OpenAI Burns Billions, Temple & Webster Gets Crushed

0:00I'm Adam Schwab. I'm Adir Shifflin. And this is The Contrarians with Adam and Adir.

0:09And we are back, episode 235. I think we had an incredible episode with the great Vicky Medbeck that we dropped on Thursday. So if you haven't listened to that audience, please do. She is probably the best negotiator in the world, if not the best negotiator in the world. And there was sort of our advice you'd pay 20, 30 grand for if you're a big business. So that is you don't get much better content than that. Well, I mean, I spend a lot of time with pro athletes, billionaires I spend some time with. I don't fanboy any of them, but I reckon like, I like semi-fanboy Vicky because like... Semi-fanboy?

0:44You're a full fanboy. Full fanboy, yeah. Her advice has pretty significantly changed the way that I behave in negotiation situations. So anyway, people should definitely listen to the episode. And where she has added so much value to us is we have her speak to our call center for 15, 20 hours of classes. And we see conversion bumps of sort of 4 or 5%. We're talking tens of millions of dollars of difference in revenue for us. And millions and millions of dollars in profit is the impact she has. So if you do want to get big into your business, she comes to Australia every couple of years. Just DM me on LinkedIn.

1:22I'll have an intro. Obviously, she doesn't work with every business. She's pretty picky on who she chooses. But obviously, she loves Catriona's listeners. so we'll do our best to put in a good word, but just DM me on LinkedIn is probably the best way to hook up with Vicky. Let's move on because we've got a couple of really big business stories to talk about that's happened the last week. I want to talk first about OpenAI. So in what was a really sort of unusually bad announcement because there was a lot of talk about OpenAI being back in the last couple of months and Sam Altman had been really bullish about its progress, but actually reported last quarter's revenue, which is obviously the June quarter, grew to$6.7 billion in three months, up from$5.7 billion in the first quarter.

2:02Now, for any other business in the world, this would be an unbelievable result. So, you're going$1 billion, which is 20 % in a quarter. That's usually incredible. This is sort of 100 % annualized. And it's a pretty big but. OpenAI's operating margin sank further into the red, pushing the company further away from profitability ahead of its alleged IPO. So, the 17 % growth was actually 90 % annualized growth. Again, really good, but a fraction of what Anthropic is growing at, which is a bit of a problem given how these two companies are compared. Anthropic, of course, grew revenue. How's this idea?

2:36Anthropic grew revenue quarter on quarter from$4.7 billion to$11.5 billion, quarter on quarter. It's crazy. What do you reckon Anthropic generated a year ago, same quarter? So Q2, 2025. 500 mil? Yeah. Or you call it 787 million. So 787 million to 11.5 billion in one year. This is just 14.6 times revenue in a year, which is, if this is true, it's extraordinary. But you know, you can feel what, I don't mean you can see, you can feel why it's happening and you can feel why I think open AI is happening that way as well. Like the entire world, I'm trying to think of something that had this level of uptake in my lifetime.

3:23time nothing like there's nothing like the internet was much slower than this i remember 94 95 96 97 98 it was super slow super slow because you have to build all this stuff and this is this is kind of built and even in like you know because people might say well this didn't start in 2026 it started in you know 2021 or something okay but that's only five years well anthropic was only like three or four years old and topic because the guys left open ai that's that he's literally three years old i know but even if you say well you know people started using the like um netscape navigator for the worldwide web i think was 95 and so yeah even if you say that yeah yeah that's the starting point okay what about 2000 2001 2002 the internet was not growing at this speed like uptake of the internet like this is there was like amazon was growing super fast in in the late 90s but i don't think it grew 14 actually no i think i think amazon had one year of like 28x growth but so amazon was going pretty quick but as an industry like i loved the internet so much because i came from bulletin boards and like yeah i i remember exactly how it grew like two people were not 2002 i don't even think most of the major corporations had websites they didn't even own their domain names like i remember what uptake was and it absolutely was not everybody's talking about the internet saying i've got to be on it a lot of people were dismissive of it no one's dismissive of ai everyone is like i gotta be on ai that's all people talk about well i think that view yes yeah but the gloss around ai from a user application perspective has certainly come off a bit like we're not like if you think think like a year ago every public company had to talk about AI 50 times an annual report or they'd be punished.

5:14Now people are punishing people for talking about AI too much and not seeing the upside from this huge token investment. So the narrative, so we've got issues at the data center's point. We've got issues across the whole stack of AI. That's a commercial argument. What you just said is a commercial argument. And let me ask you this question with AI. Do you know a single person that doesn't use Clawed? I know a few people. or Claude, or sorry, let me say like AI chat bots. I think there are people who use it scantily. Like I think there's power users like you and people who use it two, three hours a day.

5:49But I think you're the exception. I think most people use it. And I put myself somewhat in this category personally. Obviously our business uses it heaps. But I use it largely as a better search engine. I don't use it personally for operating stuff. I know I've sidetracked you in your conversation about OpenAI. No, we'll get back to it. But I want to say even when I – so I was like doing some stuff about improving the gut microbiome. It's like a big topic now, right? And so I wanted to understand which products I should buy in the supermarket, e.g. like which pickles actually are fermented versus are just full of vinegar and are not really fermented.

6:27By the way, the answer is if something's not in the fridge in general, it's not actually fermented. It's not universally true, but it's pretty true. And so I use Claude for that. I just take photos of products. Yep. And I just say, tell me where this fits into my plan. To me, that's a better search engine. No, I'm not searching. I'm saying, you know what I'm trying to do. I discussed what I was trying to do. No, but I'm saying it's a better search engine. I'm not saying it's a search. I'm saying it's a better search engine. It's a search engine that has some intelligence behind it. I'm not searching.

6:55What am I searching? You're essentially saying – I've got the product in my hand. I don't have to search. I mean, it's there. It's giving you an answer. Like, search engines give you an answer. This gives you a better answer. So in the realms of, as opposed to agentic stuff or like we look at, we just had a hackathon last week at work. We had 50 teams. I see what you mean. Each of the 50 teams are doing incredible stuff with AI, literally deploying stuff in two days that save hundreds and hundreds of hours or create millions of dollars of value. So to me, that's the real upside of AI. And that's obviously a lot of businesses do that.

7:25When you say, do you know anyone who's not using it? Yeah, lots of people are using it. But if you're just typing a search into Gemini versus into Google, yeah, it's giving you probably a better answer. but it's not markedly different. I don't want to include Gemini results in Google. Put those to the side because they're involuntary. But do you know anyone that's not using it at least once for an hour a week? Do you know anyone that's not using AI for an hour a week? Yeah, I'm sure there are lots of people who aren't using it that much. But I'm asking if you know anyone. Like anyone in your universe where you – Well, I don't ask people but I strongly suspect a lot of people I know are using it a lot less than that.

8:01Interesting. What about you, Mike? Definitely people like my parents and all of their friends, highly unlikely they'd be using it at all or even familiar with what it is. So, like, I think I agree with what Adam's saying. I find that totally shocking. Like, why don't you help them, Mike? Like, why don't you go and show them what this thing is and help them use it? You've been at DEF CON. You're hanging out with nerds. It's different people. All right. Well, by the way, I don't know if you've noticed, but nerds rule the world. Not denying that. You know what it is? It's kind of like Facebook. when you're a kid and you have Facebook, you think it's awesome.

8:32And then your parents start using Facebook and it becomes the lame thing that your parents do. Kids don't want their parents using AI. I mean, I don't know. Anyway, talk about open AI. I don't know. I think you're both crazy, but anyway. We're talking about, so obviously open AI has had, I would say a pretty good revenue result, but relatively Anthropic, not that good. But if you look at Anthropic itself, I claim they're about to generate its first ever operating profit of 500 million bucks on the back of this massive revenue increase, the 14.6 year on year. However, if you listen to what AI skeptic Ed Zitron said, and I love Ed, and he characterized Anthropics reported$559 million operating profit as an absolute swindle and an accounting trick designed to hype investors for its mooted$2 trillion IPO.

9:13Zitron argued that the profit relies on artificial cost cutting through heavily discounted compute power, masking a reality where revenue is outpaced by massive infrastructure costs and unsustainable cyclical investment from unprofitable AI startups. So if you go back to OpenAI, which appears to be certainly in a worse position from a profitability perspective, the Wall Street Journal reported the company's operating loss, which of course includes stock-based comp, widened from$9.3 billion in the first quarter to$12.3 billion in the second quarter, obviously outpacing revenue growth. So if you look, it generated$6.7 million in revenue and lost$12.3 billion, which is pretty extraordinary.

9:48So that's sort of giving away 50 cents for every dollar. pretty much. So you've got Anthropic allegedly making profits, but I don't think anybody believes it's real profits. And you've got OpenAI literally incinerating cash. And I think OpenAI's product is actually getting pretty good, to be honest. It's pretty comparable to Anthropics. I think it will start shipping away at Anthropics' growth rate and start catching up. But you've got one allegedly, but probably not profitable company, one massively loss-making company. You've got companies en masse turning away from this token maxing and just trying to reduce how much they're spending on AR because they're just not seeing the returns.

10:23Where are we with this whole thing? Well, I think it's very unlikely that anyone has any idea where things are going to be in five years' time. And I don't say that because, like, you can always say we don't know where it's going to be in five years. But on the main episode, we spoke about GYG. I mean, I think we know approximately where GYG is going to be in five years. Like, there might be some things here or there, but we know approximately where they are. And so the reason I say about this is there are just too many unknown variables involved in this. Like you flagged one, which is what's going to happen with the propensity for data center creation.

10:59Are people going to accept it? Are they going to want it? Is there going to be the energy? Are we going to see chips that use so much less energy and more efficient? So that's one issue. The second issue is just like the cost of the chips themselves. Are we going to see a huge drop with all this competition that's coming in? are businesses going to be able to find ways that they can monetize effectively off the back of this? That's a huge issue that you flagged at the moment and like at the moment what we're seeing is very early and so it's unsurprising that businesses are not getting an ROI return on investment from tokens now because mostly they're using it to mess around inside the company and their clients are unprepared to pay a decent price for agents really I mean in the most part.

11:42So I think it's just such a… Isn't that the issue? Well, that's one of many issues though. And so that might change though. Well, the main issue is the entire stack has problems. So if you look at what you talked about, you've got the application layer, so companies using the models. And at the current levels, they're cutting back spend because of the cost. And at the current cost, certainly the industry is losing a heap of money. So the only way these businesses become significantly profitable is by increasing the token cost. and even before they increase the token cost companies saying we're not going to pay this token cost so we're not even close to being profitable and maybe but the cost of compute might come down dramatically and this is the overarching view i want to share all of this that we've just said is true this is a mega hyped industry with massive problems coming down the pipe that is burning through investor money and so all of that is true but the flip side is and this is why I started off with that whole conversation which kind of went nowhere with you guys which is how much do you use AI but but but I can tell you and I'm sure that you'll agree with this even though you don't use it the way I use it like this is an innovation that you can't live without it is such a good innovation and what it does is so incredible that we're gonna have to use it we can't not use it I wish my competitors don't use this every day I pray that my competitors don't use AI.

13:07And so somehow, we're going to find a way for this to be a profitable industry. It's just unclear how it's going to be a profitable industry at the moment. And if you are going to bet who might be the winner when this becomes a profitable industry, like we have to make a bet. I certainly wouldn't bet on an ETF that buys everything involved in the industry because that is, I don't think, the answer to it. I certainly wouldn't bet on open AI for the reasons we've discussed previously because they're just like software trying to sit on everyone's stuff the companies to bet on are hyperscalers right you bet on alphabet you i think you bet on meta personally you could probably bet on amazon as well like they're the companies to bet on i think eventually they will make money from this person time will tell we'll go to a super quick break back with our next story just in a moment

14:03and we're back and friend of the pod uh templan webster shocked the market last week announcing what can be described as a relatively good profit result well in some respects but a pretty horrendous trading announcement hang on come before before you go on for even one second like what the what does friend of the pod mean exactly because i tell people we're like really good friends but now i'm very nervous about using that terminology like we're not against temple and webster we want them to succeed were you was that an attempt at humor by you when you said friend of the pod i think in this case there's a little bit of irony but i think in this case friend of the pod is someone who we talk about a lot they give us a lot of content hence they're a friend of the pod so there are actual friends of the pod like john stanshaw and joe astrid to come on all the time.

14:46And then there's, I call it my somewhat ironic friends of the pod who we talk about as our muses and certainly Temple is a company we've talked about more than any other I think. Well, you should call it Fixation of the Pod, Temple and Webster. Possibly his better name. But I will say, I do want to say because I feel so bad whenever we talk about these businesses and we haven't talked about Temple for a while because I didn't want it to feel like we were fixated. The thing is, I'm just going to reiterate this in the most defensive way. you and I, we really want Temple and Webster to succeed and we really like the people involved in the business.

15:18We've just been long-term bulls because the valuation was preposterous and I think that I understand this business as well as I understand almost any business on the ASX. I had a business that competed with these guys and these guys wiped the floor with us. So there's no question. And we talk about like some businesses are pretty easy. Last week we talked about Guzman and Gomez, really hard business to run, which is why Stephen's such a generational CEO. There are businesses that are pretty easy to run. You can run a quarter rate or a dominant gas pipeline. There are lots of businesses that have regulatory barriers that you can sort of run with.

15:50This is like a triple pike with three twists. This is almost the hardest business to run in the ASX. And that's why we think the Kogan guys are such stars because Kogan is such a hard business. And Susie's a gun. I thought Mark was a gun. Relight the bonfire and continue talking. So revenue was up 10 % in the last, I think it was over the last six months, which frankly was pretty disappointing for a business that had been trading on a super high valuation. It grew 21 % the prior year, 26 % the year before that. The year before that it shrunk, but it was 30 % the year. So this is a business that historically has grown around 20 % to 25%.

16:25So 10 % is not great. Admittedly, this is a pretty tough macro environment. We saw JB Hi-Fi, the darling of Australian retail, probably the best Australian retailer that exists now. They had a tough result. Nick Scali had a tough result. So even the sort of great business, the incredible businesses, generational retails having tough results. So it is a tough environment to be sure. So we can't be too critical of the 10%. Well, I think 10 % growth in this environment because, you know, I call Temple and Webster an undifferentiated third-party retailer, although they do sell a chunk of their own stuff now.

17:00But when I say undifferentiated, what I mean is there's nothing especially magical that sets their product apart from competitors. and so to have an undifferentiated retailer grow 10 % in the current macro environment, furniture retailer, it's not a small achievement to be honest with you. The fact that they're growing full stop is better than a lot of other players in this space. Yeah, I think in isolation of what the share price was, you'd be very happy with 10 % share price revenue growth. I think the problem was this is a business that was valued on like what, 8 to 10 times revenue multiple.

17:31So that was the problem and we obviously can't hold the company responsible for its share price. So if you go further, the next issue was gross margin appeared to drop year on year. So I think the margin dropped from 33 % to 31%, which is not great. That's probably discounting would be my guess. Oh, 100%. It would have to be discounting. And if you go to the EBIT number, like there was some tax stuff happening last year. So I'm ignoring the net number. But look at the EBIT number dropped from 15 million to 12 million year on year, which again is a pretty disappointing result for a business we thought was really starting to lead into profitability.

18:02The company did announce higher EBITDA. but so EBITDA went from$20 to$26 million on an underlying basis. This was based on stuff like excluding a new Melbourne warehouse, some early stage costs in New Zealand, some other stuff. So that's all pretty legitimate but I think you kind of need warehouses to run this business. I think we can sort of ignore the EBITDA number. The EBIT number wasn't great. I think what really freaked out investors and analysts though was the business actually reported that revenue for the first six weeks of this financial year was down 13 % year on year. I think most analysts have predicted flat first half growth.

18:40So this is a real problem. So it's gone from being a 20, 25 % growing business to a 10 % growing business to dropping 13 % in the first six weeks. What appears to have happened is the business has scaled back on marketing. But nonetheless, this negative comp is really worrying. Susie Sugden, who's I think a great CEO who I've known for many years, is a good operator. She's been put in a super hard position, sort of inherited this negative growth. She only just started running it. And I said, this is a super strong management team, but a business that's really got some big structural challenges.

19:10So understandably, the share price has been decimated in recent times. What do you think it was trading in 2025, less than a year ago, I'd say? Dollar-wise, you mean? Yeah, share price dollar-wise. Three times the current price? If you go back, literally last November, the share price was trading at$25 a share. It's now$4.30 a share. So this has been pretty catastrophic. That's 83 % drop. Oh my gosh. Okay. That's terrible. I'm only laughing because of how wrong I got it. Like, yeah. I mean, it's terrible. Well, ironically, when it was trading at that, you said it would drop to this, basically this level.

19:51So we kind of got it right. And this is no sort of great pat on the back for us because the market just got this business really wrong. So it was never a$3 billion business. It was always at an instant. Why do you say it's no pat on the back for us? You know, I don't like that. That's unfair. Like the market, everyone I spoke to when I was talking about this business for a very long time basically told me I was crazy. More polite words than that, to be fair. But like people really, really believed the story that this business told and they valued this business at$2 billion or whatever it was worth.

20:25Now it looks all obvious in retrospect. but like I know you know this but like at the time people really did not see what we saw in this business well it was just so blindingly obvious to us that like no matter how good these managers were that this is a massively commoditized business in an incredibly difficult space competing against hugely scaled competitors that was just impossible like to justify the three billion dollar valuation this business had to make call it 200 million bucks a year and be still growing and this is a business that made 10 million bucks last year so we're not like we're ballparks and it's not growing that fast so it was like even the 500 feels like it's a bit of a stretch to be fair well why do you say a bit of a stretch well quite a stretch like i think we said it would drop to 300 and like it's got to be earning 20 million bucks to hit that 300 so we're still miles away from that so i think it can get to 20 million so i think 300 still is definitely achievable.

21:2020 million of net profit you made, right? Of MPAT, yeah. Yeah. Even a bit less, maybe 15 million of MPAT. We'll get them there. Well, it depends on the growth rate, doesn't it, as well? Which is why that negative growth was so problematic in the last six weeks. Yeah, well, that's one of many reasons. But like one of the challenges with this business is that last year they made $11 million of net profit and this year they made$4 million. But it's quite bizarre because last year they had$15 million of profit before tax and paid three and a half mil of tax. And this year they had 12 mil of before tax and paid seven and a half million.

21:54So I think we need to just look at the profit before tax with this business. That's why I chose it, exactly. Yeah, no, I think you're right about that. And so let's just use that number as a multiple. So they've got$122 million in the bank. I mean, you know, they did, last week we spoke, whenever it was, we spoke about when we were talking about GYG, I said, it's one of the only businesses in recent memory that's undertaken a share buyback and seen a massive increase in their price. Like Temple and Webster is the corollary of that, right? Like they basically spent$30 million trying to catch falling knives with their share price as it's been going down.

22:30That was a bit of a disaster. That buyback was a total disaster in hindsight. They should have just bought stuff. They should have done acquisitions with that money instead of a buyback. There's nothing to buy. I mean, they could have, yeah, you're right, but I'm not sure there's anything to buy that would have actually fit synergistically. Well, I know what I would have bought, but I don't want to tell them my secrets, but I know what I would have bought in this business. And so you've got this 11, 12 mil and a 500 mil market cap, but they've got 100 mil of cash, a bit more, so it's more like a$380 million market cap.

23:02Hold on. They've also got$70 million of trade payables and$22 million deferred revenue. So I think you have to – I've also got 30 million. I think you add cash and inventory and receivables, it gets you to 160, and then take off all those genuine liabilities. Forget the leases, but you've got 112. So it's about 50 million bucks in that cash, I'd say. Okay. So it's 450. All right. And so they got 12 mil, and so they're trading on about 40 times profit before tax. And what do I think about that? I think that is very expensive. Now, admittedly, it hasn't been a good year for them, and we're in a tough economy.

23:42I think it's going to get worse before it gets better. And this isn't their fault, but them being in a tough macro climate should be impacting the share price more than it is. Like, they can't help this. This is just a... And they've been absolutely smashed by Jim and Elbow's budget here because it's just stopped the property market in its tracks. And when you go to a place like Temple, it's when you move houses, essentially. So they're a really innocent victim of the budget disaster. That's this whole industry, frankly. like the whole, yes, you're right. Like you look across this industry, they've all got issues.

24:14I want to say they do capitalize their development costs. I'm not sure if that's new. Is that a new thing? They capitalize formula of development costs. I just noticed - Oh, they've increased that. Yeah. They were always pretty good with not capitalizing much. If I become a big nerd and go all the way down into the notes of this financial statement and look at their intangibles, Development costs in 2025 was$49 ,000. And in 2026, it was$4 million. Yeah, I'm surprised at that. It's very unlike them. Well, the reason I say that is because when you look at their cash flow statement, I'm trying to look at actually how much cash they make, right?

24:48Because we're looking at them on a pre-tax basis, but this has always been a very nice cash generative business. And so they did, you know,$24 million of operating cash versus$46 the year before. but that 24 mil, it doesn't include$4 million of development costs, which I would have taken out for sure of that. And so that comes down to more like 20. Yeah, I'm shocked I've gone from 49 ,000 to 4 million. This is – and none of the analysts actually picked this. So you've picked up something that pretty much every analyst seems to have missed. I know all the analysts listen to this. So this is actually really material for a business that only made, what, 10 million bucks.

25:28that comes let's look at EBIT forget the MPAT disaster but look at EBIT that's like a third of their EBIT yeah so I'm not sure what that is like honestly there could be an explanation for that and I'm not saying that to be nice like it they haven't done it in the past so who knows but at the very least you know it needs it doesn't look right it needs some answers but they still generate 20 million dollars even if you take that out they still generate 20 million dollars of net cash but as we've said before predominantly that's timing like if a business is that's that's negative working capital basically like coming out and so that's what you see the you know again the offset of that is the prepayments or whatever they call it deferred revenue on the um on the balance sheet so i just don't think you can pay for that 20 million dollars of net cash i think you probably have to pay for the pre-text profit as a multiple if you're going to pay for something if you look at the full cash flow, there's 24 million bucks in operating cash, albeit there's some tax in there.