well, four banks.

19:22 Dom Dubois: So hopefully by the end of the year we'll have we'll have all the banks connected. So what we're aiming for, 20 banks of maybe maybe not quite all of them, but then once you've got the more connected, we can have that logic in place and hopefully be working seamlessly in the background.

19:39 Dom Dubois: We're just a bit of manual input for requesting funds back but it should take become part of their pay payment runs. Anyway, so just extra thing you bought on to a payment, run request.

19:52 Emma Sjöström: Yeah, I mean that sounds super interesting. I'm happy if if that's okay, I would love to understand like by end of year, or early next year. How that how that worked out for you guys? I think that's super fascinating.

20:03 Dom Dubois: Yeah, definitely. We can catch up and let you know how it's going.

20:06 Emma Sjöström: Yeah, thank you. But so just Going back a little bit to the cash visibility topic.

20:13 Dom Dubois: Yeah.

20:15 Emma Sjöström: So, the how how are you envisioning now that you will using your new tools? I know you already told me but just like, repeat, like your top three things around cash, visibility that you will or hopefully will get now what would those be?

20:31 Dom Dubois: The top three things.

20:33 Emma Sjöström: Yeah.

20:34 Dom Dubois: Well. One is obviously the cash visibility. We'll be able to see where that where the cash sits and then two would be to be able to access that cash. So the tool will be able to directly access it and pull it centrally and

20:49 Emma Sjöström: Right. So intercom like executing like into company.

20:54 Dom Dubois: yeah, so essentially an intercompany loan agreement between all the entities and our top, Header entity, which is in the UK and so it's a cash polling agreement, which

21:03 Emma Sjöström: Yeah.

21:06 Dom Dubois: we've drafted. Between all the different. Entities, and they can scoop from any bank and it will just be a single loan position that will be set up.

21:18 Emma Sjöström: Interesting.

21:19 Dom Dubois: Yeah, and I guess. On the back of that. And once we've got that, visibility we can Then pull historical balances as well from, from the tis tool. So we have the ability to go back and pull historical, balances of all the bank accounts and then incorporate a bit more historical.

21:40 Dom Dubois: Analysis into our cash forecasting because at the moment, it's all forward-looking because we haven't had that visibility.

21:49 Emma Sjöström: And when you say historical like, you mean to look where where were you off, why were you off those sort of things. So or what this historic,

21:57 Dom Dubois: Yeah, I mean is exactly that. So we can compare the previous forecast and do a bit of analysis around that and see what why we were off. But then we can also have a more accurate view of the trends. So use although I say it's not that useful it's still can be relatively useful to see where the where the rough trends are and move those trend lines where need be.

22:20 Dom Dubois: Whereas at the moment we've got maybe about a year's worth of very fragmented data.

22:26 Emma Sjöström: Yeah.

22:26 Dom Dubois: Which is basically just requests over email of what their balance was at the end of each month, which is not that useful. Whereas, if you've got a day-to-day

22:33 Emma Sjöström: Coach.

22:35 Dom Dubois: changes, you can actually see the proper shifts in balances. So you wouldn't be able to see monthly dips, weekly dips, in our current data, we can just see very, very general trends, which isn't useful. I think for forecasting.

22:48 Emma Sjöström: Interesting. How would you ideally so you get the data and how would you use it? Like Would it be Excel? Would you imagine another tool to help you somehow make sense of those variances or identify them, or Like in your ideal. World.

23:05 Dom Dubois: Um, in an idea world of being to move away from Excel but it will be Excel at the moment. So it's downloading. Those reports into into Excel and comparing them to the monthly forecasts that we've got from. From the opcos.

23:22 Emma Sjöström: Yeah, makes sense. And the reason I'm asking is that and I'm not trying to pitch, I'm just trying to inform what we're doing. It's just the, the holistic

23:30 Dom Dubois: Yeah.

23:33 Emma Sjöström: thing is that what we're building is the cash forecasting bit, right? And the

23:37 Dom Dubois: Yeah.

23:37 Emma Sjöström: capability in terms of also like, you know, proactive that Recommendations for positioning your cash, but we are also providing a whole like automated out of the box variants analysis, right? So you can like, visually see, your variances and click and drill down and easily spot like the driver behind variance.

23:57 Emma Sjöström: I was just I think kind of just anything stuff to hear your thoughts about it because we're also,

24:02 Dom Dubois: Yeah.

24:07 Emma Sjöström: Balancing between this, you know, exile versus not Excel. And how important is it to still allow for the user job. Excel for example, excluding data back to

24:17 Dom Dubois: Yeah.

24:19 Emma Sjöström: Excel or input, you know?

24:20 Dom Dubois: I think that's always a useful tool to have the ability to to manipulate in Excel because I mean it's an accountant and and treasurer's best friend either. Everyone knows how to use it. But ideally for the visualization piece it's good to see in a tool like having those abilities to drill down and and really identify what the differences are fairly like automated variants analysis, would be really useful at the moment.

24:46 Dom Dubois: We don't do any but it's something that we will do in the future.

24:51 Emma Sjöström: exactly, you're coming to that once you have because not now

24:53 Dom Dubois: Yeah, we're very much like in a in a journey then we're right at the start at the moment but we've got I think we've got good ideas but it's

25:01 Emma Sjöström: Yeah.

25:02 Dom Dubois: It's putting it together to build a bigger picture which is, which is the interesting part.

25:07 Emma Sjöström: Yeah, I think it's super cool. I think. Yeah, you guys are really forward thinking in my in my opinion. And yeah. So you're building out your

25:13 Dom Dubois: What. Yeah.

25:16 Emma Sjöström: connectivity. Yeah, it's gonna be really cool to to recycle back to you if you're up for that in the future just to hear how you guys are getting along.

25:24 Dom Dubois: And have you, have you done any of these chats of anyone else in the In the events industry and have any sort of feedback on what they're doing differently to us.

25:34 Emma Sjöström: No, you're the first one in terms of events industry.

25:37 Dom Dubois: Okay.

25:39 Emma Sjöström: But happy, happy to share. If I if I find it

25:42 Dom Dubois: Yeah, and if you do would like because I say we we are really looking to improve things and looking for advice as well, we're open to hearing how other people do things and seeing if there's any ways that we can do things better.

25:57 Emma Sjöström: Yes, if you don't mind, I could forward that to one of our founders. I know he has really interesting thought on on the internal catch pools. Now, you already have that and it seems that control but I just think it's

26:07 Dom Dubois: Yeah.

26:09 Emma Sjöström: I don't want to like solicited advice your way but maybe, you know, I think.

26:10 Dom Dubois: Yeah.

26:15 Emma Sjöström: Yeah, that could be interesting exchange of ideas there.

26:17 Dom Dubois: Yeah. I mean, any advice we will take it's free advice.

26:22 Emma Sjöström: Yeah. Exactly, right. That's true. But I thought about that like also I enjoyed like

26:27 Dom Dubois: Yeah.

26:30 Emma Sjöström: the Treasury community. It feels like you guys all pretty open-minded and generous with kind of sharing knowledge.

26:35 Dom Dubois: Yeah no I agree with that. I think I'm yeah. All my sort of People I've passed work with and and my previous boss as well is more than happy to sit down and just talk things through me. So I'm hope to do the same as well because I think it's good to learn off each other and bounce ideas off each other.

00:00 :

26:53 Emma Sjöström: Agreed.

26:54 Dom Dubois: That's why I think these ACT conferences are so good because it gives you the opportunity to to broaden that network and learn what other people are doing.

27:01 Emma Sjöström: 100%. Hundred and for me, I'm still learning. So it's obviously really, really nice domain to be in. Okay, so it's a going back to cash forecast. I'm just looking at my questions here. This was a really, really nice chat but What. So, if if You could dream. Like I know you're on your journey already.

27:27 Emma Sjöström: So it could be like the the dream and state of that journey or just like in general, but focusing on forecast. What would this perfect forecast system or wait a forecasting do for you? That your current one just doesn't, or doesn't have the capability to do. It's a very broad question but

27:54 Dom Dubois: I guess it would be something that can be quite dynamic in agile so we can have it on say a system where Every time the we update it. Say, if a ticket sales go better, it can automatically move the the dial across a bit so it can automates early.

28:13 Dom Dubois: We can move. This increase in funds across a bit is like, if it kind of updates itself I would say that would probably be the ideal situation that I would like to see myself in. But then also it can do its comparisons to our banking data relatively quickly and easily and it not automatic, because I think it's never gonna get that get that way, but it can give us like think where we need to look at can so once we've submit it and go right his, his new major differences, these are the things you need to investigate and then I guess kind of like doing that sort of underline analytical.

28:56 Dom Dubois: Donkey work underneath so we can just do the more high level. so like a high level changes up top would be ideal really

29:06 Emma Sjöström: Cool. So, sort of being a bit proactive in drawing your attention to things that matter.

29:11 Dom Dubois: Yeah, yeah, exactly.

29:13 Emma Sjöström: spend your time on and then also like easy like relieving you of some of the day-to-day manual like never-ending kind of

29:21 Dom Dubois: Yeah, because I mean at the moment it's it's super difficult to even collate it. We getting So we've got I think 23 different operating businesses and they send back templates, we are some not to change it, but obviously in Excel, they make certain changes thinking they're being helpful and then all you look up to look wrong.

29:43 Dom Dubois: And then you're getting errors and you're doing all that sort of side of things. You spend half a day collating, all the data To get to something that should read done as soon as you cook a submit button,

29:54 Emma Sjöström: Yeah.

29:55 Dom Dubois: there's some sort of standardized and it can do that analytics. And Repairs in the background and that's probably a pipe dream at the moment. That, that would be ideal.

30:06 Emma Sjöström: I think this is such interesting to just a question, tell us because it's so I'm noticing. It's very common to have this model where your taking a bunch of different spreadsheets from different entities, or organizations or projects, or however, you're set up, and then you merge them into this one.

00:00 :

30:22 Dom Dubois: Yeah.

30:25 Emma Sjöström: Master file like that. No full process but I think and I think that's a made

30:25 Dom Dubois: Yeah.

30:29 Emma Sjöström: great, right? Because I do believe in the fact, like, how getting a forecast, that's, you know, numbers been massaged by a human. There are Built in assumptions. All of that. I think that's beautiful and I'm really interested in like, comparing those forecasts too and don't get me wrong, but the machine learning based forecast and I get your point on the ticket sales for being learning that's gonna be a hard one, but what would be your like general opinion when it comes to the other forecast? Like Costs, or the other.

00:00 :

31:02 Dom Dubois: I mean, that could easily be done with some machine learning. You could a way of just putting in inputs on on Increases and stuff and then it machine learning can set your assumptions and put that all underneath. It's sort of what we have, not machine learning, but it's just Excel assumptions that we put in.

31:22 Dom Dubois: So we, we asked the Hot Coast to put in what their budgets are and for expenses, etc, and it automatically puts it into the certain lines and spreads it across. Same with taxation etc. So we've got some stuff like that, but definitely stuff that could be improved with. some proper AI machine learning underneath, but I think that the revenue side of it, Whilst it can be sort of spread out of the machine learning.

31:49 Dom Dubois: It really needs that human touch to to really Get the vibe of like how things how things are looking, especially in our

31:54 Emma Sjöström: Yeah.

31:56 Dom Dubois: business.

31:58 Emma Sjöström: Yeah, for sure. Because where I'm thinking this is quite interesting, too, right? Because we are building. I'm just being transparent with you. Like we're currently building our forecast from the account level up.

32:10 Dom Dubois: Mm-hmm.

32:11 Emma Sjöström: Whereas of course like a a focus coming from another part of an organization, it's not necessarily at the account level, right? So,

32:20 Dom Dubois: You say account? You mean the accounting? Yeah. Okay.

32:23 Emma Sjöström: The bank. Account. Yeah, we go bank account by bank account and then By category.

32:31 Dom Dubois: Okay.

32:32 Emma Sjöström: And that's kind of the bottom up approach we have when it comes to the machine learning models. And it works really good for the outflows.

32:39 Dom Dubois: Yeah.

32:40 Emma Sjöström: But where I'm where I'm looking for us to innovate. A bit more, is to reconcile like the The inflow forecast with that human knowledge.

32:52 Dom Dubois: Yeah.

32:52 Emma Sjöström: And I think that's really challenging given that we are currently at least and we might pivot, but we are, we are really, like, a bank account upwards focus. But how can we use that human input, kind of sanity. Check the models assumptions in terms of the

33:08 Dom Dubois: Hmm.

33:10 Emma Sjöström: Enclosed. So either kind of salary to check. or, you know, act as a kind of influence or input to that forecast, even

33:19 Dom Dubois: Yeah, yeah. Trusting.

33:21 Emma Sjöström: Do you have any like General? Like I get That's a bit of a general question but you get, do you get any thoughts? If I, if I talk about these things? Does that spark any thoughts or ideas in you?

33:35 Dom Dubois: yeah, I mean, in terms of the inflows So, how would you predict the inflows based off a bank account level? Because surely you need that you need the budgets to get inflows. I can see you can just look at historical.

33:48 Emma Sjöström: Yeah, exactly. So it depends so much on the nature of the business, right? So if

33:52 Dom Dubois: Yeah.

33:52 Emma Sjöström: it You only have online collections. Like, then we could look at your historicals and do a pretty good prediction in terms of, like, what can be expected? Depending on what month, you're like, all of these things like recycling, but it could also be about then letting a human kind of mildly influencer forecast.

34:12 Emma Sjöström: It could be like, Really a new product. We're pretty sure it's gonna have this effect on

34:14 Dom Dubois: Yeah.

34:18 Emma Sjöström: collections here and then you could just I call it influencing the machine learning model. But that's that's easier case, right? For you guys. This is why I'm so interested because that's just not that data is just not there. We could look at. Okay, so, how much Ticket sales was the last year.

34:35 Emma Sjöström: But no, if you, you know, increase the scope of the festival or like something, you know, you got that thing about. Hey, the lineup is perfect, like all of

34:45 Dom Dubois: I mean, you I mean, in an obvious situation, you could, you could have it linked

34:47 Emma Sjöström: those things, right? So how do you allow humans to

34:55 Dom Dubois: to many different data sources and so, if you can get a data lake in place and

34:56 Emma Sjöström: Yeah.

35:00 Dom Dubois: you're seeing these these surveys in in a live day so they can see if it's matching what we've, we've actually produced in the festival then I guess you could use the machine learning to Make a more accurate assumption then and I mean if you want to go step further, where I would see it, maybe in, in 20, 50 Or whatever.

35:25 Dom Dubois: It's it's linked to The. the weather forecast as well and it's like looking at that and it can and it can judge like How we expect to been a week's time, if it's going to be sunny, are we going to be getting more?

35:44 Emma Sjöström: revenue from from drinks or whatever and stuff like that but I don't know if it's ever going to get to that stage and I mean BBC news can't predict the weather right now even though like yeah so Exactly.

36:01 Dom Dubois: it is just it's just super difficult for us us to forecast and hence, why we haven't really been doing it so much until until now, which has been mainly a driver from our private equity owners to Basically because we've got quite a lot of acquisitions that we're looking to fund as well.

36:19 Dom Dubois: They want to try us to try and forecast cash a bit better so that we can forecast the interest payments debts etc which we weren't looking at before.

36:29 Emma Sjöström: Right.

36:31 Dom Dubois: And we kind of explained them as it is so difficult to forecast, but now it's become more and sort of a necessity, We are looking at ways to improve it and I think we are making ways to improve it and it's been an interesting topic to look at

36:46 Emma Sjöström: For I bet I bet.

36:48 Dom Dubois: Yeah.

36:50 Emma Sjöström: Yeah but it sounds to me also like one way, you're going to deal deal with it is your internal cash pool set up, right? Which also kind of

36:57 Dom Dubois: Yeah, I mean exactly.

36:59 Emma Sjöström: Today, cash management perspective is going, it sounds like super helpful.

37:03 Dom Dubois: Yeah. Yeah, I know exactly. And I think that's what we're really keep. We're really keen to implement It became quite apparent that when I joined we needed to get some Some tech in place to do this because when I first joined it was very much a manual deposits. Which are quite time-consuming.

37:24 Dom Dubois: It can take a week to get all the approvals in place. And you have to get bored board approval, sign off, then all the legal documents signed up, transfer price, every single every single loan. If the interest rates change, etc,

37:40 Emma Sjöström: Well.

37:41 Dom Dubois: Whereas, if you've got this internal pooling structure in place, you only do that once. And then it's all flowing in the background. You might have to adjust the transfer pricing as our credit ratings change etc, but that can be done on.

37:54 Emma Sjöström: Yeah, for sure. Sounds great.

37:55 Dom Dubois: A yearly, maybe even even every two years so it's much more fluid approach. And I think, yeah, the implementing all this text would be really helpful for us to better optimize our cash management, but even on that as well, it will help with the forecasting as I said, go and going forwards because we'll have that historical days.

38:18 Dom Dubois: That's reliable. Not just monthly data and hopefully we can get to a place where we're incorporating both and they're bouncing off each other. and so, Is Palm, isn't it? The Is it? Yeah. So is is it a TMS that you've got? Or is it just a forecasting software?

38:38 Emma Sjöström: Right now we're not positioning cells, our positioning ourselves as a team, as in that sense.

38:43 Dom Dubois: Yeah.

38:44 Emma Sjöström: We're a bit more. Exactly. So we're much more heavy on the forecast and cash management side of things.

38:50 Dom Dubois: Okay.

38:51 Emma Sjöström: For sure. It's it's a bit like our philosophy, right? It's a little bit of like Do one of you things right now and do them really well.

39:00 Dom Dubois: Yeah, and then build out.

39:03 Emma Sjöström: So what we've noticed is that we we work really well alongside with TMS most of our customers actually is kariba.

39:12 Dom Dubois: Yeah.

39:12 Emma Sjöström: but one example where we sit really nicely at like together with career bar, the kind of enhances the forecasting and the visibility and the variance analysis and also we have like these Reporting capabilities in terms of like building your own dashboards and like creating your own disability dashboards and also automatically sharing those via email and whatever.

00:00 :

39:36 Dom Dubois: And is that?

39:36 Emma Sjöström: With yeah.

39:37 Dom Dubois: Is that via an API interface with Curry bows. It all accessed on the TMS or do you have two separate? Portals that essentially need to log on.

39:48 Emma Sjöström: So it's no. It's so we, we do ingest the data from creative directly into our system, but set up can differ a little bit depending on the organization like

39:53 Dom Dubois: Okay.

39:57 Emma Sjöström: the client organization.

39:59 Dom Dubois: Yeah.

40:00 Emma Sjöström: Sometimes we just bank statements or as well.

40:04 Dom Dubois: Okay. And is it agnostic to TMS or is or is the software only?

40:10 Emma Sjöström: That's the current goal to be to be agnostic in terms of teams. And that would be ideal. So we've been all depends like we're still learning and exploring what it would

40:19 Dom Dubois: Yeah.

40:22 Emma Sjöström: mean to to work, you know, alongside this tms versus that humans. But I I I say so for sure.

40:31 Dom Dubois: It's cool.

40:31 Emma Sjöström: Yeah, yeah. So but that's where we're at right now and we don't have these amazing cool cash like pulling solutions in place but we have not internal project just around the visibility into the cash pull in the traditional sense, like we have customers who have the Cash pro sitting with a bank like that structure being very Visible.

40:57 Emma Sjöström: And the noise that can, you know, come of the transactions, like, all the, in the company transactions and correctly, labeling, and categorizing. Those is super interesting stuff in my opinion, at least. Yeah. So that's a

41:05 Dom Dubois: Yeah. Yeah. Definitely.

41:13 Emma Sjöström: little bit. What we're what we're doing currently. But yeah, so obviously AI, native all of that. But very like, Mindful of where we use AI and where we don't.

41:28 Dom Dubois: Yeah, yeah. Okay.

41:29 Emma Sjöström: Yeah. So, if I interviewed many um, and a couple or and I have a couple more

41:32 Dom Dubois: Awesome. and if you have you interviewed many, many other treasures since the act

41:43 Emma Sjöström: interviews schedule in the coming weeks. I think that was super nice. I really

41:45 Dom Dubois: Okay.

41:49 Emma Sjöström: enjoyed the going to the A city and like being out there talking talking to you guys because I feel like that's typically what product function has to do.

41:58 Dom Dubois: Yeah, why is similar to us? We need to know, you know, the customers. So you then you actually building a product for them so that they'll come to you rather

42:06 Emma Sjöström: Exactly.

42:08 Dom Dubois: than you having to keep pushing for them. So that's the sort of model that we're trying to work towards as well, by putting out these surveys and trying to improve the product that we have for them being closer to our customer. So then you don't have to be spending so much on advertisement.

42:25 Dom Dubois: You have that loyal customer base. Like one of our festivals. Wackers is A sales are faster than Glastonbury because they just know. They know, their their base. They know their niche, and they just always pounded to the client and they've got that perfect beast set up. Maybe sometimes too perfectly need to cut their own costs a little bit but like it is, it's amazing how quickly they sell out and how loyal the fans are.

42:50 Dom Dubois: And we're just trying to push to make to keep that fan base and try and build those sort of fan bases for the, for the other festivals as well.

43:00 Emma Sjöström: Yeah, that's super cool. Honestly that's actually no small feat, right?

43:05 Dom Dubois: Yeah. Yeah.

43:07 Emma Sjöström: Yeah.

43:07 Dom Dubois: It's a long-term goal. But is there any anything from from your discussions that you have so far that might be relevant to us or that you could that you could share obviously, nothing confidential or anything? Any sort of methods that you've encountered from speaking to any of the other treasurers that you think might be beneficial to the setup that we've got, or is it similar sort of stuff that they're doing as well?

43:31 Emma Sjöström: Let me think of that, that a little bit if that's fine now cycle back.

43:35 Dom Dubois: Yes. Yeah. That'd be great.

43:38 Emma Sjöström: Yeah, for sure. I don't have anything more right now. I think this was a great conversation.

43:49 Dom Dubois: Okay. Yeah, awesome. And yeah, let's let's keep in touch and, and yeah, more than happy to one way a bit further in the project give you an update and that, you know, where we're at. And Yeah, if you think of anything that from other discussions with any other treasurers that you think might be helpful to us,

44:04 Emma Sjöström: Of course.

44:04 Dom Dubois: I'll be really helpful if you could reach out to us and let us know.

44:08 Emma Sjöström: Of course, of course.

44:10 Dom Dubois: Amazing.

44:10 Emma Sjöström: Also. Thank you. So nice again.

44:12 Dom Dubois: Amazing. You too Emma, thanks to the shop.

44:17 Emma Sjöström: And thank you. Take care.