58. Texas bets big on tech. Will it steal the pot?

0:00Welcome to The Difference Engine, the show for tech founders, investors and innovators.

0:11all right so we're telling it straight once again and we'll be grinding our gears against some of the most common misunderstandings about llms and ai we'll also be taking a deep dive into some new data what does it reveal about the future of sass well first go grab your finest stetson we're riding our financial horses west yeehaw

0:42all right so we are heading west west of main street west of london what's happening down in the lone star state certainly having been there recently i can tell you the place is hopping and And you know that when you get into a taxi and the taxi driver can't wait to boast about, you know, their son's new job and how much construction is going on, you just get a buzz from a place in a way that I don't think you have in Europe for quite a while. Yeah, well, there's clearly a lot happening there. And the issue is, is this going to impact on a global basis? We think it probably will. But let's talk you through why.

1:24So in March 2025, which is only a few months ago, NYSE, that's the New York Stock Exchange, relocated its Chicago-based exchange to Dallas. And they rebranded it. And this is really clever. I wonder how much they paid for this. NYSE, Texas. That's an oxymoron, though, isn't it? NYSE stands for New York. Stock Exchange. SE, Texas. Lovely. Yeah, lovely. It is a fully electronic equities platform. And it's sitting there in the big Dallas-Fort Worth conurbation. Its first listing, to our minds rather disappointingly, was the Trump Media and Technology Group, which, of course, has the ticker symbol GDJT.

2:08Donald Trump. Donald Trump. Nice. Anyway, so that's happening. Meanwhile, the tech-heavy New York exchange, the NASDAQ, has actually responded, announcing a regional headquarters in Dallas, reinforcing its presence amongst 200 Texas-based listed companies, which have a very tasty$1.3 trillion in market capitalization. So why did they do this? Well, we think they sense they're about to be caught by a power shift that spells big trouble for the Midwest, shall we say, and the Northeastern exchanges. The financial world was voting with its feet, or should we say it's cowboy boots what has appeared we asked to cause all this consternation well it's the texas stock exchange the txse again see what they did there's no confusion with ny here though at least it's pretty clear txse right it's very texan the interesting thing about this exchange is it's backed by the big boys uh blackrock citadel securities charles schwab and others and has already raised over$160 million and that makes it the best capitalised new market entrant.

3:29Now, they think they're going to start trading early in 2026 with some initial listings later that year. And what is particularly interesting about this exchange is its emphasis on the stuff that the institutions, i.e. the Black Rocks, etc., want and aren't getting from the Midwest and Northeast exchanges. So, it's about cutting costs in bureaucracy, it's about raising quality through stricter listing standards, it's about offering a streamlined, low-latency electronic matching engine, basically an exchange that's aligned with tech, of which there is an increasing amount in that part of the world, as you pointed it out, Paul.

4:15So, we have to ask ourselves, are we witnessing a shootout at the Not Okay Tech Caron? Well, I'll tell you what, the job to have, given there's two brand new exchanges in town, the job to have there is a recruiter for stock exchange people, isn't it? Yes, it probably is. Can you imagine the amount of CVs flying around? I think we should get on a plane and start hiring people. I can just see a whole line of U-Haul vans going from the northeast down to... Well, it's true. I mean, if you look at the names that are relocating. I mean, you've got Toyota moved, I believe, I need to check this, but I believe it's like global HQ to Texas.

4:50Tesla are obviously there with Mr. Musk, Schwab, as you mentioned, Fidelity, and even podcaster, I don't know if we call him a rival, but amusing podcaster, let's say Jason Calacanes, he's moved to an estate in Austin. The place is booming so much so that Universal Studios is building something that was literally as far as my eye could see over the horizon. I can really see that this is a, it's not a flesh in the pan, as you say, U-Haul trucks, livelihoods, talent, the people who own the stock themselves all moving. Yeah, and for those of you who aren't aware of what's been happening is that whole Dallas-Fort Worth area is now the second largest US financial hub.

5:36Is this the growth that we need in Europe, I wonder? Well, it could certainly do with 3.7 trillion in NY listed market cap sitting down there. But also, you know, the reason those U-Haul trucks are heading south is that you've got a very pro-business, low-tax climate driven by the current governor. You know, no personal intacts, favourable regulations, business-friendly courts. You know, it's no wonder Texas is luring the corporations and the innovators. You have to contrast that with our government here in the UK, which is high tax and 13 ,500 millionaires, whatever that means these days, leaving since they came into power.

6:17You have to contrast one is a magnet for capital and talent, the other one, maybe not so much. Yeah, and that is reflected in what's happening on NYC and NASDAQ. You've got rising costs there, you've got greater disclosure demands, and heavy activist activity, ESG board requirements, and so on and so on and so on. So it's no surprise that companies are seeking a more streamlined, lower threshold alternative to that. So we have to say, well, this is really interesting if you're in America, but is it interesting for the rest of the world? So can Texas attract global tech listings? So I think we think that the prospects are promising.

7:05Sure, absolutely. They can start pursuing dual listing strategies. And both TXSE and NYSE Texas are currently positioning itself. And this is really interesting. This is a really interesting entry positioning strategy as complements to NYSE and NASDAQ. Clever, clever, clever. Not replacements, but complements. Okay, so global issuers can maintain existing listings, and that's quite good for not scaring the investment horses. Undoubtedly, if you're tech and you're backed by VC or private capital, you are going to appreciate, inverted commas, reduced cost and bureaucracy, and your backers are going to like it.

7:50And there's definitely an issue around quality here. and TXSE's stricter listing standards, we hope, are going to attract prominent firms and the high-growth tech names looking for that endorsement that comes with those stricter listing standards. I mean, I think that's fascinating given your predominance of VCs in California. If they are incubating companies you know round about them or trying to and then end up listing in texas why not just move over a little bit closer yeah i mean you know dallas may not yet rival uh ny or nasdaq's international brand but frankly those burgeoning texas headquarters uh could could look pretty favorably uh towards uh towards a local exchange or local exchanges and and we know from small acorns.

8:50Mighty oaks grow given the right conditions. But I do think we have to reality check as well on this is that liquidity and visibility are pretty much everything when it comes to exchanges. And those new exchanges are going to need critical mass. They're going to need market makers and awareness to overcome those entrenched incumbents, no matter how shabby they are. And there is the elephant in the room, which is a track record. So those regional exchanges appearing don't have a particularly shiny track record. So things like IEX and LTSC. Euronext, remember that? Limited success in gaining large listings.

9:38So can TXSE go national international? Well, we'll see. But for those seeking liquidity, there's got to be some positive impacts. And on that very point, you know, isn't the reason that our friends at Wise, currently listed, much to my pleasure, on the London Stock Exchange, are looking to go overseas? They haven't said they're going to NASDAQ or NYSE. This has to be an option. Stripes IPO, could they take a risk and go here? It's very interesting times. Yeah, and it is a wake-up call because undoubtedly these guys are going to put increased price pressure into the market. Yeah, right. That level of competition, if the Texans get it right, will force traditional exchanges to cut fees, probably simplify the listing process, and we hope be more businesslike and offer more flexible terms.

10:36There's going to be a loss of friction here with those disgruntled or feeling that they're over-compliant tech founders opting out of NYSE and Nasdaq in favour of Texas, undoubtedly with the encouragement of the current Texas governor and his pretty aggressive administration. And I think this will also make the regional hubs expand because already these exchanges are showing the willingness to go on the fence and actually nurture the local ecosystems. And we hope this would be a global knock-on effect. Yeah, and also let's not forget the London Stock Exchange. You know, went to a dinner with a senior member of that team, made a point that she didn't particularly like about how hard it is to list over here.

11:32And could I get a private word with her? No, she was too busy trying to chat somebody else up to get onto the, join their exchange. I think there's a bit of a wake-up call, as you say, that needs to ripple through all of these rival exchanges. And so in that respect, we are very pro this move. You know, we think that the emergence of the Dallas-based NYSE, Texas, and TXSE, and to a certain extent, other exchanges, does really reflect shifting issuer needs. Right. And it's pretty simple, lower cost, regulatory relief, and aligned ecosystems. The thing is that they're not going to topple Wall Street or major financial centers overnight, but they will force incumbent exchanges to adapt.

12:20So we think the future of listing venues will increasingly be shaped by regional presence, flexible regimes, and tech-first offerings. Hello, LSE. Are you listening? Hello? Hello?

12:41Think there are too many lookalike tech companies chasing the same markets? We do. We think it's because so few tech startups are able to build and lead their own categories. What matters most is to be different. Stop following and start building your own unassailable leadership position today. Working with the category design gurus at Categorical brings decades of differentiation expertise to your team. book a consultation from our website today and we will send you our one pager detailing how to start designing your own highly differentiated category

13:20you know what this really grinds my gears

13:27right then jonathan what is grinding your gears this time oh well you won't be surprised to know it's llms and ai it seems people fundamentally misunderstand two key points when it comes to the potential impact of the two technologies right um yeah that doesn't come as too much of a surprise uh i watched or uh you know esteemed minister for uh technology talk about this on the tv the night and it was embarrassing yeah yeah it was as embarrassing as somebody trying to actually say what their budgets were live on TV. Yes, that was dreadful. You know, what grinds my gears is particularly because these points fundamentally affect how we view AI and its future.

14:07Now, they can also answer the constant question, if it's so great, why hasn't it taken our jump yet? Right? So the first thing is AI's current inability to deal with multitasking is protecting us. And second, the current iteration of AI is a tool, not a solution. Both are likely not always to be the case though, right? So it's been clear that compared with humans, LLMs are more than capable of completing certain challenging tasks. For instance, a couple of years ago, we saw GPT pass difficult exams. Since then, AI models have been getting better and better and better. And anybody who uses them will know that.

14:51But it's important to understand that an LLM's ability to perform a task and its usefulness in doing so is a function not so much of how intellectually challenging the same job would be for us, nor the level of specialist skill required, but of how long it would take a human and how unstructured is the workflow. Now, basically, the messy stuff we all do on a day-to-day basis. And as we know, Paul, Paul, some of us are messier than most, right? I'm going to ignore that, but just a slight side diversion here. We both know people who are currently in academia, let's say. And when asking one of these people, I happen to know rather well, about their favorite use of AI, which was, of course, to answer any questions they were being asked in exam situations, all of the big guys didn't come out.

15:45What did come out, and this is maybe slightly more economically worrying, was, oh yeah, I quite like the deep seek AI. So yeah, it is definitely being used for things that humans traditionally found hard, but LLMs find easy. But it may be that the help might be coming from somewhere a little further east. So yes, indeed it might. And one presumes that your contacts in academia may not be on the staff side of academia from that little description. Let's get a little bit closer. Yeah. So this all means that what we think of as the simplest of current jobs done by humans in terms of task and structure.

16:21Now, a good example with bookkeeping, where you're just basically filling up a ledger over and over again. These are still beyond the capabilities of even cutting edge AIs. The current problem with AIs at the moment is that they struggle to keep track of multiple streams of information, respond to a dynamic environment, work with unclear or challenging goals, and multitask, which, let's face it, most attention span challenge humans these days have problems with mastering these skills. And of course, there's that running joke about male multitasking or complete lack of it. But this is not to say that these jobs will remain a human domain forever.

17:04But currently, we see AI best proving itself at coding, writing, and gathering research. Ironically, these are areas traditionally considered to be more demanding, requiring high levels of intelligence and education. They're sort of graduate level jobs. So sorry coders, writers and researchers, from the point of view of AI implementation, the execution of these jobs can be considered as one dimensional. The workflow is linear and can be broken down and performed in a sequential, predictable and recurring series of white-collar tasks. So if you're in those areas, you'd better get a plan B or plan to integrate AI if you rely on one of these skills for your living.

17:49To take this to the next level, I think you mentioned research. So the sort of writing that one does that I would say you have to put your fingers in your ears to do or lock yourself away to focus, that sort of stuff is done very well by AIs. But the thinking part, the conceiving of it, is not so strong right now. So the plan B could look a little bit like plan A, but with a couple of mods. You know, really, the good news is it's not that AI isn't inherently powerful, as you've just pointed out. It's just that our messiness might have given those support type jobs we talked about earlier a year or two's grace and the rest of us perhaps a little bit more time.

18:26But the thing here is we should never mistake delay for immunity. The only certainty in life is change. It's always just a matter of how and when. Now, the second thing, which has been grinding my gears, as I said, the current iteration of AI is a tool, a tool, not a solution. But again, this is likely not always to be the case. So, you know, don't judge the future power of AI and its eventual impact on your life and work by your own limited experience right now. You know, how crap the output from GPT has been so far, for instance. You know, we've been around the IT industry a long time, and that old IT adage of Geico, garbage in, garbage out, still applies to the output.

19:14And it will always be regarded as crap if we expect the wrong thing out of something. So, you know, we all know that just even basic prompting at the moment is an art still to be learned by most. Yet AI is simultaneously held up to impossible standards, expecting it to be flawless straight away and dismissing it entirely when it's not up to somebody's imaginary perceptions. So many act as if imperfect output means the whole uber category is a fail. You know, nonsense. You know, this mindset blinds us to constant incremental progress, which is going on as we're doing this podcast. And AI doesn't need to be perfect to be transformative.

19:59Early locomotives weren't. You know, cars were far from it. Computers weren't. Well, they were just basic calculating machines. And remember dial-up internet? Well, we do. Perhaps some of our listeners won't. They're not old enough. But, you know, when we dismiss tools like this, we're missing the broader shift. AI will continue to get more powerful every day. That's a certainty. Well, AI is still a tool right now. we must understand the true potential it holds. It's not about wholesale replacement in most jobs, it's about accentuation of the capabilities of individuals, so we'll need less of those individuals to deliver the same output in these tasks.

20:38Now we hope that when this happens, the wealth created by our new machine helpers will in turn deliver new opportunities for augmented human skills, as has been the case in every single industrial revolution so far. Could it be different this time? Only time will tell.

21:01So we are now going to have a little think, and we're going to consider what the latest Q1 2025 numbers reveal about the SaaS software market. market. Now, new figures from Jameen Ball of the Valley's Altimeter Capital on the cumulative revenue growth of all public SaaS companies does not make for comforting reading if you're in that business. The aggregate cloud software market just delivered its worst quarterly performance in years with net new AIR additions plummeting nearly 30 % year over year in Q1 2025. Is it a blip? Is it the end of the cycle or is it just a screwed economy? Just to make the point, these are the guys that are public, right?

21:47And God knows what's happening in the world of private SaaS. So the aggregate of quarterly net new ARR added across the cloud software universe dropped to just$1.65 billion in Q1 2025. That's a big number,$1.65 billion. but it's down from 2.33 billion in Q1 2024. Again, that's a 29 % year-over-year decline. But this isn't just about a bad quarter. The trend has been building for over a year, with growth rates fluctuating wildly between Zeniths and Nadirs. The market experienced a brief resurgence in Q2 2024 with 50 % year-over-year growth, only to crash back down to negative territory by Q1 2025.

22:39Now, this is more yo-yo than year-over-year. And it is what is most striking about the current environment, the extreme volatility in growth patterns. So looking at that year-over-year growth chart we see the stuff of nightmares for software sales executives and investors alike if you go back q2 2024 there was 50 growth now that was pretty much the peak of the whole cess environment the next quarter q3 2024 it declined 29 and in q4 of 2024 it went back up again by 15%. And then Q1 2025, it went down 29%, which is of course a bigger relative decline. This isn't the truly steady, predictable growth SaaS companies have historically enjoyed and investors and M &A processes have mandated.

23:33The SaaS plane appears to be still flying, but the pilot, and indeed the autopilot, seem to have long since bailed out. So what's driving this flight of fright? Ultimately, it's a perfect storm based on several factors. We think the big issues include, first one, enterprise budget scrutiny. We've talked about this before. CFOs in particular are demanding higher ROI thresholds for software investments and continuing to push for vendor consolidation, i.e. we want less vendors in our staff. The days of that endless nice-to-have SaaS tools getting approved are over. Every pound, euro, dollar spent on software now requires copper bond justification, or frankly, a sneaky allocation to the AI budget.

24:26Yeah, and it probably doesn't help that the people who used to buy those, typically not IT people, typically line-of-business people, are slightly concerned about layoffs, right? They're the people who buy it. If they're not around, who uses it? Absolutely. And, you know, we did say this is a perfect storm. You know, we know we're at the end of the cycle and we have seen market saturation. Many of the categories that drove explosive growth in previous years are reaching maturity. That low hanging COVID ripened fruit of digital transformation has been picked and companies are being a lot more selective about additional software investments, particularly in what is a very volatile trading environment for everyone at the moment.

25:13Now, that is part of overall global uncertainty. Despite some positive economic indicators, many businesses are still operating in stasis or indeed survival mode rather than with a growth attitude. This means maintaining or in fact reducing the existing software stack rather than expanding it. I guess we have to, in our perfect storm, add to that AI anxiety. Well, surely that's it, right? The big theory is that people are spending money on AI and not digital transformation. In other words, give the guys whose processes would be digitally transformed their own AI tools, and we don't need to spend all that money on digitally transforming software.

25:57That's sort of true. You can summarize that as companies aren't investing in traditional B2B SaaS solutions. What they're doing is waiting to see how AI will disrupt and reshape their operations, whether they indeed will have an operation, because there's been so completely disrupted by somebody else. So, you know, why buy, say, a traditional CRM or even an AI enhanced one at this stage when AI native alternatives might emerge pretty quickly? all that incremental budget anyway is mostly going to a ai uh you know which is foreshadowing as we've said before in the difference engine the move from sas to ass that's agents as a service any of you in doubt um and and if you talk to people who are out there trying to sell at the moment on a on a day-to-day level just ask anybody carrying a number and they'll tell you Net revenue retention, that's NRR, is down.

26:57And it's much, much harder to upsell. Again, all incremental budgets are being spent on AI. Now, we really shouldn't be surprised. In a maturing market, new sales slow. There's a limit to how much you can shore up expected growth from existing customers. But it's grim to be in the middle of all that. So what do you do? fiddling around at the edges while sass rome burns and the orange nero fans the flames isn't going to cut it at this stage in the cycle return to you know vertical as normal ain't going to happen it's time to adapt or even for the brave pivot but you have to act doing nothing is a death sentence so what do we think invest in ai first of course because that's where the budgets are and deliver some form of AI integration.

27:47You can't compete with AI if you're not already in it. Even if it's just that the customer needs to be seen to be doing something with AI, something AI-ish, you've got to help them. So incorporate AI capabilities into existing platforms to buy some time while you build new AI-based categories for customers to migrate to. Meanwhile, and this may be counterintuitive, realize that consolidation is actually your friend. Customers want to reduce their software vendors, so expand your platform capabilities, again, with AI, or integrate more deeply with fewer tools so that you stay put in the customer environment.

28:28Probably lastly, focus on category-defining mission-critical use cases. Position your product as a strategic, not tactical, essential, not optional. Can it increase revenue or reduce operational risk? That would be a starting point. And of course, exceptional customer experience is a competitive moat. So I'd love those to death that show off your category leadership best. Right? So the numbers tell us that the enterprise SaaS boom is over and it's only the laggards that remain to be converted. Otherwise, the only software spend currently is AI-fueled. But for companies that can adapt to this new reality, there's still significant category creation opportunities ahead.