Why the best companies are scaling faster than ever - Ep 1: The Business Stack of the Future

0:00Propel Fitness Water with Gatorade Electrolites, Zero Sugar, and Vitamins. Propel hydrates better than water to help you get the most out of your workout and get back to your best self. What propels you? Propel with Gatorade Electrolites. Clash it up with Crocs. You know back to school is coming in fast. So why wait to find your new faith footwear? Step into a local Crocs store and step into your new look. Try it. Style it. Make it yours. Because the right pair doesn't just show up. It shows off. First day fits, handled. Walk out ready for whatever's next. Visit your nearest Crocs store today. This episode is brought to you by Palm Olive. Family time isn't just the big moments. It's weeknight dinners. Sitting around the table, everyone talking all at once. So when the plates are empty and the sink is full, use Palm Olive Ultra. Palm Olive's most powerful formula removes up to 99.9% of Greece,

1:02leaving your dishes sparkling clean. And the new convenient pump makes cleaning even easier. So you can spend less time tackling dishes and more time together. Shop now at palmolive.com. Welcome to the business stack of the future. A five-part race series brought to you by Stripe. This is episode one, the Great Acceleration. In this episode, we'll explore why companies are scaling faster than ever before. Here is a ridiculous statistic. At the beginning of 2025, and Thropic, which is the company behind the Claude family of artificial intelligence tools, was reportedly generating annualized revenue at a run rate of around US $1 billion. By the end of 2025, it was around US $9 billion. When we prepared for this episode, I thought that was the whole story. Except the research became outdated while I was working on this episode. By the end of July 2026, Reuters reported that anthropics annual revenue run rate had passed US $65 billion. Now there is an important distinction here.

2:03Annualized revenue is not the same as audited revenue collected over the previous 12 months. We'll talk more about that later in this episode. But even with that qualification, the direction is remarkable. The company was growing so quickly that my research couldn't keep up with it. And I think that tells us something important about business today. The clock is getting faster. Hey there, here's a quick note. This podcast contains general financial information only. That means it's not specific to you, your needs, goals, or financial objectives. So don't act on the information until you've spoken to your financial planner. You'll find our full disclosure, disclaimer, and link to our financial services guide in the show notes. G'day and welcome to the Business Stack of the Future, a five-part race series brought to you by Strut. Across this series, we are going to look underneath the products and brands we see every day. And examine the infrastructure allowing the next generation of companies to operate, monetize, and compete at enterprise scale. And this first episode is not really about payments. It's about speed. It's about why some modern companies

3:05can build products faster, find customers earlier, reach global market sooner, and generate revenue at a scale that would have been difficult to imagine only a few years ago. My central argument is this. The timeline for building a company is collapsing. And as that timeline collapses, the world of business may become increasingly power-lower-driven. That means the winners do not merely grow a little faster than everyone else. In some cases, they grow multiples faster. Their advantages reinforce one another. A better product attracts more customers, more customers generate more usage, revenue, and data. That allows the company to hire better people, buy more computing capacity, improve its product, and earn more trust. That improved product then attracts even more customers and the flywheel turns again. But none of this happens in isolation. Behind the product is an increasingly sophisticated collection of infrastructure. We've got cloud computing, artificial intelligence, data storage, sub-security, payments, billing fraud detection, and tax and treasury management. Today, we are going to examine four business clocks

4:07that appear to me to be accelerating. The time required to build, the time required to distribute, the time required to monetize, and the time required to scale. Then we are going to ask the most important questions for investors. Does moving faster create a durable competitive advantage? Or does it simply help everyone move faster? Something strange is happening to company formation. Let us begin with some data. Stripes 2025 annual letter said businesses operating on Stripe generated US $1.9 trillion in total volume during the year. That was up 34% from 2024, and was equivalent to approximately 1.6% of global gross domestic product. Stripe also said its infrastructure now powers more than 5 million businesses either directly or through its other payment platforms. That matters because it gives Stripe an unusually broad view of what is happening across these new companies, global enterprises and the internet economy. The businesses joining its platform are scaling faster.

5:07Stripe says the group of businesses that joined in 2025 groomed around 50% faster than the equivalent 2024 group. The number of companies reaching US $10 million in annual recurring revenue within three months of launching was double the previous year's number. And among businesses established through Stripe Atlas, 20% charged their first customer within 30 days. Back in 2020, that figure was only 8%. This is a large shift in a short period, but the more interesting data sits beneath the headline. They studied approximately 23,000 companies incorporated through Atlas in 2025. The median company in that group generated 39% more revenue during its first six months than the equivalent company established in 2024. At the lower end, companies around the 10th percentile produce 18% more revenue than the previous year's equivalent. And at the upper end, companies around the 90th percentile generated 52% more. So nearly the entire group improved. But the businesses at the top improved the most, or to put it another way, everyone got faster.

6:09The winners got faster faster. Stripe also found that 56% more Atlas companies reached $100,000 in revenue during their first six months than in the previous year. And the companies that reached that mark did it in an average of 108 days down from 121 days. This is Stripe's own data. So the patent is difficult to ignore. Company formation is accelerating. Time to revenue is compressing, and the gap between an average outcome and a breakout outcome appears to be widening. So the question is, what's really changed? The first collapsing clock building. The first clock is the time required to build a product. Historically, starting a serious software business could require building an enormous amount of infrastructure to yourself. You needed servers, databases, authentication systems, analytics, security, payments, billing, customer support tools, finance systems, and of course, the actual product. Modern companies do not begin with a completely empty sheet of paper.

7:10They assemble a stack of existing infrastructure. They rent computing. They use APIs to connect to specialist services. They use open source software rather than building every component internally. They can use AI tools to help write, review, test, and document software. And they can use established infrastructure to perform tasks that once required entire departments. The modern digital company increasingly looks less like someone building a factory from the ground up and more like someone assembling highly specialized Lego. That does not mean building a great company is easy. The difficult work has a shifted. It is less about creating every technical component yourself and more about choosing the right components, integrating them properly, building something, customers genuinely need, and moving more intelligently than the competition. But when less time is spent rebuilding common infrastructure, more time can be spent on the product itself. And I've got some examples in just a minute. And that can compress the time between an idea

8:12and something a customer is willing to pay for. The second collapsing clock distribution. The second clock is distribution. For much of corporate history, international expansion was highly sequential. A company might build a strong position in Australia. Then it would enter New Zealand or vice versa. Then perhaps the United Kingdom. Then maybe the United States, then Europe or Asia. Each new country could require local officers, banking relationships, tax, marketing teams, distribution agreements, and custom technology. For a new generation of internet businesses, that model is changing. Stripes annual letter puts it nicely. The domestic market for many modern internet companies is increasingly the internet itself. More than half of the businesses joining Stripe in 2025 were outside the United States. Among Stripe businesses that already generated most of their revenue internationally, approximately 30% of that international revenue came from countries that were neither companies' home market,

9:14nor one of the 10 largest economies in the world. That is important. Global growth is not only coming from the obvious large markets. It is spread across a long tail of countries that would previously have been difficult or un-economical for a young company to serve. A good example is ROK, a platform for building websites and applications. ROK reached paying customers in 69 countries during its first month accepting payments, following a viral social media post that generated a $100,000 in revenue in five days. Think about how strange that would have been, say, a generation ago. A young company can put something online, attract attention through a single piece of content, and start receiving money from customers spread across dozens of countries. The product travels almost instantly. The challenge is making sure the infrastructure underneath it can follow. The third collapsing clock monetization. That brings us to the third clock monetization, reaching a customer and successfully charging that customer are not the same thing. A business can have a brilliant product.

10:16It may have strong brand awareness. It may even have customers actively trying to buy it, but demand can still be lost between the moment someone decides to purchase, and the moment the payment succeeds. A customer may not have the right card. The price may not be displayed in a familiar currency. The checkout process may feel untrustworthy. The business may not accept the local payment method the customer normally uses. The payment may even be incorrectly declined, or the customer may simply decide that the process was just too difficult. This is why payments should not be treated purely as an administrative function. The monetization system sits directly between demand and revenue. And the story of Gamma gives us an excellent example. Gamma, the growth hiding behind a payment method. If you haven't heard of it, Gamma is an AI-powered platform used to create presentations, documents, graphics, and websites. The company launched in 2020 and found a stronger product market fit after introducing AI-powered

11:18design tools in 2023. Customer adoption accelerated. But Gamma initially launched the AI product as a free service with users receiving a limited number of credits. You might remember those. People began using all of their credits and asking how they could pay for more. That is a nice problem to have, but it's still a problem. At the time, Gamma had fewer than 10 employees. The team needed to introduce subscriptions quickly, manage different payment plans, track AI usage, and serve customers around the world, without diverting the engineers away from building the actual product. Gamma launched paid subscriptions within weeks. Within two months, it had passed US$1 million in annual recurring revenue and became profitable. By the end of 2025, Gamma had grown to US$100 million in annual recurring revenue and 70 million users. But the most interesting part of the story is what happened over in India. Gamma began accepting UPI, a popular local payment method used by Indian customers.

12:20Its revenue from India increased by 22%. But the product did not suddenly become 22% better. The presentation software did not change overnight. The Indian market did not suddenly discover that presentations existed. The business removed friction. Customers were already interested. The missing piece was allowing them to pay in the way they wanted to pay. There is a valuable investing lesson here. Observed demand is not always the same as potential demand. A company can appear to have weak demand in a country where the problem is distribution, localization, or monetization. That forces us to ask better questions. Is the product failing? Is the checkout failing the market unattractive? Or has the company failed to localize pricing and payment methods? Is revenue growth slowing because customers do not want the product? Or because the infrastructure is getting in their way? The monetization layer does not merely collect the value created by the product. In some cases, it determines how much of that value that business can actually capture.

13:20The fourth collapsing clock, scaling the organization. The fourth clock is the time required to scale the organization. It is one thing to find 10 customers. It is another to serve 10 million. As a company grows, complexity starts arriving from every direction. More customers, more payment methods, more currencies, more pricing plans, more invoices, more taxes. More fraud attempts, more refunds, more enterprise contracts, more reporting requirements, more countries, and more things that can go wrong. The old response to growth was often to hire more people and build more internal systems. But modern infrastructure can allow parts of the organization to scale without headcount increasing at the same rate. You may have seen this yourself.

14:20Class it up with Crocs. You know back to school is coming in fast. So why wait to find your new faith footwear? Step into a local Crocs store and step into your new look. Try it. Style it. Make it yours. Because the right pair doesn't just show up. It shows off. First day fits, handled. Walk out ready for whatever's next. Visit your nearest Crocs store today. That is the real importance of the business stack. A company should not need to rebuild its billing engine every time it introduces new pricing. It should not need to recreate its checkout every time it enters a country. And it should not need to design a new fraud system from scratch. Every time criminals develop a new tactic. Stripe is one of the companies building this financial infrastructure across payments, billing, invoicing tax, fraud, prevention, and money management. But the value is not simply that the payment can be processed. The value is that the company can keep moving

15:22while the complexity underneath it increases. This is what infrastructure does at its best. It makes enormous complexity feel almost invisible. When the flywheel spins violently. Now let's return to Anthropic. Anthropic is an extreme example. I think you'll agree. But extremes can help us see what is changing. Reuters reported that Anthropics annual revenue run rate increased from US$1 billion at the beginning of $25, to approximately US$9 billion at the end of that year. By April 2026, it had passed US$30 billion. By May, it was around US$47 billion. By the end of July, it had passed US$65 billion. Reuters linked much of the growth to enterprise usage and developer products such as Claude Code, where customers can consume large amounts of computing capacity while completing valuable work. This reveals something important about the economics of AI. The number of users does not tell us everything. 10 million people occasionally asking an AI tool what to cook for dinner, maybe less economically valuable

16:25than a small number of engineers using it continuously to create tests to maintain production software. The economic equation is closer to number of customers multiplied by frequency of usage, by intensity of usage, by price. Two companies can have similar user numbers and completely different revenue outcomes because their customers use the product differently. This is why usage-based business models are becoming so important. But before we get carried away with the headline, let us explain what an annual revenue run rate actually means. A simple version of the formula is annual revenue run rate equals current monthly revenue multiplied by 12. For example, imagine a business generates US$5 million as a revenue in July, its simple annual revenue run rate would be US$5 million by 12 equals US$60 million. That does not mean the company collected US$60 million during the previous year. It just means that if July's revenue continued at exactly the same pace for 12 months, the business would generate 60 million. For a rapidly growing usage-based business,

17:27that assumption can be fragile, as we're seeing. Reuters has reported that anthropic calculation annualizes very recent consumption activity, meaning sudden changes in usage can have an outsized effect on the headline figure. So we should not interpret a US$65 billion run rate as $65 billion of trailing, audited revenue. Speed makes measurement harder too, but the broader story still matters. Anthropic appears to have built a reinforcing loop. A strong product attracts developers and enterprise customers. Those customers use the product intensely. Usage generates revenue. Revenue supports greater investment in computing capacity, research, distribution, and talent. That investment can improve the product, and an improved product attracts more customers with more usage. That is the acceleration of flywheel. Better product, more usage, more revenue, more capital and talent, greater enterprise trust. Better product again. The question for investors is not simply how quickly is this company growing?

18:29It is what causes its growth advantage to reinforce itself. WiseTech, boring infrastructure, and extraordinary economics. Anthropic is an extraordinary AI native example, but the underlying economic idea did not begin with generative AI. Australian investors can see a more established version of it in WiseTech Global and its cargo-wise platform. Cargo-wise is software used inside global logistics and freight forwarding operations. From the outside, they may not sound as exciting as a frontier artificial intelligence lab, but some of the best infrastructure businesses do not look exciting from the outside. They sit inside workflows. They become connected to data, processes, reporting, and customer operations, mission critical. Removing them becomes difficult. And as their customers grow, activity through the platform can grow as well. In FY25, WiseTech reported total revenue of US $778 million. Recurring revenue represented 98% of total revenue. WiseTech also reported EBITAR,

19:30excluding the one-off E2 Open Acquisition Costs, of US $409.5 million, equivalent to a margin of 53%. Why can a logistic software company produce economics like that? Because software can separate revenue growth from labor growth. If a freight forwarding customer processes another shipment, WiseTech does not necessarily need to hire another employee to process that transaction. The software handles the workflow. As usage expands, revenue can rise without cost increasing, and precisely the same rate. That is, operating leverage. And because cargo-wise is embedded in complex, mission critical processes, its value is not simply the code itself. The value includes the integration, accumulated workflows, organizational knowledge, and the cost of changing systems. AI did not invent scalable software economics. What AI may be doing is increasing the speed at which companies can reach those economics. Why the AI winners can pull away? This brings us to power laws. A power law environment is one in which a small number of outcomes

20:33account for a disproportionate share of the total result. The best company does not earn 10% more than an average competitor. It may become 10 times, 100 times, even 1,000 times more valuable. We already see versions of this invention capital, entertainment, professional sport, social media, and technology. The important point is that modern infrastructure can increase the size of the market that a leading company is able to serve. A software business is not restricted by the number of customers within its driving distance of the office. The product can be distributed globally. The cost of serving an additional digital customer can be relatively low. And when quality differences matter, customers can concentrate around a smaller number of leading products. Economists have been studying the rise of what they call super-star firms for years. A major paper published in the quarterly journal of economics examined how technological change in globalization can shift sales towards the most productive companies. Increasing concentration as those businesses

21:34serve a larger portion of the market. The research predates the current generative AI boom. But it provides a useful framework for understanding what may be happening now. This does not necessarily create a winner-take-all market. I prefer the phrase winner-take-more. The leading company may not eliminate every competitor, but it can capture a disproportionate share of customers, talent, capital, and profit. And once that lead begins to reinforce itself, the gap can widen quickly. There is, however, a major tension in this thesis, the same infrastructure that helps the winner's scale also makes it easier for challenges to start. Cloud computing is widely available. AI models are available through APIs. Payment infrastructure is available to new entrants, open source software, can be accessed by almost anyone. Distribution through the internet is not reserved for incumbents. The barriers to starting a company may be collapsing at exactly the same time that the rewards for winning are becoming more concentrated. That leads to one of the most important ideas I want you to remember from this episode.

22:36The cost of starting may be falling, while the cost of winning may be rising. Starting may require fewer employees, unless capital than it once did, but winning may demand enormous computing capacity, exceptional talent, global distribution, customer trust, and continuous product development. The tools are becoming more democratic. The outcomes may be becoming less equal. Speed is not the same as durability. Investors also need to avoid assuming that rapid growth automatically creates a great investment. I see this all the time. Speed and durability are different things. A company can grow quickly while having wheat customer retention. It can generate enormous revenue while consuming even more cash. It can attract millions of users without establishing pricing power. It can benefit from temporary excitement, promotional credits, or unusually high demand that eventually normalize. AI businesses can also face substantial variable computing costs. If every new dollar, for example, of revenue requires a large increase in inference costs,

23:39the operating leverage may be weaker than the revenue growth suggests. Customers may switch between models. Competition may force prices lower, a newer model may make today's leader look ordinary. Regulators may intervene, and private company revenue figures may use definitions that are difficult to compare with the audited results we get from listed companies. So the lesson is not find the company growing fastest in buyer. The lesson is understand the mechanism behind the growth. Five questions for investors. When I look at a rapidly scaling company, I would ask five questions. First, what has actually become faster? Is the company building faster, occurring customers faster, monetizing faster, expanding internationally faster? Or is it simply spending more money? What reinforces the company's lead? Does more usage improve the product? Does scale create better data? Does the company earn greater customer trust? Does it attract better talent?

24:39Or can a competitor reproduce the advantage relatively easily? Third, what happens to the economics as usage increases? Do margins improve because software is spreading fixed costs across more revenue? Or do computing support and acquisition costs rise almost as quickly as sales? Fourth, is the infrastructure itself an advantage? Has the company assembled a stack that allows it to test localizing scale faster? Or is it using exactly the same commodity tools as every competitor without creating anything distinctive on top? And fifth, what number am I actually looking at? Is it audited revenue, annual recurring revenue, a recent annualized run rate, payment volume, bookings, or cash collected? The faster the business moves, the more disciplined investors need to be about definitions. So the big idea from episode one is not that every company will be anthropic. They will not. It is that the basic clock speed of business appears to be increasing. Products can be built faster.

25:41Customers can be reached globally sooner. Revenue can begin earlier. Pricing can be tested more quickly. Infrastructure that once required huge teams and years of work can increasingly be accessed through software. And when you combine those forces, the strongest companies may have the potential to pull away at extraordinary speed. Stripe starter suggests that the entire distribution of new companies is moving faster. But it also suggests that the businesses at the top are accelerating the most. That is the power law dynamic. For investors, the challenge is not merely finding growth. It is identifying which growth is self-reinforcing, which company is building a genuine flywheel, which infrastructure choices are helping it move faster, which advantages deepen as the business expands, and which apparent advantages disappear as soon as a competitor catches up. This episode was not really about payments. It was about the infrastructure that allows a modern company to monetize earlier, serve customers globally,

26:42experiment with its business model, and keep scaling without rebuilding its foundations every time the company changes. And that leads us to the next question. What happens when the customer is no longer necessarily a person? What happens when AI systems can search for products, compare prices, negotiate, and eventually complete the purchases on our behalf? In episode two, we are exploring the AI shopping cart. And what happens when machines start buying? I will see you then. Thanks for tuning into this RASC podcast. As a reminder, this episode contained general financial information only. It's not personalized financial advice, like you get from a financial planner, so don't act on the information until you're spoken to one. And if we've mentioned things like financial products, they come with something called a product disclosure statement or PDS, and a target market determination or TMD for short. These documents are essential to read and understand before you acquire or dispose of that financial product. You'll find our full disclaimer, a link to our financial services guide,

27:42and a range of free education by following the links that are available in your show notes. Thanks for tuning in to this RASC podcast. Don't forget to share this episode with a friend or family member that we can help. All right, class settle down. Today's lesson is on the ARCO Rewards app. Try to stay with me. The fundamentals are simple. Earn at least five cents a gallon in rewards, then redeem them later for up to $1 off every gallon. Now here's where it gets complicated. Oh, wait, it doesn't. It's as simple as downloading the ARCO Rewards app to get started. Class dismissed! Say these about $1 dollar for the gallon redeemable with $20 rewards dollars in your law of the account by participating locations, terms, and conditions apply.