Subscribe
Share
Share
Embed
Founder Reality with George Pu. Real talk from a technical founder building AI-powered businesses in the trenches. No highlight reel, no startup theater – just honest insights from someone who codes, ships, and scales.
Every week, George breaks down the messy, unfiltered decisions behind building a bootstrap software company. From saying yes to projects you don't know how to build, to navigating AI hype vs. reality, to the mental models that actually matter for technical founders.
Whether you're a developer thinking about starting a company, a founder scaling your first product, or a technical leader building AI features, this show gives you the frameworks and hard-won lessons you won't find in the startup content circus.
George Pu is a software engineer turned founder building multiple AI-powered businesses. He's bootstrapped companies, shipped products that matter, and learned the hard way what works and what's just noise.
Follow along as he builds in public and shares what's really happening behind the scenes.
New episodes every Monday, Wednesday, and Friday.
George Pu (00:00)
Hey everyone. Happy Friday. Welcome back to the founder Valley podcast. I'm your host George pu and starting from this Friday, we're going to dive down a little bit different format for Friday. So we're going to do the weekly Friday winners and losers. So basically winners and losers from tech finance, both strapping or anything that I curated, I think it's very useful for you, the founders. And then we're going to see the wins. We're going to see the fails and eventually, most importantly, we're going to share why this is very important to you, the readers. So without further ado, let's get into it.
So the first story basically is that chat gpt just introduced shopping experience within the chat gpt interface. Second story is basically FICO, one of the most popular credit store company in the US and also globally has changed their licensing model and that's sending a few stocks dropping by more than 10, 20%. And third story is that European founders, especially the AI founders are leaving Europe and actually going to San Francisco. So my personal take about, you know, those three stories, which I'll get into in a bit.
is that they're actually the same theme, right? They're all basically eliminating middleman. For example, the chat GPT, they're eliminating the Google ads and also the middle persons there. FICO is eliminating the middle person and European founders are eliminating European VCs, which can be seen here as the middleman. So we see the companies that are winning, they control the core IP, they control the core algorithms and they control the rails and they are systematically eliminating middleman to reclaim their margins.
And also reducing friction and also making themselves of course, bigger and bigger and bigger in a tech sense. So if you're building a company in 2025, you need to understand that because only infrastructure and only the end to an experience is going to be even more and more important. let's dive into the first story about chat gpt So the first story actually is not chat gpt. The first story is about FICO. So FICO is the one of the most popular credit score agencies in the U S. So when I was working on Simple Direct,
Obviously because we were working secure financing loans, so we actually have to deal with a lot of with FICO. Even though simple direct ourselves, we don't directly deal with FICO. We don't run people's credits. So all of our lenders do. So think of like companies like Upstar, SoFi, know, SunTrust Bank and all that. So we work with JP Morgan Chase So we work with different banks and every time somebody does apply, they do have to run the credit scores, right? So those lenders, they oftentimes go to
Experian, they go to like TransUnion and they go to Equifax. So those are the three big ones in the U S and basically they charge a lot of money of the banks. Right? So we'll go into, we'll get into exactly how much they do charge, but here is a company that people do seem to forget and they're fighting back and that company is called FICO. So if you know, in the U S we call it FICO score, right? FICO credit score. And that is a score that determines whether you can get a lower rates of mortgage and all that. So that is the company FICO that creates a credit scores.
So they just launched something called the Mortgage Direct License Program. So it's a very complex name, but the market reaction was brutal. So Equifax, one of the credit agencies, their stock dropped 8.4 % just after the news and TransUnion crashed 10.1 % in a single trading session. So here's what FICO actually did as in those stocks dropping. So for decades, they actually licensed their...
scoring algorithms of like, you know, how they determine if someone is credit score is 750 versus as someone's credit score is six 50, there is an algorithm that actually belongs to FICO. So FICO is the one that invented the algorithm. They're the ones who was the IP, the patents, trademarks, everything, right? But for decades and decades, every consumer only knows about Equifax, TransUnion and Experian, because those are the quote unquote duopolies, right? The bureaus will basically compile the data together.
they basically calculate the FICO score and they basically just resell it to lenders. It could be mortgage lenders, it could be personal lenders or credit card lenders. So those companies make a lot of money. And of course, as you can see, they're all public companies directly listed on the US stock exchanges. So they're very popular. So what happened? why is FICO doing this? So that is because, let's get into pricing a little bit. These bureaus are charging close to 100 % markups.
So FICO actually charges those bureaus about $5 per score. And those companies will actually turn around and charge lenders $10. So that's basically $5 per credit score, which is pretty significant. And I know if you're a listener from the US or somewhere else, you might be asking, oh, George, but I do use Credit Karma, I use Borrow Well, I use different types of free tools to check my credit score.
So those companies, they don't work with FICO. They actually work with one of those credit agencies like Equifax and TransUnion and they pay those guys $10 per check, right? Maybe a little bit lower because of volume, but those companies are losing money because they are actually betting on you going back, checking on your score and also, ⁓ clicking other advertisements, getting one of the products from them. That's their business model, right? So as a consumer, you might say, ⁓ but I'm not paying the money for it.
But you know, credit karma, know, borrow well and also JP Morgan Chase or credit companies, your personal loan companies, and most importantly, your mortgage companies. Those are the financial institutions that pay those credit agencies and make their stock, you know, balloon by quite a lot. So FICO has finally had enough, I guess. And they said like, we're kind of done with that. And we're directly going to license with the resellers. We're going to kick out the middleman and we're going to directly work with companies that want that data. And we're going to cut the middleman out.
So the new price is actually $4.4 and 95 cents per score, which is basically a 50 % reduction from what those companies are, middlemen are charging, right? So I think it's really interesting why FICO is choosing to do this now. I think there are several factors, but I think most importantly, those companies have been working with FICO for so many years and FICO probably have seen that, you know, they have...
dual-opalize the market, they are cutting the middle man, they're trying to be monopolized in the market, just like what Visa and MasterCard are doing for the payment markets, right? They're basically squeezing everyone out. And the credit bureaus themselves, know, as unfortunately as their stocks dropping, at the end of the day, FICO can make decisions that rock their businesses. And I'm kind of surprised that they only drop about 8 10 % per day on a daily basis for that day, because the credit bureaus themselves, they actually don't have anything valuable. Sure, they do have a brand that people do know about.
When people say, oh, I need to my credit score, they might have thought of, oh, my TransUnion score, my QuickFact score, et cetera. But at the end of the day, it's still FICO that's presenting the algorithms to those people, presenting the data to those agencies, and they cannot exist without FICO. They're basically just a gatekeeper with a smart markup. It's basically like Coca-Cola and Pepsi. So some people say it's the same, others say it's different. And that's basically the relationship between FICO and the credit agencies.
And that's how I see it. eventually FICO owns the IP. They have the scoring model and the U S mortgage industry, which is $12 trillion, by the way, rely on that specific model. So FICO has the moat. So I don't know. I'm not sure if they're public or private company, but they must have come together and think, we can actually make our market a lot bigger than relying on people to resale it for us. So FICO has had enough, right? But I think this is a theme.
That's actually bigger than what it is. It's basically the infrastructure owner saying like, you know, saying F you and screw you. We're going to take kick out the middle man and we're going to go directly to consumer. So I think in this case it's smart. And I think, you know, there are so many, my personal thoughts on this as well. It's like, you know, when we were building some product financing, if you were know a few episodes ago, I basically announced that we're actually sunsetting what our products cost and product financing and focusing on more other products that allows us to own the end twin experience.
And you might be asking, George, but like, what is that? Why do you have to give up, you know, a product and some sort of projects because you don't own the end to an experience. And I think the answer is actually simple. It's because if we don't own the end to an experience and in our experience, we have a significant amount of customer support tickets from customers who went to our one hour lending partners because, know, background information for a simple like financing. We are the middleman. Actually, we do not make the lending decisions decisions.
because we're a very small bootstrap company. So we essentially send all the decision, the data to the lenders and they make a final decision. And usually that is instant, but sometimes it could take one or two days. So many times we have the merchants, our business customers asking us what's going on. We have homeowners and consumers where the borrower is asking us what's going on. And we have to be this middle person connecting information in silos, very back and forth with nothing being our fault. Because if we have to do ourselves,
Obviously we had the APIs, we have everything figured out and we wouldn't need to actually, you know, like paying everybody all the time and creating a horrible experience because that's what eventually turned out to as the volume increased significantly, we realized we create a silo business. We realized we're the middleman and we're not able to own the engine experience and we're never going to be able to make our customers happy. And we're essentially just creating a business that sends volume to, you know, to the lending partners.
So eventually that's why we pull the plug and we're saying like we're sunsetting by the end of this year, 2025, and we have already started developing new products which will be heading the market that allows us to own the entire experience. So if you ask me, oh, George, like, what do you think about this FICO situation? I just think eventually if you don't end, if you don't own the IP, if you're just the reseller, or if you're just using someone else's API in order to offer a service, it's always going to be risky, right? Think about it this way. If you're like a credit karma, you're in the same shoe.
potentially FICO can control or sorry, they don't work with FICO, they work with Equifax TransUnion. Those agencies can decide whether to increase the price or like to make it harder for you to access it, right? Or introducing specific requirements that you're not ready for. So you're always on the realm of like basically other people's decisions. And while I do think there are some instances where you do need to have API connections, for example, you just strive for payments. I think that's a no brainer.
because it's just like infrastructure at this point. So you obviously need them to be able to take payments, but there are so many apps that do rely on external partners, external APIs in order for them to deal with the core value to their customers. And those startups, I think are the ones that are not owning the Intune experience. And as a result, they will be the ones losing out in this trend of businesses cutting middleman and the investor owners are actually going directly to the consumer. And in this case, you
That is essentially what we're seeing with the second story, is basically chat gpt introducing Shopify integrations, basically allowing Shopify merchants to be able to enable Shopify merchants to directly sell the services directly in chat gpt, which is pretty interesting. So let's see what happened. So OpenAI and Shopify a few days ago just announced the instant checkout inside chat gpt. So I'm sure if you have worked chat gpt,
You had the experience of basically asking chat gpt to maybe compare a few items, give you advice on shopping, et cetera. So for example, if you're like, you know, I really need a cool sunglasses that, can help me protect my skin on my trip to, you know, Europe or South America. chat gpt can now actually work with Shopify and basically instantly recommend products from a selection of Shopify merchants. But the most important thing is that chat gpt and Shopify actually share data together. So Shopify will use their API.
to tell chat gpt exactly how much the pricing is, if there's any inventory, right? What are the images or descriptions and what are people saying about it in terms of the review? And basically as an end user, you can check out right there on the chat. You don't have to leave the conversation, go into the merchants website, fill everything again, creating accounts, create your credit card information, and you don't have to do all of that, right? So chat gpt eventually is going to try to become a company that have your credit card information and directly allows you to buy different things on the internet, which is pretty cool.
I personally do not know if they store it in your credit card or maybe those prompt you to basically enter that in ChachiBitE which is secure in their perspective. So this news is huge as well. Shopify stock jumped more than 6 % on news and it seems like they're connecting more than 1 million of their merchants to ChachiBitE's basically 700 million weekly active users, which is significant.
And I think Shopify is calling it the agent commerce, which basically it's a complex word, but basically AI agents acting as personal shoppers, autonomously handling the entire purchase flow, which I think is pretty interesting for Shopify because I personally have tried e-commerce before I haven't used Shopify. And I personally have the experience of basically starting a broke on Shopify and also on Amazon. And the benefit of actually starting Amazon is that yes, you don't have a lot of control over how your product experiences because.
And you're, you're, you're really facing a lot of competition for most of time because you're selling on a marketplace. But a benefit of Amazon is that it really sends traffic, which is like one of the world is like most visited website in the world. And the traffic is, you know, it's basically very, very huge amount of traffic, even though you're selling a small product on page two, on page six, you'll still get traffic and you'll still be able to sell. And that is a power of Amazon. So for a long time, I thought about Shopify merchants that do have the best technology used to do have the easy technologies that you have the.
inventory system figured out, the payment phase figured out and everything. But the problem with Shopify is that their merchants actually have to deal with marketing themselves, right? And they do not have to give 20%, 30 % to Amazon. But the marketing part, at least for me at a time when I tried it for fun, it was pretty difficult. So I do appreciate how they are trying to figure out this, like, you know, using AI to be selling this to customers. I think it's a win for Shopify. But I do think it ⁓ raises some questions about
LMS as well, right? Because like we've basically seen that LLMs are desperate to monetize and and and we have Recently just saw the news about Elon Musk's company XAI, which is grok, which is pretty intelligent I use grok sometimes on a free tier And I really enjoy it just because it has instant information direction information. It can pull information from X directly It actually can give you a quick perspective over Twitter
content and that's something that the other AI agents and platforms are able to do. But I do realize something that I just announced, I think two weeks ago is that XAI just announced that they're going to be adding advertisements in those models, which personally I think is pretty interesting because obviously we know that Grok's basic tier is free, ChachiPT's base tier is free and Anthropic with Claude's base tier is free and Gemini as well.
they're offering AI for free. So a lot of those AI, of course, they're not free. The training is not free. The training can cost hundreds of millions of dollars to billions of dollars per training completion for a large language model. And also on a database, if you're running LLMs, you know that every message users send costs money on a server basis, right? Like the companies will have to pay Google Cloud or AWS or Microsoft Azure to basically burn the energy to give that response.
So obviously it's not free. And I always ask the question about, how are companies be able to subsidize that? And for longest time, I think the responses that we've gotten is that, you know, like the VC monies are getting there. The Saudis, the Qataris and the UAE, the Middle East are pouring money into this thing. And we have the largest pension funds and we have the largest VC funds, pouring money into this thing. So how can it ever go bad? But for the longest time, I think my main concern is that these AI models, are they underpricing?
just in order to gain more users and are they, are their money losing streak eventually become eventually even possible breaking even. For example, like I do pay Claude a lot of money because I use Claude for coding. So I'm not, I, I'm not sure about Claude's unit economics, but you know, in terms of chat, I use it quite a lot. I'm a daily user as well. It's, it's currently my secondary platform and I use it probably a dozen times a day asking different questions, using a generic data. I'm using it to learn many things as well using their advanced voice mode.
And, I'm only paying them $20 per month, right? So I'm hoping opening eyes, not, not listening to this, but I do think personally the value I'm getting and the willingness I am as a user to be paying for this service is definitely more than $20 per month. However, just because of market competition and also other fear opening eyes, not raising the price just yet. They're trying to lure people into paying for a higher tier, which is $200 per month, which is of course substantial. And most people are not going to switch without easier persuasions.
So the question and also they also offer chat GPT like GPT-5 and events voicemail and other things basically an interface for users for free as well So there are always the free users who are not paying anything and getting a usage cap So that always happens and I think it does bring me the question always about how those companies are being are potentially eventually Profitable do you have to charge people an enormous amount of money to do that to use it? Right, and that's not even including API's as I clean enterprises and other places
of spending money. So that's interesting, I think.
I do want to also bring in another interesting perspective from one of my founder friends who I've just talked to a while ago, short while ago. He's actually building a company without getting too much detail into it. He's actually building a company that works with AI platforms are using those LLMs and they actually work with a group of advertisers who are willing to pay money for the LLMs in order to show those advertisement in the text directly. So they're a little bit frontier on this. They're a little bit like early on this and I didn't get too much deep into what they're working on.
But I do think it's a very interesting story because for companies like that, you do have to be quite early in order to get in the market share. You do have to have a lot of funding. You need to have probably an existing relationship with advertisers in order to pull this off. And you also need to have to find a list of like AI marketplaces who are currently probably losing money and desperate to monetizing, right? So if they're doing that with smaller or medium sized AI companies, I'm sure ChachiBt, Claude, Anthropic and...
you know, Gemini and XAI, they're figuring this out as well. And unlike my friend startup, they have the direct distribution advantage. So they're able to cut out a middleman again and go in directly with the advertisers because they have the platform, they have the leverage, they have the IP, which they all am, and they have everything else. So, you know, widening the middleman, right? So we're entering another era. you know, basically the, the second story, I think it's also as well in terms of Shopify. Well, the second story, I think we can see in two ways. The first is like,
OpenAid is working with Shopify to cut out middleman in terms of Google ads, Facebook ads, Amazon ads, right? That basically holds those Shopify companies hostage. And using LLMs, they actually can distribute those things for free. OpenAid gets something in return, which is the traffic. And Shopify gets something in return as well, which is traffic, right? And I'm not sure how the unit economics works, but that's pretty interesting. And a second is that Grok is already integrating AI. So I think...
Sorry, it's already integrating advertisement in their ads responses. So I'm not sure how advanced they are, but they're definitely trying that. So that's another way of cutting on middleman, right? Which is basically, ⁓ come to us. The platform will work with directly advertisers. We have the consumers. So the losers are basically medicine avenues, ad agencies, and different types of, you know, like advertisement distributors and losers are Google ads, other sorts of like ad distributions, for example, YouTube. So that is pretty interesting. Second story as well.
So moving on to the third story, which some of you might ask, George, how this relevant to the middleman? But I do think personally is enormous. It's like EU founders, so AI founders who are in the European Union, which I guess even includes the UK per this story, they're fleeing to the US, they're leaving Europe, and they are basically proving that capital infrastructure wins at the end. So this story.
is interesting because I personally just got back from Europe. So I actually got a little bit experience of my own, you know, over there and understanding nuances between US capital and also European capital. essentially the Wall Street Journal published an article last week that basically says that the US investors are now dominating European as funding rounds, right? American investors are now accounting for over 71 % of AI deals in Europe by value this year in 2025.
Right. So it's up by, you know, it's up from 57 % last year to 71 % this year. And here's the kicker. European founders are not doing Delaware flips, know, which is interesting word, dollar where flip flops, guess. It's basically saying that those founders, they're basically living in Europe, but they're setting up us holding companies in Delaware just to have access to American capital.
So my personal take is this, it's like, this is not actually obviously nothing new. It has been happening for a long time. And, you know, with me personally in Canada, I also understand companies here in Canada in the past five, seven years of when I was doing this business, they have also incorporated in the U.S. just to have access to capital there. It's not everybody and, you know, you don't have to have a U.S. corporation or to raise capital from the U.S. But I do see a lot of these things still happen, which is, you know, and obviously with Europe, European AI startups obviously cannot
raised high valuation in Europe, they're of course fleeing to the US. That's why we're seeing more capital and tunnel flights from Europe to the US. So one founder in the article basically said it perfectly, in London, we will probably end up half the price of our valuation. In Silicon Valley, investors can actually price the market potential and they can allow companies to grow more aggressively. Another company actually just raised half a million dollars in SF within a week of landing and he had been trying for months in London.
So those are all the cool stories. So this is the same pattern again, I think cutting on the middleman. European VCs are the middleman in this story and capital infrastructure wins, right? The US investors, basis Sequoias, the interesting Horowitz, the white combinators, they have built an infrastructure for themselves, not just for US founders, but for global founders. They're able to write bigger checks. They're able to move faster. And sometimes they can commit to capital within a day or two of meeting.
Right. As opposed to being, for example, being Europe or Canada or the UK, where you probably have to wait for weeks and months for the investor to make a decision because the competition there is not as high as they're in the U S right. So I've been to both sides of the aisle and understand how fast you as investors do move and that you have a looser due diligence processes and they don't ask stupid questions, which for founders are the best, best things. And in this case, as I said, European VCs, they are the middleman.
They're slower, they're more conservative and they write smaller checks. And just like the credit bureaus in the first story and just like the Google adsense in the second story, they are being bypassed because they're the middleman. But here's what people are, I think, missing from the story is that it's not just about startups raising money. It's not just about the money. It's about the speed, right? In AI, especially you need speed to move past your competitors. And European funds take months, American funds take weeks. And that is a different
my personal opinion, and that's not even including the basically the regulations and funders frustrations about capital structures and also how the society is happening right now. So this week, obviously, as you have known, the French government collapsed again within 14 hours of being empowered. Those are all just like uncertainties that adds up to the cloud of basically Europe as a continent, whether it's going to be relevant in the AI stage, I think we're still yet to see. But my personal take is that, okay, like, it's not just Europe.
Right. So I personally from Canada, so there's a huge problem here as well. And I personally run him comp. I rent companies both in Canada and the U S and I can tell you that Canadian VCs don't lead rounds. Right. They only write support checks basically for companies and they tell you, you need to go to the U S to find a lead investor and then we'll lead. And I always see the same thing. I always say, why the heck would I need you? If I can go to the U S and find a lead investor, I'll just find a supporting investors there who will treat me nicer than you guys.
So, you know, even though I have never, I've not raised a VC round, but I have attempted it. So three years ago in 2022, and my lead investor at the time was from San Francisco. And my secondary lead investor was actually from New York. So it was, it was actually a pretty, pretty good experience. And those guys have money. They understand. And it's not just VCs. have a few gentlemen who are really nice enough to almost offer us angel check as well. So, and they're really risk takers, right? They have the capital, they have money and are just in risk. And I think they really understood the space really well as well.
And that's not even their day job. Their day job is running their own companies and they're here writing an Android check. Um, they're basically a quarter of a million dollars each, which is pretty generous. So I do think having gone through both sides of the aisle, you know, in Canada, obviously we've gone through those due diligence being ghosted by VCs and they keep asking, have you have American investors? Do you have a lead? Because I know from my heart and from my knowledge that they cannot lead around and their LPs tell them you don't, cannot write a check unless a lead investor has been found. So.
Those are the stories of my personal experience. So I do think Canadian VCs, just like European VCs, they are becoming the middleman and they're being cut out. Because for a founder who are successful, there is no reason for them to remain in those countries if they do want to raise venture capital and becoming, know, especially in the AI, where speed is everything. And they cannot allow themselves to be slowed down by bureaucratics.
they cannot allow themselves to be slowed down by people who cannot move as fast as them. So I think InfoHood wins, so Y Combinator wins, know, San Hill Road wins at the end. So I don't think European VCs and Canadian VCs or whatnot or UK VCs are bad at their jobs. They're just operating a system that's designed for personally thinking industrial age and not the AI age, right? So the regulations process and risk tolerance is quite frankly, just not there.
So, but the interesting thing, you know, for both staff founders is that you actually don't have to move to SF in order to have that, right? So that's a choice for both staff funders as well. I've talked about this many times on the program. You can actually set up in Delaware and it can be somewhere else and you can go to Delaware. It's right. can go to San Francisco for months and to get your capital, get your introduction there. And then you can move out and keep building your startup elsewhere. I've seen many successful founders do that way and continue on raising series A, series B and series C.
So it is a choice of what you want to do. And if you are a boost ref founder, this is irrelevant. You can build anywhere you want. And with AI can probably build faster. But I do think venture capital is still relevant for many things in the tech space. For example, deep tech, AI tech, and some things that you need velocity, you need capital burning in order to reach that velocity. And I think that's important. So that's the third story. So that's basically the three stories for this week.
Right. I think, you know, for the first one with FICO versus credit bureaus, we're seeing that companies that have IPs, companies that have infrastructure, they, there are eliminating the middleman. They're eliminating the markups and they're taking the whole pie to themselves because the bureaus don't own the score. They got cut out and that's bad for them. And a second story with Shopify and ChachiPT, I think the lesson here is that the commerce infrastructure owner, which is Shopify, eliminated the website friction.
So you no longer have to have a website in order to sell something. LMs are desperate for revenue, so LMs win and Shopify wins. LMs are becoming the new storefront.
And in the third story, we're seeing that, you know, basically the, funding infrastructure in Europe and also in the UK and also elsewhere in the world, they are eliminating the middleman who are not fast enough, who are not efficient enough, right? And who just cannot be good enough for founders. So founders are leaving and they're going towards the best infrastructure in the world, which is the U S and American investors. They are not just sitting on their asses in the U S they fly to all around the world. They go to Europe, they go to Canada, go to UK.
they go to South America, they go where the founders are and they write checks. So I think that's also very cool. They're basically not just saying that, you we're not just American, we have foot, we can go anywhere we want and we'll be throwing checks whenever we see a good fit. So I think all those three stories, in my opinion, the pattern is pretty clear. In every layer of the stack, whether it's data, whether it's commerce, whether it's capital, the infrastructure owners are killing middlemen right now.
And I think what that means to you as a founder is that the middle ground is dead. You cannot just be successful and make money as simply a marked up seller, right? You cannot just sell be selling something as a middleman and be successful anymore, especially not in AI age. You cannot be a slow local fund. You cannot be a direct website just to make money, right? LMs, AI, everything else, infrastructure owner, IP owners, they're going to squeeze you out.
So here are two options I think are realistic for startup founders to be thinking about how to position themselves in this AI age. And I think the option one is very apparent, right? Which is basically, can you be the infrastructure owner? Can you own the IP? Can you own the algorithm? Can you own the core value creation, right? Think about, and this is not something you had to take years or decades in order to achieve, right? Be FICO, don't be credit bureau. But in this case, you can actually own the IP.
try to own the end to an experience and do not leave one party experience entirely to another API, to another provider, et cetera. That is slow suicide and that's going to get your business stuck somewhere eventually. That is the first option. The second option is to be so lean that you don't need a middleman at all, right? Which is basically try to both strap as much as you can, try to build a global teams. For example, teams in India, teams in Middle East, teams in South America, and also teams in North America.
and try to optimize for efficiency. And this way you don't have to rely on gatekeepers because gatekeepers are getting eliminated. You're able to build so free that all the gatekeepers are going to be away. So, but of course, what you cannot be building is be building a business that depends on being the channel between the infrastructure and the owner. I personally actually have to make this decision very recently about whether or not to go into a business like that. And I have made the exact same decision ⁓ that I'm just telling you about.
I do see myself getting blocked again, right? Even though it's going well right now being blocked again by those infrastructure partners. And I'm very afraid that's going to happen because infrastructure partner, eventually they have the incentive to squeeze you out eventually. Maybe now they're early, they're not doing that, but eventually they might. So I personally think try not to think about the question is if I try to think about it as when, right? So this week's action is something I think.
ask yourself these following questions and see how to answer it. The first question is like, do you own the IP and algorithm or are you selling it or reselling it as someone else's? And this happens more than you think. So it's okay if your answer is yes, but if you are just like white labeling something or you're just charting a markup, you're pretty vulnerable because like we're seeing AI is only in the risks and cost of those IP providers to go indirect.
So you're vulnerable, right? And just think about what you can do to mitigate yourself from this. The second question is like, Oh, am I in a destination or can I be eliminated by another better interface? So this is basically coming back to the chat GPT and Shopify story. If you're a website that are selling something, right. And you're going to be replaced by a chat GPT plugin, because let's be honest. This is going to squeeze companies who are not using chat GPT, sorry, chat, sort of, who are not using Shopify out. If you're using Wix, if you're using some other.
platforms to be doing your e-commerce and you're selling there, of course you're going to lose volume because people are now using Chachapiti as their Google's. And your traffic is going to drop because they're going to bind directly from people who are using Shopify, selling on Shopify directly. So you will be squeezed out. So if you're just a website that can be replaced by Chachapiti plugin, you need to own the data layer and you need to own your customer's relationship in a way that's defensible, right? If you're not, you have to think about how can I do that?
The third question is, I structured for speed or am I structured for slow? For example, a Delaware entity, global team, or am I stuck in one geography? So this is something interesting. think it's just something to think about. If we're only accessible to local slow capital, right? For example, if you're not in the U S or even if you're in the U S you're not in the big centers like Silicon Valley and New York, you're at a disadvantage because you're
because you're not structuring or thinking globally, right? Just like those European startups, they have successfully raised startup capital by getting close to American investors. Is that something you can do? Or if you're both strapping, is there some ways you can lower your costs significantly that doesn't really matter if something goes wrong tomorrow because you have a very low cost of running your business. So, and if you're a middleman in any of those scenarios, right, you are the credit bureau in this case. And in my humble opinion is that you need to fix it now.
So as simple direct and ANC, which are two of my companies, like I've personally earned the hardcore, the hardship of basically not being able to own the end-to-end experience. So it is extremely, extremely hard. I think, and I think that's something that you do not understand at the moment, but eventually three, four or five years on the roll, you come back and you realize, shoot, I messed it up. And I didn't know that from coming. And that exactly happened to me with simple direct financing with not ending up.
end of the customer experience, right? It looks so easy at the end, but it doesn't look easy when I just started and seeing that I was, oh, okay, those are the biggest companies in the world. They pass, they figure it out, but the core is that they don't, you know, so in 2025, try to own the infrastructure, try to optimize the geography and try to eliminate the friction. The era of cutting out the middleman is happening and make sure you own the infra, you own the IP, you own the customer, you own the algorithm and don't be the middleman. So.
that's it for this week. hope it's valuable. If this valuable, obviously share with the founder who needs to hear it. Share it on Twitter, share on everywhere you see this. you have a founder network and of course you can always find me on Twitter as the George pu. And as you kind of see the show notes, transcripts and other important things that don't share, you know, on the podcast is because it's a lot longer frameworks or thought processes and other updates. You can find that on founder reality.com and until next time, I'll see you then.