Canva’s Profitless Profit, Smiggle Slumps, Catapult Flying, CGT Debacle, the Incredible Benefits of Autonomous Cars, Firmus and Adam Gets Run over by a Golf Buggy

27:37Now, will it go higher? Probably in future it will, right? But we are picking a moment in time that is quite unusual. If you look at Atlassian, for example, I'm comparing to... Atlassian's back down to... And Seek as well. Seek's back down to its 2013 price and Atlassian's back down to its almost 2018 price. So this is not a moment in time. This is a 10-year journey to irrelevance. The difference between... I'm actually not going to talk about specific companies, but the difference between, this is maybe what I'd look at to separate some of these tech companies as well. How many are at these lows based on a long, slow slide over a year or two?

28:10And how many just are like going down a cliff face since the SaaSpocalypse? That's also one way to kind of divide these up a bit. I'm not sure that's that relevant. Why? The degree of overvaluedness or undervaluedness is kind of, look at the long-term price over a long period of time. But the market's wrong about some of these valuations. You think they're too low now? Yes. I think they're still too high. Depends which, right? I think there are some that I think, I mean, obviously I'm going to say I think Catapult is undervalued. I don't have to make the whole argument on this podcast. But Catapult's up 72%.

28:41You're up 72 % in five years, which is a great return. I'm talking about Atlassian, which is down 70%. So you can't compare the two. I think, look, Atlassian, we've spoken about it a lot. They've got many problems. Overvaluation, not making cash, and they're the most imperiled by AI in a lot of people's eyes. Possibly including ours, right? So yes, I definitely take those points. I think this is how I've started looking at, this is another prism through which I've now started looking at AI. I divide it into these three questions for every company. Question one, because I worry about the negative before I worry about the positive with AI.

29:17Question one, is our industry going to be negatively imperiled by AI? And so Catapult, for example, is in sports. that's not going to be negatively imperiled neither are sports teams um and like atlassian that's software developers probably there's an argument for that being negatively imperiled although i'm not going to talk about this now but there's an mit study that just came out that really refutes how much software developers will end up being replaced in the long term by these ai models so but we can say at the moment software will certainly become more efficient that's right we we had a hackathon last week which is amazing and it was really ai driven hackathon for everyone and we had 64 different teams and the 12 best teams one one team created an automatic bug checker so we have team of people that go on qa stuff and bug check stuff and automatically bug check stuff and we another team created effectively an automatic agent that fixes stuff so obviously not super complex stuff so you could automatically find the bugs and automatically fix the bug like not instantaneously but pretty quickly this will take some of these bugs weren't getting fixed at all i agree with you so you know what i'm going to talk about this study in a second.

30:21So I'll tell you these three things. So one is, is the industry going to be negatively affected? Yeah. Like you're in travel, it's not going to be negatively affected by AI. Like the industry. Yeah. The makeup of it might be. Yeah, but the industry, people are not going to, well, travel's going to, people having holidays is fundamentally going to change because of Gemini. How they book them might, but me going to a hotel and going to a pool and doing whatever. If I was being really pessimistic, you could say, if AI destroys a lot of jobs, people have less money to travel. That would be the ultimate pessimistic view all right but like i think that to me is too long a bow to worry about today it's a third order effect i know people are saying that but people say lots of things anyway a futurist has to say something dramatic about the future or they're a presentist yeah and so like i think i would just largely ignore that stuff so number one is the industry going to be negatively affected number two is the product that i'm offering to companies to clients going to be negatively affected and so one of that we've had lots of conversations about what type of SaaS businesses will and won't be negatively affected.

31:21Personally, I like long contracts with complicated products. Switching costs. Lots of high switching costs, lots of proprietary data that's in there, all those kind of things. Specialised sector. Specialised sector, thinking it's mission critical, so I'm not going to trust it to somebody. So that's the second group. Is my product suite going to be negatively affected? And then I'll say, is there an opportunity to have margin expansion and or value creation from AI? I'll get to that as number three. And that's kind of how I've started looking at the world as well. And I think I reached that in a brainstorm with Will and Sean Holthouse, the other two Catapult founders, him and Igor.

31:59And so no surprise in an investor day that I think it's an analyst day today. That's one of the slides in there to talk about it in that way. Let me talk about this MIT study quickly because you'll find this very fascinating. So this study was basically looking at what's actually going on inside vector space, which is the brain of large language models, let's call it. And like, can we, is it really going to fundamentally displace the need for human beings acting as software engineers? And it doesn't look at it from the perspective of the tasks to be done. It looks at it from this perspective. There are these things called hallucinations, which, if we were honest, we would just say is bugs.

32:49We don't call it bugs because we try to, what do you call it, anthropomorphize? Is that what it's called when you turn something into being like a person? So we try to make AI, we say, oh, we send, I spoke to her or I spoke to him. It's just a software, right? And so it looks at these, because really what's stopping the permanent mainstreaming of large language models to displace workers is these mistakes that they're making, these hallucinations. So this MIT study said, let's look at how this actually exists and let's look at when we're going to get rid of it. And this is what they came up with.

33:26Now, it's a bit complicated. I'm going to say it very simply. So there's this idea of tokens. Now, I really tried to understand long and hard what they mean when they say current llms operate with um 40 000 tokens a bit like the currency of ai so what they mean in this context is there are 40 000 different ways piece there are 40 000 discrete ways that it understands information that's what it means so when you tell it something or when it trains on a piece of data, it will chop up that data into different segments so it can understand each segment of the data and how that data relates to other things.

34:14And at the moment, there's 40 ,000 tokens, 40 ,000 different discrete chunks of understanding that it has. Okay, so that's fine. It's hard to conceptualize that. I feel like I'm not 100 % of the way there, but that's as far as I've gotten in understanding. it okay and now the question is what do we do with those 40 000 chunks of information when we have conversations and the answer is that in an ideal world so do you know what it means in maths for something to be orthogonal no okay it means it's at right angles that's all it means and so on earth we've got three dimensions that are orthogonal length width height that means you can run each one as far as you want length width depth or whatever you want to call them yes fine yes and so you can run them as far as you want in either direction those lines will never touch because they are at perfect right angles so orthogonal why is there a special word i don't know and so and so that's beautiful they never touch but the minute i had a fourth line let's figure about time as a dimension for a second the minute i had a fourth line if i run it in one or the other direction it's eventually going to start interfering with one of the existing lines we can call those three things dimensions so this is why we use the word they not we use the word dimensions llms are now up to 4 000 dimensions what that means is that you can have 4 000 chunks of information that could be interpreted in such a way as they would never be confused with any other piece of information because no matter how far you go in each direction they're never going to interfere with each other but there's a problem i just told you there's 40 000 not 4 000 yeah and so you got 40 000 t-shirts that you're trying to get into a drawer that holds 4 000 t-shirts or maybe we can say a hundred t-shirts you're trying to get into a drawer that holds 10 and the thing about llms if they've developed really smart ways to pack stuff in like they fold t-shirts in a certain way and put them on an angle and jam them in but no matter what you do they're going to be overlapping and so when i say i want to know about a piece of information let's say i'm talking about like a riverbank maybe that when it wants to talk about a riverbank it wants to pull out a white t-shirt but when it pulls out a white t-shirt there's all these other t-shirts touching it that are going to come out a bit as well yeah and it doesn't know it's also pulling those t-shirts out and maybe instead of talking about a riverbank it talks about a riverbank and then it starts talking about a bank like did i go and put money in now that's a dumb example because it wouldn't make that kind of mistake right but like um that's an example of how having too many different types of information packed into too small a space with not enough dimensions means you're going to get overlap with these and then the next question that mit looked at is so can we increase the number of dimensions we're not even talking about increasing the number of tokens that would be fantastic as well the more stuff chunks of information the better it can engage just increasing the dimensions and so it turns out that basically you have to increase the dimensions by massively increasing the size of the model let's just say roughly if you double the size of the model you can double the size of the dimensions and the problem is that we are running out of power and cooling and like money to be able to keep doubling the size of these models like we're running to the end of doubling the size of these models and so that is a fundamental limitation to the operation of these llms and it's extremely likely at the moment according to this mit study very respected that um we're not really going to be able to do much better with these hallucinations than we are right now because of this fundamental structural issue with the way these models are built.

38:17I found that very fascinating. Jason's talking about a million times the number of tokens. So obviously he's speaking his own book to a great extent. And he's using tokens in a different description about these kind of things, right? And he's talking about how many tokens an instance of the model can consume. And so, like, yes, that's true. But if you look at the amount of data sent a new space that's come online in the US in the last quarter, you know it's massively down on the previous quarter. Yeah, it's quite fascinating to look at. So that's a bit of homework for you to do for the next episode because I know you'll be fascinated by that.

38:53But I do think we're in a moment where we're saying basically how can we increase the productivity or efficiency of development teams using AI? And I think we should still retain a high degree of cynicism based on this study about whether we're really going to see this 50 % reduction in staff count if these hallucinations cannot be overcome. Totally. Can I pivot to CGT, one of your favourite topics? Oh, CGT. I was thinking about what could that TLA mean. Do you know what TLA is? No. Three-letter acronym. Oh, there we go. Yeah. Very good. So the Australian government has undertaken a masterclass in taxation reform, gaining support from fellow MPs for a reduction in the CGT, capital gains tax discount.

39:40And while details remain scarce, with the government not confirming whether existing investors will be exempt from the proposed reduction in the CGT discount from 50 % to 30%. And I can't work out, is this like just an investment property thing or any asset thing? That's the question we don't know. No one knows anything, right? So investment property, I think most people would sort of have a degree of sympathy for. I think if we're talking other assets, it starts becoming significantly problematic because it's just a disincentive people to start businesses and employ people. but the government hasn't confirmed.

40:08And while few suggest there are significant issues with that grossly unfair tax system, as usual, the government is finding a way to make a bad problem even worse. Are they talking about, I know I'm interrupting you, but are they talking about, surely they're not talking about removing the discount, they're talking about restructuring the discount. They're talking about reducing it by 50%, as in from 50 % to 25%. Which would be the single dumbest thing they could do with this. There's a few dumbest things they could do, but this is definitely up there. Because before this, I think, am I right in saying this structure was brought in by the Hawke Keating government brought it in in the 80s I think because the previous thing was you basically oh no it was 90s it was 90s you worked out what the CPI was it was indexed essentially and so when you had a high inflationary environment the discount would be higher was it capped at 50 % I'm not sure if it was capped I presume it was capped at something but I'm not sure it was capped at Respect to the Communist Adam Creighton astutely noted that the capital gains that's discount was not concessional as you said during periods of high inflation like currently and paring it back, as the government is talking about, would penalise property and share investors by overtaxing them when inflation is elevated.

41:11Creighton noted, reducing the CTA discount 50 % to 33%, especially in today's higher inflation environment, would see many investors taxed purely on illusory gains. So what do you think the objective of this is? Well, it's to raise money. Do you think that's the objective? Absolutely. It's probably a side objective of making housing cheaper. That's definitely not their main objective. So I feel like you're not being cynical enough. So we all know the order of politician motivations. One, get elected slash stay elected. That's one. Two, get even. Three, everything else. And so you're basically giving me two reasons in group three, but surely the number one reason is they need votes of young people that don't own property, and this is the primary motivation.

41:59Yeah, I think that's right. Because they don't need more money, because the amount of money you spend and the amount of money you collect in tax are not necessarily correlated. Totally. And so - No, you're right there. You're right. I don't disagree with that. And this is the 50 % discount on capital gains, actually introduced by the Howard Government, no, nine, sorry, replacing the Paul Keating era indexing of capital gains. The old labor indexation method was still designed to avoid taxing inflation on assets held for multiple years. The other issue with Jim Chalmers' great attack on investors is it barely scratches the surface of government spending compared to the, obviously, the huge discount on real estate, which is the principal homeowners.

42:32Yeah. So I can buy a house for$500 ,000 30 years ago and sell it for$15 million today. And as long as I've lived in it, I don't pay one cent of tax on that. Yeah, exactly. So even if you reduce the discount on real estate, it only raises$4 billion a year. If we take the discount on all assets, it's only$6.5 billion a year. We're wasting$50 billion a year on the NDIS. $10 billion goes to autism. So more money is spent on autism. and this whole job-killing tax increase will do. But some of that money has to be spent, or I think very strongly I feel should be spent. The question is how much of the 10 bill is it?

43:09Yeah, exactly. But do you agree we're doomed? Because there is no way to fix a tax system without fixing the whole tax system. It's the spending that's the problem. But there's all these piecemeal bits, like there's still payroll tax because fresh fruit and vegetables don't have GST. Payroll tax is going up. Yeah. And there's a GST, which is probably the wrong number. And there's an income tax that probably isn't right. And there's company tax, which is also wrong. And there's superannuation concessions. And there's no minerals. And there's no petroleum. Some people like Santos and Woodside are frantically drilling oil outside the Australian ports and paying basically no royalty.

43:45Well, you know, when they run ads that say, we pay as much tax that covers the whole of Medicare, so you should love us, I would say, yes, but you're digging up my gold from the ground, right? And oil and whatever. I own that as an Australian citizen. At least the gold miners pay royalty to the states. Whereas Woodside and Santos and many foreign-owned companies, like Chevron and all the other guys, pay nothing. It's called the Petroleum Rent Resources, Petroleum Resources, PRRT, whatever it's called. That does basically zero. And government after government leave this thing completely stuffed.

44:17So this is my point. If Kevin Rudd, remember that guy? That was a minerals, the MRRT. I know. if he couldn't even get a tax through to take more money from highly profitable hole diggers, which I always say, like I have the utmost respect for miners because they're the only reason we have a good quality of life in this country. But like, really he couldn't even sell. We want to take more money from rich hole diggers. Okay. And like, so he's got spooked around a campaign. Remember how they were Andy frozen in the back of a truck? Super smart. Yeah. Super smart. And so, and Kevin Rudd was a generationally terrible prime minister as well.

44:52Well, Will you say that about all of our promises? No, I don't. We had this conversation. Well, Morrison, generationally terrible. You've been the worst. Okay. Well, what do you think about Elbow? Pretty bad. So you're three from three. I thought Turnbull was fine. Gillard was fine. It's been a bad generation. We didn't ask us anything episode where we had this discussion as well. I'm definitely not anti-every problem. It's just the terrible ones. I know. Just a few of them are bad. And so the reason I say we're doomed is what elected politician can ever go - That's the problem. to the electorate and say, I'm going to overhaul the whole tax system.

45:25And even if some of you will be overall worse off, some of you will be overall better off, but even you in some way will be worse off, you're only going to focus on the way that you're worse off, that you have to have someone that basically doesn't want themselves or their party to be in power after the next election. I'm not sure it's for, I think the Australian government, the shame people will understand a brave politician who tells it how it is i think there's this i know what brave means in politics if you're a chess minister brave means suicidal that's what brave means in politics i don't think i just think it can't happen yeah well i think eventually you get to the point where you get the pitchforks people get the pitchforks out and you have no choice this is like if you look at this cgt thing it's just ridiculous like but should we focusing on the spending issues the spending on 50 billion on nds and tens of billions on orcas and all these other garbage they waste money on.

46:18That should be the first point of call. Then fix up the tax system after, which I agree isn't perfect, but this isn't the solution. I say this before, I say this constantly, I say it again, I hope I'm wrong about this. It's hard for me to see how the wholesale changes required across Western liberal democracies are going to be happening before some episode of violence. Either an internal war, an external war, like these look like fairly intractable problems in this system. and I think the and like elbow instead of doing this political junk that he's doing now and he's not alone in doing that like he's not the only party doing it but he's in power I think the focus should be on how do we save western liberal democracy from itself at the moment on that note we'll go for a super quick break back with some great business stories in just a moment

47:14and we're back and billionaire rag trader solomon lou said he'll introduce low-cost beauty items at smiggle as he attempts to turn around the once hugely popular kids stationary chain baring a return to growth next year after a disastrous calendar year 25 the company posted a first half net profit of$101 million for the December half. But once the disposal of a major stake in Meyer was accounted for, profits actually fell by 13%. Smiggle was an absolute horror story. Sales fell by 11%,$140 million for the six months of January. Sales at the store peaked at$320 million back in 2013. So that's more than 50 % down, Smiggle, which is a disaster.

47:54Premier shares briefly peaked at$27 in 2024 and have slumped back to only$12 now. with Lou's net worth being only$2.8 billion, according to Bloomberg, down from almost$5 billion in last year's AFR rich list. Lou does, of course, own a couple of very, very successful, hugely profitable private businesses, including Seed House, all kinds. So Soli's doing pretty well for himself, so I don't have to worry too much about him. But if you look at Premier Investments as an entity, it's had a really tough time. Certainly, I was bullish about this business about nine months ago when we talked about it. You weren't bullish about Smiggle.

48:25No, but I was bullish about Peter Alexander. Well, I think you can still be bullish about Peter Alexander. Even Alexander only grew 4.9%. So it's not... We are pretty close to the feeling of a recession in this country. Yeah, for sure. And consumer confidence now is terrific. So I don't know what this... This is before around. This is December last year. I know, but I don't know what their same store sales growth was. But it's not an easy time in retail right now. I think Smigel is one of the few times where I think Solly didn't get the timing right. Yeah. He should have offloaded that business.

48:57or spun it out to its own thing and kept a bit of it and sold it down. I think it's hard for me to see what is going to make that brand special and unique again, but it does have a brand. So it's got an opportunity, right? It's a very commoditized product. The brand is, I'd say, on the weaker side of brands. Yeah, but it's more well-known than lots of other stuff playing in that space. And so it's tried licensing. That hasn't set the world on fire. I did a Minecraft license. That's right. I think, you know, what people – so you can sit in an office and look at numbers and be a finance person and say, we've got to do licensing.

49:35Let's try licensing. And then you can be a commercial person and you go and do that licensing deal and you get the terms right. But then it comes down to being a product person. And if you make ugly product, no matter how good the finances are and the deal is, nobody buys your product. and when I've looked in, I'm not sure if you're going to be happy about me telling you this, but I basically stopped people coming out of Smiggle and asked them what they think about the stuff because I thought it was really ugly. Serial pest, basically. Yeah, and basically people were like, yeah, it's not good.

50:07People didn't like... We bought Smiggle stuff. I've had no issue. Did you buy their Minecraft stuff? No. Most of it's not very good looking. Lunchbox, maybe? Maybe Minecraft lunchbox? I think that if they would have... I think their main opportunity, maybe this sounds obvious, but I think their main opportunity is actually on the product side. I think they used to make this quite differentiated product. There's too many competitors now doing the same thing. I know, and so what you need to do is, if you're running in a race, I'll do something that you'll understand. If you're running over the Balti Bridge with no shoes on, dodging potholes, and you're ahead, but someone is running fast behind you, in order to not be overtaken, you need to speed up.

50:46And it's not a hard concept, and I think Smigel, they never sped up. They never found the next thing. They lost John Chester, who was the previous CEO. Maybe that's the problem. Come on, you can't be that thin on talent that if you lose the CEO, it's over. Well, they've had real trouble filling it. They haven't filled the role yet. So there's some pretty serious issues in this business by the looks of it. As much as you can say colourful plastic is innovation, let's call it innovation, like they lost their innovation edge and people caught up to them and they didn't excel. Well, it's not like they weren't designing Vera Rubin chips here.

51:18They're designing lunchboxes. I know, but they still found a hole in the market. I don't think they started Smiggle right it was acquired is that right I bought it when it was tiny so someone had this good idea and then Soli the master of scaling scales it up ultimately when I was taking my kids in there six or seven years ago my daughter she found stuff she loved that was different and so we can say okay it's not AI chips but it's better not to make AI chips it's easier to make this stuff in China I'd rather own NVIDIA than Smiggle you would but you didn't want to have to go on the journey to get to where NVIDIA got to and like this is capital intensive and so the problem is there's no competitive advantage no differentiation well there was there was product differentiation but that's not that's not competitive advantage though that's just some that's some operational excellence well you know what you're right basically I would call that there's no corner and resource there that's innovation and they had the opportunity to build competitive advantage long term competitive and they didn't in brand in terms of brand in anything in whatever what else could they build they couldn't build there's nothing else I'll tell you what you would like me to say you would like me to say process power yes they could have built process power in fundamentally making because they had no process power.

52:21They could have bought other stuff that they thought, this would be really good running through a Smiggle brand channel, for example. I don't know. I'm not solving their problems here. What I'm telling you is they had an opportunity to strengthen the business. I think your point before was right. The Soli should have sold it five years ago when it was flying. Well, that's definitely right in hindsight. Yeah. Pretty easy in hindsight. Yeah, everything's easy in hindsight. I don't think it looked terrible two years ago. I think it just started two years ago, right? Yes, exactly. Exactly. It has been a tough couple of years for Previa.

52:51The whole Maya thing, jury's still out there. The markets are still unsure. I think Peter Alexander's a big opportunity, personally. It's just not growing anymore. That's what inflation is. It's basically growing in inflation plus a percent. But again, there's all sorts of product complaints coming out of that business. Go and read online. Quality complaints. Are you just defeating your own argument? No, what's defeating my argument? They have to fix product. Smiggle, make better products. Peter Alexander, fix your product. I agree on the product issue. I just don't think that's going to make that much difference.

53:18I think the ship's out on this stuff. You know, if you're charging a premium, so Peter Alexander charges a premium. And so obviously I talk a lot about brand equity nowadays. So Peter Alexander is not going to be able to maintain what's called willingness to pay a premium if the perceived quality of the product diminishes. Yeah, absolutely. It takes a long time to cause that damage, but I think they might be ticking that box now slowly. That's very problematic. They have to fix the product quality perception. Yeah. So I'm not – what's the market cap of this thing? Like what's – tell me some metrics about this, other than that sales have gone nowhere.

53:54Share price is$12 now, so about$2 billion market cap. Yeah. It's dropped off from its high of – I briefly touched at a 28. That was really brief. It sort of reached – I always go to pre-COVID level, so it's basically about 15 % below pre-COVID. What's the – give me like an earnings number or something on this thing. It's a half yearly, right? So you don't have the full year earnings number. Looking at their numbers now, so their total revenue down 1 % to$460 million for the half. Profit was basically flat at$101 million. NPAT. Yeah. Yes, NPAT for the half. Yeah, yeah. And so - So it's making a couple hundred million bucks and capped at$2 billion.

54:36So it's not expensive in that sense. But it might not be making a couple hundred million dollars because it's retail, super seasonal. And what is the leasing issue? Oh no, the leasing issue is probably captured. No, that'll be, well, not really. NPAT is pretty bad for capturing leasing. Of course it is. They do talk about lease rental expense in the P &L, so I'm just not sure if that captures all of it. And no, they wouldn't be capitalised in tech or anything like that, so it's not that sort of issue here. So when I browse through this, it seems like it's trading on about 14 times earnings. Before tax.

55:10No, PE. That would make sense. That would make sense. Yep. And so what do I think about this thing trading on 14 times earnings? It's not cheap. It's not cheap for a business that is delivering flat impact. Yeah. And has one business, Smiggles in real trouble. Peter Alexander's not in trouble, but not what it was. I'm going to assume they don't have a bucket of cash on their balance sheet that's accounting for most of that$2 billion. Yeah, they only got$360 million in cash and$180 million in current liability. so yeah it's not cash laden by any stretch so basically what you're being asked to do is to buy a business that's got one growing a bit growing part of it one going backwards part of it there's some other stuff in here as well there's breville in there breville thing which is actually doing pretty well and so would you pay 14 times you can buy breville directly as well yeah i think my view on this is it doesn't feel cheap yeah you can't short solly no you never never short it's too risky to short it.

56:11I think this is a wait and watch. Yeah, no, I agree. I'm generally bearish, but not to the point that I have a short it now. I think you're a brave person going against a great retailer like Sally Lou. Can we move on? Speaking of great business, Australia's most valuable startup, the still private graphic design business, Canberra, finally lodged their financial statements for 2023 and 2024. This is the calendar years, and it makes for very interesting reading. The US$42 billion business delivered outstanding top-line revenue growth in 2024, 45%, hitting$2.1 billion, almost all of which is subscription revenue, of course, so very high-quality revenue.

56:49Is this all of their revenue? We're not sure. This is the Australian financial statements. It could be higher. There could be some other revenue. They claim it's$4 billion now, AR. It could be. So if you look at growth, that implies we've got most of it in here. So let's say it's grew another 40%, hits$3 billion, and then looking at AIR rather than sort of caught actual revenue. So I think it's most of it, but not all of it. I'm going to say something just back of the envelope, but like if it's four bill USD and it's growing at four… I'm quoting these numbers in Australian dollars. Four bill AUD.

57:24Okay, whatever. It doesn't actually matter. If it's four bill and it's growing at 40%, then it grew from what? 2.8 or 2.9. And so that means their revenue is probably mid-ish threes for that 12-month period. Well, I told you the revenue because I got the revenue until 2024. No, but that's from 2024. Yeah, so I don't have last year's revenue. I don't know if you're aware of this, but we're in March 2026. We don't have last year's revenue. What we do know, they've claimed$4 billion AR now. So we still know that. So I'm trying to figure out what I think the trailing 12 months might be. Yeah, I think three, whatever, three is about right.

57:54I think the trailing 12 months might be 3.4. Okay. That's what I think it might be. Yeah. Which is quite a bit. I don't think, that would be, you see, You're saying it's gone from 2.1, 3.4 in a year? No, I said trailing 12 months. But we're only three months past the end of the – this was calendar years, not financial years. So this is the year to December. So yeah, I think it might have grown. What was it then, 2.8? 2.1. Do I think – You're saying it's grown 70%. No, I don't think that's 70%. I'm saying it's grown – 1.4 on 2.1. So 1.3 on 2.1. That's 61 % growth. So what do I think? if you say to me do I think I could have achieved 61 % growth in 15 months not 12 yeah maybe like I think so and so I think that could be the case but whatever it is we can assume that there's that it's somewhere north of 3 and so I think the top line growth is very impressive and the least of any problems they could possibly have and it's subscription revenue almost all so it's good there's a lot of enterprise revenue in there now so a lot of big ticks there and we're assuming these are all AUD given it was obviously filed with ASIC.

59:02So cash generation increased from$1 ,3 million to$263 million and the business ended with$700 million in cash. So what does that mean? They're making 11 % or 12 % cash on revenue? Yeah, but let's not worry too much about cash. Well, you don't like cash? Well, I like cash. You send all your cash towards me if you don't like cash. I like cash. I'm happy to take any cash you don't want. With an asterisk. So the business had 220 million active users up and won 70 million. So that wasn't – it was good user growth, but not – so there's a lot of expansion revenue there is the problem. Well, not problem.

59:33It's good to have expansion revenue, but the problem with expansion revenue is you can only expand so much and you start getting sort of hitting a wall. So you'd rather have – So it's a 32 or 30 odd percent or something. And also there was lots of good news. The financials were far from a slam dunk. Expenses rocketed in line with sales. That meant the business recorded a huge loss of$241 million in 2024 up from$232 million in 2023. co-founder cliff obrex told a afr last week co-incentedly that we're growing revenue at over 40 which we agree with at rare scale this is our eighth consecutive year of profitability on a free cash flow basis and something only a handful of software companies in the world have achieved in response to a question i don't think they have to sell their book that hard to be honest with you like this is such a good business i think um modest i mean the us that doesn't typically love modesty as a concept but i do think a bit of modesty with this kind of business would actually sell it harder uh in response to a question from uh jason lemkin in his uh podcast last year we should just provide some perspective the fanboy of all fanboys for sass literally the founder of sesta their whole conference and ecosystem of sass businesses well i'm not talking about what jason said i'm talking about what cliff said to him um it's gonna be a pretty friendly podcast Oh, no doubt.

1:00:52And Obrecht said on Lemkin's podcast that the company was not only cash flow positive, but also profitable. Well, that's good. You have a billion in cash, you're profitable or cash flow positive. I don't care which one. Probably both, Lemkin said. Both, Obrecht replied. Well, you know what? That's great. Well, it's not great because it seems like Cliff's been spending way too much time with Scott and Mike, his mentors, and he's trying to claim that cash flow profitability is real profitability when it's not. See, this is the problem. The numbers you've got, what was the last period included in those numbers?

1:01:22The 2024. But he's saying it's been profitable for eight years. June 2020. No, he didn't. He's saying it's been profitable on a cash basis. Yeah. We could rephrase that in accounting speak. In the interview, he's claiming it's profitable now. Now. Now is March 2026. No, this interview was like months ago. Okay. December 2025. That's a long way after June 2024. So you're saying the business lost$230 million in 2023, $240 million in 2024, and suddenly it's profitable. I'm saying that's what he's saying. And by the way, if you're making three - I think he misspoke being generous. Well, I don't think so.

1:01:56I mean, he might have, but if you're making three bill of revenue, you could easily turn around minus 200 to plus 20, and that would be profitable. They've talked about being profitable repeatedly, and they're not profitable. You read me a quote that said, we are profitable on a cash basis. He added that recently. But the last thing guys would say, about being profitable we're profitable on a cash basis no they talk about their non-gap bullshit and now the market has worked out Atlassian and said no you guys are completely for shit and we're going to mark you down 90 % Atlassian's problem is the dilution from share based payments was too high and also eventually caught that and also the nature of the share based payments got the market offside but I was saying for years about it and everybody just ignored me and suddenly they didn't well you're very avant-garde Adam you've got to get used to the same people are going to be a bit behind you and you're behind me now on campus I might be or I might be Or you might be wrong.

1:02:47I could be wrong, but it probably not. Those moments somehow seem to fall into the amnesia state with you. In 2024, Canberra spent$356 million on share-based payments to staff, up from$282 million in 2020. This is a lot of dilution, to your point just a minute ago. It's a lot of dilution. This is huge amounts of dilution. Well, hang on a second, just to challenge that. Let's call it$350, an evaluation. Is that USD or AUD, that$350? AUD, I think. Okay, AUD. A valuation 60 bill? Yeah, at the moment. So you can work out what that is. Yeah, but look at Atlassian now. Half a percent. Look how much dilution Atlassian.

1:03:23You said, oh, Atlassian dilution is nothing. It's nothing. Forget about dilution. Now the share price has dropped 90%. It's only dilution is actually quite a lot. That's what happens when the share price is overvalued. So that could be the case. I think this business, do you think this business at its best looks better than Atlassian looked at its best? I was thinking about this this morning, actually. I don't think so. It's a more well-known brand, but it's a consumer brand. Atlassian was immovable at its best. Like, they were just growing. Well, Atlassian made money for a while, which was. But that wasn't even its best.

1:03:58Like, I think its best was post-IPO when it looked like it could pull a few levers and start generating tons of cash. It just didn't, right? But, like, at its best at that moment, it was growing like crazy. It was not having to aggressively sell. sell it hadn't really pulled the lever yet on price increases it felt when you talk to people about it like i know you always say you hate it but like people yeah it's a necessity you couldn't get rid of it like i think that business did not feel worse than canva at its best the tam might not have been as big i think what alassian does how to know what tam canvas tam is how do you even measure it the thing i like about alassian more than canva this is the one thing is alassian kind of has the market to itself yeah there's monday.com and some other stuff but really I would say TAM is a total addressable market.

1:04:41The beauty of Atlassian is there aren't any real competitors. Like, they're all pretty irrelevant. Whereas Canva's competing against Google and Microsoft. Their competitors are all the hyperscalers plus Adobe. Adobe and Figma, maybe. No, you're right. So it's got these amazing competitors. Even forget about Figma. Like, if I say, who would you not want to have as a competitor? Maybe Alphabet and Microsoft. Well, you've got both of them. And so that's not great, right? And remember when we talked about - But they're doing a good job against them. Remember when we compared Canva to Figma, and we were saying and we said oh you were saying and I somewhat agreed that Canva's valuation of 60 or 40 million US was more justified because Figma's at 10 but Figma's not making money and we said oh Canva's making money so Canva's better but Canva's not making money either so if Figma's 10 billion I think you should be less Canva's 3x Figma I reckon in terms of sort of scale and is losing I think the same or if not more money or roughly the same Figma's down to 10.5 billion so I would have thought that USD USD alright This whole long soliloquy Not that long, you've been interrupting constantly.

1:05:42Gets us to, instead of being worth 42 bill USD, they should be worth 33. 20 to 30. No, you said three times. You said three times Figma. Don't change it now. Yeah, but I've looked at the reflective profitability so I've got Canva's latest number. You know, if you generate, so they've got great gross margins, okay, Canva. If you generate this much revenue at great gross margins, You can... But they've got a lot of costs in there. I know, but you can throw out 500 mil pretty quickly. Like, you know about their costs. You used to have to last in every year. If they only last in, stop tiring and stop doing this, stop doing that, and make a billion dollars, they'll never make a billion dollars.

1:06:18But they could have. Well, we say that, but I'm not sure they could have. No, you're changing your story to try and have an argument with me. You know they could have. You argued that they should have. And, like, you know Canva, what? They fired some people out of Leonardo, right? Yeah. Okay. Was it, what, 20 people or something? Yeah. Like, who knows, right? Yeah. So they haven't really fired anyone. No. And they're still hiring a lot of people. We know the game that they're playing here. They can adjust that game if they want to. I'm not anti-camber. The question is do they want to adjust that game?

1:06:45I'm not anti-camber. I just think, I hate the fact that I go talking about profitability and we're profitable. We're not lasting. We're profitable. These guys aren't. They're talking this cash flow bullshit. Stop talking about this shit. Nothing makes me more annoying when they're not lying to people. They're just not being truthful. Well, so what I think is this. most of us unfortunately do not get the luxury of running a business as great as canva in fact none of us gets that opportunity and so i think that like when you are running that business you shouldn't talk the way the rest of us have to talk you should talk in some exalted much more modest way buffett style yes because it's going to get you more mileage totally you don't have to convince anyone that you're great we know yeah what you should convince us is that you are so great that whenever you say something, we should assume it's even better than you're saying.

1:07:37And I think that's the mistake they make with their public pronouncements. And look at the balance sheet. It's far from pristine. They've got $700 million in cash. I'm thinking of last year. Well, that's handy. $870 million in cash-like assets. They've got$900 million in good assets, let's call it. Debts of$1.4 billion, though. So this is not a business that's... Could be old convertible notes. I don't think so. It's racked up losses of$1 billion over its lifetime. So again, not Atlassian$3.6 billion, but this is a chunk of losses. From a financial perspective, it's looking a lot more like Atlassian than we thought it was.

1:08:12And Atlassian's seen its business value drop from$100 billion to$16 billion, playing this game of we're making profit, but we're not really making profit. No, I think it's hard to disagree with you on that point. I think you would have thought a business like this would have looked at what Atlassian has done and said, that's one way to go and it was good for this period of time. Let's be honest. It was certainly good for the founders selling down. Atlassian founders? Yeah. It was really good. Like that model worked, okay? And the investors wanted that. But they've pummeled Atlassian now. And I think, I keep saying to you, we've seen data until the end of June 2024.

1:08:48I think it's calendar year 24. All right, December 2024. And so they've had now 15 months to do something different. We don't know what they've done. It's opaque. They're not going to IPO anytime soon. Delaware, so that, like, again, like the fact that, and again, I love the fact that these guys reported so late. These are a year overdue, these numbers. I'm much less cynical than you about where this business is, and I'm much more positive about their ability to go and generate a billion dollars of profit. Remember that song, Don't Stop Thinking About Tomorrow? Don't stop thinking about tomorrow.

1:09:22Ah, that Bill Clinton used. Yeah, yeah. And Atlassian had that, like, Atlassian always said, we're losing money now, but we can make money in the future. Give us great valuation on the fact that we will make money now. We're going to keep investing. They didn't invent that idea. But they kind of were the poster boys. Have you heard of the dot-com boom? That was a big thing then. How did that turn out? Very badly for most companies, but well for Amazon. One company. Well, not just Amazon, but they're the poster for it, right? Right, right. And the others. Because they said, don't stop thinking about tomorrow.

1:09:50And what does Amazon say today? Don't stop thinking about tomorrow. No, but AWS makes heaps of money now. Without AWS, which is frankly, I think Jeff Bezos says it's the luckiest thing in business history. like without AWS, Amazon would have been in a pretty bad position now. So like it's, and so look at the last year and they were, we're going to make money tomorrow. We're going to keep, forget that, we're losing money now. We're building this, you can say we're investing in scale and we're going to make, and it feels like Canva's been doing the same thing. And last year ran into the brick wall of dilution that we've talked about.

1:10:17Canva's ran into the brick wall of Microsoft and Google. So they've got, and they've got their own problems. This works well in a world where Canva has no competition and I'd say that's the right strategy. In a world where they've got a maelstrom of competition. I'm not sure this is the right move they made. They should have been focusing on real profitability and I'll be singing their praises, which is what I thought they were and which is why I did sing their praises. Well, you're right. There's definitely a risk attached to it. You've got this song in my head and I don't stop thinking about it.

1:10:46It's like a Fleetwood Mac song. It is Fleetwood Mac, yeah. There's a song where I think the B-Naked Lady is called Sell, Sell, Sell. I would be telling Canva they should be thinking about that song, not about the Fleetwood Mac song. Yeah, I think, listen, I've said it for a while. Like, it's time to sell this business. I wouldn't IPO it. I would go and find - Who's going to buy it? Who can buy it? Microsoft alphabet. They can't buy it. They can't buy it. Because you're worried about antitrust. It's Trump is in power. Yeah, I'm not sure. Give him a gold clock or something. I'm not sure Trump likes these.

1:11:20What did Tim Cook give him? I don't know. A gold something, remember? I thought it was a gold clock. I'm not sure. this is going to be able to get through. I don't know. I think, by the way, Adobe, who can no longer afford them, could now, I know they couldn't buy Figma, but the world has changed. No way Adobe can buy. The world has changed a lot. No, you've got EU issues, don't forget. It's not just US, you've got to get through this through the EU, which is they're the biggest socialist in them all, so I can't see that happening. Maybe a massive private equity firm? What are the, yeah. If you were, but private equity's not paying you 42 bill for SAS today.

1:11:53No, forget 32. If you were Clifford Mail, would you sell for 30? Yes, of course. So I think probably they could pay 30. Yeah, but they weren't like an intergenerational business, I think. So you think they're never going to sell? I don't know. I don't know. But I think that's my feel based on what they're saying. Yeah. So I'm not sure about that. But I definitely think – well, let me ask you this. Let's do the ultra bull case. Yeah. What are the chances that AI becomes completely commoditized on the model side, these frontier models? and their brand and user base, and they are very astute. Someone said to me about Mel, someone who knows who I really trust, said she's the best 20-year-out thinker that they've ever come across.

1:12:35And Cliff is really, really good operationally. So what if, could they be the big winner in the consumer design space in the era of AI? What's the chance? Why would they not be commoditized like everyone else? Yeah, it's a good question. I don't know. The bull case doesn't really resonate with me, but I think they're very good at confounding sceptics who always call their demise. And they build a great business. They raise money against the odds. They're being amazing leaders. If I could get 30 billion, I'd be running for the hills. The question is, would you pay 30 billion for this? If you had 30 bill, would you pay it?

1:13:08No way. But I think you can make a billion dollars of profit out of this. Do you want to pay 30 times earnings? No. 20? I'm going to be too worried about Microsoft and Google. So you're five? I'll pay five. Oh, and you won't get your price eventually. So somewhere between$5 and$30 you pay? Something like that. Probably between$5 and$25. So you're not sure where it is. It's somewhere between$5 and$25. Big range. Still a great – 25 is 50 % more than Atlassian, don't forget, at the moment. Yeah, I hear you. I hear you. I know it's an interesting question. It's pretty remarkable that on its current valuation, it's 3X Atlassian.

1:13:42That's just staggering to think. I feel like – yeah, it's amazing. I feel, so I'm very much a believer that Canva is extremely exposed to AI disruption, like one of the most exposed. But I haven't heard a lot of people express that view. And definitely the company talks about how they're an AI winner. Okay, well, I'm not shocked by that. But like they might be. But I just can't figure out whether what's super obvious to me is just something that other people aren't seeing or if I'm totally wrong about it being super obvious. But to me, it's super obvious. This is not trading on the public market.

1:14:14The valuation is opaque. the financials are opaque we're sort of we're going off a 15 month old data here so that said I'm much more disappointed than I thought seeing these financials I thought they'd be generating heaps of profit real profit not this sort of make believe profit the Scott Mike form of profit I thought it'd be a real profit I'm disappointed it's not but anyway they have built a great business we don't want to disappoint you that's always a recipe for disaster thank you I dear great episode as always we'll be back on Saturday with a very special episode we'll see you then

1:14:45Thank you.