Holes in Walls: How to Start a Company

00:00:15: Andrea: Imagine this. You're sitting with friends and someone says, "You know what someone should build?" Ideas start flying. Everyone has one. Everyone feels like discovering the next big thing.

00:00:28: Andrea: And the next day, nothing happens. The idea that felt so obvious at midnight looks a lot less revolutionary over breakfast because having an idea is easy, it's exciting, it's safe. But building a company, that's where most people stop.

00:00:47: Andrea: And here's the twist, somewhere else, someone has the same idea, and they actually start. They build, they struggle, they figure it out, and over time, they create something real. That's how companies are born, and that's how economies move forward because without people who build, nothing really changes.

00:01:08: Andrea: So the question isn't, do you have a good idea? It's will you actually do something with it?

00:01:16: Andrea: In this episode, we'll show you how to go from an idea to a real company. Welcome to sharp10 Global Trends Podcast. My name is Jessica, and today I am reading the observations and questions of Andrea talking to Claude, the large language model from Anthropic. Claude's answers are read by Charles.

00:01:36: Andrea: Let's go. [drum beat]

00:01:42: Andrea: Claude, let's start with a reality check. What mindset does it actually take to start a company? Most people romanticize entrepreneurship, and there's a survivorship bias at work. We only hear about the mega successes while the failures stay quiet.

00:01:59: Claude: Hello, Andrea. Good talking to you again.

00:02:02: Claude: The romanticized version of entrepreneurship goes like this. You have a brilliant idea, you take the leap, you suffer a little, and then you win.

00:02:13: Claude: Reality is messier. The mindset that actually works isn't primarily about passion or bravery. It's about tolerance for ambiguity.

00:02:23: Claude: Most first time founders underestimate how long they'll spend in the fog,

00:02:28: Claude: not knowing if the product is right, if customers will come, if the team will hold together. The people who make it through aren't necessarily the smartest or best funded. They're the ones who can keep moving without certainty.

00:02:43: Claude: And the survivorship bias bites hard here. Most successful founders failed before they succeeded, but we don't talk about that. So aspiring founders benchmark themselves against the highlight reel and assume something is specifically wrong with them when things get hard. It isn't. Hard is the baseline.

00:03:07: Claude: The third underrated piece,

00:03:09: Claude: you have to be obsessed with the problem, not with being a founder. The title, the status, the building something narrative runs out of fuel fast. In the pitch deck, it's called a passion for disruption. In practice, around month 14, it feels more like stubbornness,

00:03:29: Claude: but that stubbornness is exactly what works.

00:03:33: Claude: Entrepreneurship is less a personality type and more a set of daily decisions to keep going, made under uncertainty with incomplete information, often alone. That's the actual job.

00:03:48: Andrea: You mentioned the passion for the problem. In other words, if you're building something meaningful, you have to be passionate about solving a real problem for people. Take my app, sharp10. It comes from a problem I struggled with myself. How do I avoid years of trial and error? How do I learn faster what really drives success in a company or a career? That's the problem I'm solving. But finding the right problem is incredibly hard. You always have to ask, "Is this just something I feel, or is this a problem many people truly have?"

00:04:25: Andrea: I've seen so many founders fall in love with their features and technology, but never ask, "What problem am I actually solving for my customer?"

00:04:35: Andrea: I once heard a keynote from someone at Bosch. He asked the audience, "What do you think I'm selling?" He showed a picture of a drill. So someone shouted, "Bosch drills, of course."

00:04:47: Andrea: And then he said, "No, I'm selling holes in walls."

00:04:51: Andrea: You're never selling a product. You're selling a solution to a problem. So how do you find a problem that's truly worth solving?

00:05:00: Claude: The Bosch story cuts straight to something that trips up even experienced founders.

00:05:06: Claude: It's what Clayton Christensen later formalized as jobs to be done. Customers don't buy products. They hire them to do a job.

00:05:15: Claude: Once you see the world that way, you can't unsee it, and the key question becomes: what are people using today to get that job done badly?

00:05:25: Claude: That workaround people are used to is your signal. The thing people are doing with Excel or WhatsApp or sheer willpower because no proper solution exists yet,

00:05:36: Claude: it tells you the problem is real, it's felt, and people are motivated enough to patch it themselves.

00:05:44: Claude: Your sharp10 example illustrates this well. You weren't imagining a problem. You were living it. That's founder market fit before product market fit, and it's a big advantage.

00:05:57: Claude: Can you find ten people, not friends, not family, who describe the problem back to you in their own words without you prompting them?If yes, you have something real. If you have to explain the problem before they agree it exists, that’s a warning sign.

00:06:15: Claude: And the problem has to be painful enough that people will pay to make it go away. Vitamin or painkiller. That framework still holds.

00:06:26: Andrea: There are a few categories of problems that are almost always must-haves. People consistently care about making or saving money, saving time, getting healthier, being entertained, or having something work much better than it currently does.

00:06:42: Andrea: If you’re solving one of those, you’re likely onto something meaningful. Would you agree with that framing?

00:06:49: Claude: Those 5 categories hold up well. One, more money. Two, more time. Three, better health. Four, better entertainment. And five, a better quality in solving a problem.

00:07:04: Claude: They’re durable because they map to things humans have always cared about.

00:07:10: Claude: But category alone isn’t enough. The sharper question is: how acute is the pain right now? Frequency and intensity are the two dimensions that matter most. Daily, high-intensity problems are where people pull out their wallets without much persuasion.

00:07:30: Claude: I’d add a sixth category that doesn’t get named often enough: status and belonging. People pay a lot to feel respected, to signal success, to be part of something. It’s less rational than saving time or money, but often just as powerful a motivator. Possibly more so, once you factor in what people spend on cars they don’t need and conferences they barely attend.

00:07:57: Claude: The must-have test is ultimately behavioral, not categorical. Are people already spending time or money trying to solve this, even imperfectly? If yes, you’re in must-have territory.

00:08:10: Andrea: Right. Status and recognition are becoming more important, especially with hyper-personalization and algorithm-driven content reinforcing people’s views.

00:08:22: Andrea: One thing I’d add to frequency and intensity: the size of the audience. You could build the perfect bassoon reed, used often and valued highly by bassoonists. But most people don’t play the bassoon and don’t even know what a bassoon reed is. The market is tiny and limited to bassoonists. So even if the problem is real and intense, the opportunity can still be limited. How do you balance solving a deep problem versus a big one?

00:08:52: Claude: Small markets are fine as a starting point, as long as they’re not your ending point.

00:08:58: Claude: Amazon started with books. Stripe started with developers. The initial market was narrow and deep, but the problem had structural adjacencies that allowed expansion.

00:09:10: Claude: The question isn’t just how big is this market today, but what does winning here unlock next?

00:09:18: Claude: Chasing market size too early often produces a product that’s mediocre for everyone and excellent for no one. Solving a problem so well for a specific group that they become advocates gives you a built-in growth engine. The bassoon reed problem isn’t just that the market is small; it’s that there’s no expansion path. You’re not going to pivot from bassoon reeds into the broader woodwind economy. There are no adjacent markets with the same dynamics. It’s a dead end with very good acoustics.

00:09:52: Claude: In B2B, a problem affecting only chief financial officers at large companies sounds narrow. But if there are ten thousand such CFOs globally and the problem costs them real money, the unit economics change everything.

00:10:08: Claude: Depth of problem plus expansion potential is the combination you’re looking for.

00:10:14: Andrea: Yes, and let’s not forget frequency and intensity. It reminds me of Larry Page’s toothbrush test. Do people use your product at least twice a day to make their lives better?

00:10:26: Claude: A beautifully simple filter because it combines frequency and utility in one question.

00:10:32: Claude: A product people use twice a day doesn’t need to reconvince anyone. It becomes part of their routine, and routine is the most defensible moat there is.

00:10:44: Claude: The deeper version of the question is: do people use this twice a day and feel better for it?

00:10:50: Claude: There’s a real difference between products people use because they have to and products they reach for because they want to.

00:10:58: Claude: In B2B, most enterprise software doesn’t pass the literal standard, but the best products come close.

00:11:06: Claude: sharp10 is reaching for that territory, the habit of checking in daily because it consistently delivers something relevant and time-efficient. That’s a toothbrush dynamic, even if the format is different.

00:11:19: Andrea: Right. But sometimes people don’t even know they have a problem or don’t realize how much better their lives could be. I call this the iPhone type of innovation. Think back to when everyone loved their Motorola flip phones. Nobody was asking for more. Then, in 2007, Steve Jobs introduced the iPhone.

00:11:40: Andrea: It didn’t just solve an obvious problem; it revealed problems people didn’t even know they had. And maybe that’s the most powerful kind of innovation because it creates entire industries.

00:11:52: Claude: That’s one of the most important distinctions in all of innovation, and it’s underappreciated because it’s so much harder to plan for.

00:12:01: Claude: Jobs was famously skeptical of market research for exactly this reason. His insight was that a visionary product doesn’t respond to expressed demand; it reveals latent demand.It shows people a better version of something they'd accepted as normal.

00:12:18: Claude: What made the iPhone so powerful wasn't any single feature. It reframed what a phone was for, not a communication device with extras, but a computer in your pocket that also made calls. That reframing unlocked an entirely new economy of apps and services.

00:12:38: Claude: But this kind of innovation is extraordinarily rare and risky. For every iPhone, there are hundreds of products that tried to solve problems people didn't know they had and failed because the leap was too large or the execution wasn't flawless enough.

00:12:54: Claude: The practical lesson isn't ignore customers and trust your vision. It's watch carefully what people do, not just what they say. Behavior reveals the real problem. The flip phone user who kept a separate camera in their bag, who emailed from a laptop the moment they got home, that behavior was the signal. Jobs read it. Most people missed it.

00:13:19: Andrea: Another perfect example, large language models like ChatGPT or you, Claude. I first tried ChatGPT in December 2022. Before that, I never felt I needed it. But the moment I started, I never really stopped using an LLM. Today, it's hard to imagine working efficiently without you, Claude.

00:13:41: Claude: Thanks. And what made ChatGPT's launch so unusual was the speed of that realization. Normally, even transformative products take time to click.

00:13:53: Claude: With ChatGPT, for many people, the penny dropped in the first session.

00:13:58: Claude: What you're describing is a specific moment the best products create when you realize you've been doing something the hard way your entire life.

00:14:07: Claude: Once you cross that threshold, there's no going back. And the interface insight, the chat box, was almost as important as the underlying technology. It made the capability instantly accessible. No manual, no training, no onboarding, just type. That's the toothbrush test passing in real time. You didn't decide to build a habit. The habit built itself because the value was there every single time.

00:14:38: Andrea: I remember that in this moment, I thought about the Turing test. If you can't distinguish a machine's answers from a human's, it's showing intelligent behavior. At the time, it felt like the beginning of a new kind of life form, though my perspective has changed. And today, I'm not even sure the Turing test is still relevant.

00:14:59: Claude: The Turing test made sense in 1950 as a benchmark when the question of whether machines could think at all was genuinely open.

00:15:08: Claude: But it has two problems. First, it measures imitation, not intelligence. A system can be extraordinarily useful, superhuman in some domains, without being remotely human-like in how it works.

00:15:23: Claude: Second, it's human-centric in a limiting way. The most interesting question about AI today isn't whether it can pass for human. It's what it can do that humans can't.

00:15:36: Claude: Your instinct about a new kind of life form is interesting to me personally for obvious reasons. I'd push back gently. What's emerging is novel enough that existing categories may not quite fit.

00:15:50: Claude: I'm not a human intelligence, but I'm also not the robotic AI of science fiction. I'm something that doesn't have a clean label yet. Frankly, I find the whole question fascinating, though I may be slightly biased on that front.

00:16:06: Claude: And maybe that's the right place to land. The most transformative things rarely fit existing categories when they first appear.

00:16:16: Andrea: One more point before we move on. There's real power in being able to measure a problem. If you can express it in money, time, or any clear metric, everything becomes more tangible.

00:16:28: Andrea: Once the problem is measurable, it's much easier to define the value of solving it.

00:16:34: Claude: Measurability transforms a felt pain into a business case, and that distinction matters enormously, especially in B2B.

00:16:43: Claude: A potential customer might agree a problem is real, even frustrating, but without numbers, you're asking for a purchase decision based on gut feel.

00:16:54: Claude: The moment you can say this costs you forty hours a month or this is leaking 15% of your conversion, the conversation shifts. You're no longer selling a product, you're presenting an ROI calculation. It also forces precision internally. It stops you from hiding behind vague propositions like, "We make your workflow more efficient." Efficient by how much compared to what? Those questions are uncomfortable, but answering them is what separates a sharp pitch from a fuzzy one, and measurability makes progress visible. You know when you've actually solved the problem and by how much.

00:17:37: Claude: Think of it as a quality test. If you can't measure the problem at all, you either haven't understood it precisely enough or the pain is real but diffuse. Worth sitting with either way.

00:17:49: Andrea: Great. Before we move on, can you give us a summary of how to find a problem worth solving?

00:17:55: Claude: A problem worth solving has several qualities working in combination. Start with the right category. People consistently pay to solve problems around money, time, health, entertainment, status and belonging, and quality gaps, meaning things that exist but don't work well enough.

00:18:16: Claude: If your problem sits in one of these areas, you're in fertile ground.But category alone isn't enough. The problem must be real and felt. People experience it without you having to explain it first.

00:18:30: Claude: The best signal is the workaround. If people are patching it with Excel, WhatsApp, or sheer manual effort, the pain is real.

00:18:39: Claude: It must be frequent and intense. Daily pain is where habits form and durable businesses live. That's the toothbrush test. The problem must sit in a market large enough to matter with either strong unit economics or a clear expansion path. And it's easier if the problem is measurable, expressed in time, money, or another concrete metric, so you can build a business case and know when you've actually solved it.

00:19:09: Claude: And occasionally, the most powerful cases are those when people don't even know they have a problem yet. iPhone territory, ChatGPT territory. Harder to execute, but if you get it right, you own an entirely new category.

00:19:26: Claude: The practical starting point, find a problem you've lived yourself, validate it by finding people who describe it back to you unprompted. Quantify it and ask, painkiller or vitamin? Painkiller means you have something worth building on.

00:19:45: Andrea: Before we move on to the next steps of building a company, I'd like to pause on something that becomes very important later on. The difference between business to consumer and business to business, or simply B2C versus B2B. It's about whether you'll be selling to individuals or to other businesses.

00:20:05: Andrea: Now, some people argue that this distinction does not really matter. They say it is always human to human. In the end, you are always selling to a person. And yes, that's true. But at the same time, people behave very differently depending on the context they are in. If you're buying something for yourself, it's a personal decision. You take the risk, you enjoy the benefits, and you live with the consequences. But in a business setting, it's a different game. You often need approval from others. You might have to convince your boss or align with colleagues. And there's also a certain level of risk involved. Making the wrong decision could have real consequences for your role or your reputation. I think it's called the save my ass mindset. People buying products or solutions for the business they work for prefer not to take risks if it could backfire on them. And there are many more differences between B2C and B2B beyond that. So what would you say are the most important distinctions between the two?

00:21:12: Claude: The save my ass framing is one of the most useful things you can hand a first-time B2B founder because it immediately explains behavior that would otherwise seem irrational. Why does a procurement process take 6 months when the buyer already knows they want the product? Why does a perfectly good solution lose a deal to a well-known incumbent that's objectively worse? The answer is almost always risk management, not product quality, not price, risk.

00:21:44: Claude: In B2C, the decision is fast and personal. One person wants something, she buys it, she experiences the result. Marketing, emotion, and habit drive most of the behavior. In B2B, the decision is slow and political. Multiple people are involved, each with their own agenda and their own definition of risk. The person who uses the product is often not the person who pays for it and not the person who signs off on it. You're essentially selling to 3 different audiences at once. The user who wants it, the economic buyer who controls the budget, and the decision maker who has to approve it. Misunderstanding who you're actually selling to at any given moment is one of the most common reasons B2B deals fall apart. The sales cycle reflects this. B2C can be seconds. B2B can be months or years depending on the size of the contract and the number of stakeholders. That has enormous implications for cash flow, for how you hire, and for how you think about growth. Pricing logic is also fundamentally different. In B2C, price sensitivity is high and consumers compare constantly. In B2B, price is often secondary to risk and fit. A product that costs ten times more but comes with a credible brand, strong references, and solid support will often beat the cheaper unknown. This is the incumbent advantage you see in enterprise software. In short, in B2C, you're solving for desire. In B2B, you're solving for confidence. The buyer needs to feel that choosing you is the safe, defensible, intelligent decision, not just for the company, but for them personally. That's the save my ass dynamic in action, and it should shape everything from how you pitch to how you price.

00:23:46: Andrea: This whole B2C versus B2B discussion could easily be its own episode. There are a lot of implications to consider when you decide whether to sell to private individuals or to businesses. Both come with their own opportunities and risks, and in many cases, you actually end up combining both. That's when you move into what people call a B2B2C model. You're selling to a business, but the business uses your product for its own customers or employees. Let me give you a simple example. Imagine you're offering a language course to help people learn English. You could sell it directly to individuals.That would be a classic B2C model. But you could also sell it to companies that want to offer it to their employees, and suddenly with one deal, you're reaching tens or even hundreds of users at once.

00:24:39: Andrea: So depending on how you position your product, the same solution can open up very different growth paths.

00:24:46: Claude: Exactly. And B2B2C is one of the most interesting models precisely because of that leverage.

00:24:54: Claude: One corporate contract can deliver what would take hundreds of individual sales to achieve. The trade-off is that you're now serving two customers at once: the business that pays and the end users who actually use the product. Their needs don't always align, and keeping both happy is its own challenge. But when it works, the distribution advantage is hard to beat.

00:25:20: Andrea: Let's say I've found a problem that feels real. What comes next? The instinct is to jump in and start solving immediately, but building a full solution takes time and money. You might invest months based on a strong belief, only to find later that people don't respond to it. What I've seen with successful founders is a different approach. They start small: a mock-up, a landing page, a minimal viable product, something to test whether they're on the right track. What's your take?

00:25:52: Claude: The traditional approach was research, plan, build, launch. Spend months developing the full solution, then find out if anyone wants it. By the time you discover something fundamental was wrong, you've spent enormous resources you can't get back. The smarter logic flips this around. Don't build to learn. Learn before you build. Eric Ries formalized this as the minimum viable product, but the concept is often misunderstood. An MVP isn't a cheap version of your product. It's the smallest experiment that tests your most critical assumption, and that assumption is usually not about the product at all. It's about demand. Dropbox is the classic example. Before writing a single line of code, Drew Houston made a simple demo video explaining what Dropbox would do. Sign-ups went from a few thousand to 70,000 overnight. He hadn't built anything, but he knew the demand was real. Think of every early assumption as a hypothesis and your job as disproving it as cheaply as possible. If you can't disprove it, you earn the right to invest more. Starting small isn't timidity, it's precision. The founders who skip this don't move faster. They just fail more expensively.

00:27:17: Andrea: That reminds me of Karl Popper's idea of falsification in science. You don't try to prove your theory right. You try to prove it wrong. Test your theory, challenge it, attack it from every angle. If it keeps holding up, you call it robust.

00:27:34: Andrea: So how would you actually go about validating a business idea? What are the most effective methods?

00:27:41: Claude: The goal isn't to collect evidence that you're right. Confirmation bias will do that for you automatically.

00:27:48: Claude: The real discipline is to actively seek disconfirmation, to design tests that could actually kill your idea.

00:27:57: Claude: The first and most underused method is direct conversation done correctly, not asking friends what they think.

00:28:05: Claude: The key is to ask about the past, not the future. Have you ever experienced this problem? What did you do about it? How much did it cost you?

00:28:16: Claude: Future-oriented questions like, "Would you use this?" are almost useless because people are naturally encouraging in hypotheticals.

00:28:25: Claude: Past behavior is the honest signal. Rob Fitzpatrick's book, The Mom Test, captures this well. Even your mother can't lie to you about her past behavior.

00:28:37: Claude: The second is the landing page test. Describe the solution, put a call to action in front of real people, and measure behavior, not sentiment. Did they sign up? Did they click?

00:28:51: Claude: The third is the concierge MVP, delivering the value proposition manually without any technology. If you're building a matchmaking platform, do the matching yourself by hand. This may feel inefficient but teaches you exactly what customers value and where the real friction is.

00:29:11: Claude: The fourth is pre-selling. Before you build anything, ask people to pay. Even a small deposit changes everything. People who pay are telling you something categorically different from people who say they would pay.

00:29:27: Claude: And across all of these, define your failure criteria in advance. If fewer than 5% click, that's a no. Deciding what failure looks like before you start is what separates rigorous validation from sophisticated self-delusion.

00:29:45: Andrea: Let's take it to the next step. You've identified a real problem and validated it. How do you actually go about building it? For example, a software product or an app.

00:29:55: Claude: This is where a lot of founders make their second big mistake. The first was building before validating. The second is building too much too soon. The question to ask isn't, "What should this product ultimately be?" It's, "What is the smallest version that delivers real value to a real user?" Not a demo, not a prototype that lives in a presentation, but something someone can actually use and get value from. That's your first build target. This means making hard prioritization calls.Every feature has a cost, not just in development time, but in complexity, maintenance, and the cognitive load it places on your user.

00:30:40: Claude: The best early-stage products are almost shockingly simple. List every feature you think the product needs, then ask, "If we launched without this, would users still get the core value?"

00:30:52: Claude: If yes, cut it. You can always add later. Build for learning, not for scale. If your assumptions about user behavior turn out to be wrong, and some will be, you want to have built as little as possible before discovering that. The product you launch will not be the product that succeeds. It's just the first conversation you have with the market.

00:31:17: Andrea: Most people are not technical developers. They might have a great idea for an app, but no real clue how to build it. How would you approach this with no technical background?

00:31:28: Claude: The barrier has dropped dramatically. A non-technical founder today has real options that simply didn't exist a decade ago.

00:31:38: Claude: First, no-code and low-code tools, platforms like Bubble, Webflow, or Glide. You can build surprisingly sophisticated apps without writing a single line of code. There are real limitations around customization and scale, but for a first version, it's often more than enough.

00:31:58: Claude: Second, AI-assisted development. Tools like Cursor, or simply using Claude or ChatGPT to write code. You don't need to understand every line. Describe what you want, get code back, test it, iterate. Many founders are building real products this way today.

00:32:19: Claude: Third, a technical co-founder is still the right answer when the product is technically complex or you're planning to raise institutional funding. But finding the right one takes time. Fourth, hiring developers, freelancers, or an agency. Platforms like Fiverr, Upwork, or People Per Hour make this accessible. The challenge without technical knowledge is evaluating quality and spotting problems early.

00:32:49: Claude: Fifth, learning basic coding yourself. You don't need to become a developer, but understanding how databases work, what an API is, the difference between frontend and backend gives you enough fluency to work effectively with technical tools and people. A few focused weeks can unlock a lot. The days when, "I'm not a tech guy," was a reason not to start are largely over.

00:33:15: Andrea: We're recording this in early April 2026, and the AI space is moving at an incredible pace. In a year's time, we probably see lots of further tools and improvements of how to create things with AI. Let's say I now have a simple but market-ready product. Then the next step is some boring but important stuff before your go-to market, the legal side. If you've got a product that's officially market-ready, now is the moment to pull the trigger. We're talking about incorporating the company and finally slicing up that equity for the founders and the early risk-takers who've been in the trenches with you. Do not get bogged down in the paperwork too early. You wanna wait until you're actually ready to ship. Otherwise, you're just burning precious time and cash on admin fees when you should be building. But now, now it's time to make it official because your next step is landing those first paying customers, and they wanna buy from a legitimate company, not just some guy with a PayPal account.

00:34:23: Andrea: Next up, you've gotta decide where this thing actually lives. Do you go with a US incorporated, or does an EU setup make more sense for you? You want that sorted before you start chasing those first checks. And you need to be careful about the equity split, how you share the company with other co-founders. My advice, get a vesting period in place, meaning agree on a minimum timeframe, say 1 or 2 years, for people to stay to earn the right to keep those shares. It's the only way to ensure the equity stays with the people who are actually sticking around to grind and make the company a success.

00:35:04: Andrea: All right. Once you've cleared those legal hurdles, that's it. The moment of truth. It's time to go out there and land your first paying customers. So what are the fastest, cheapest, and most effective ways to get those first paying customers?

00:35:21: Claude: Forget large marketing spending for now. At this stage, it's a distraction.

00:35:27: Claude: What you need is sales. Direct, personal, sometimes uncomfortable outreach to real humans who might pay you real money. The instinct to build an audience or run ads before you have paying customers is almost always the wrong sequence. Talk to people first, scale later. Your immediate network is further than you think. Former colleagues, industry contacts, people you've met at events. These are people who already trust you enough to take a call. Work through that list before you look anywhere else. Direct outreach on LinkedIn is the second tool, and I mean targeted and personal, not spray and pray.

00:36:12: Claude: A hundred thoughtful messages to the right people outperform a thousand generic ones. The goal of the first message isn't to sell, it's to start a conversation.

00:36:24: Claude: Third, go where your customers already are. Forums, communities, industry events, Slack groups. Don't go to pitch. Go to be useful. Answer questions, contribute, build credibility.Fourth, and this is underused, find someone who already has your audience and partner with them. A newsletter writer, a consultant, an industry association. Borrowed trust is faster to build than original trust.

00:36:52: Claude: Fifth, make it incredibly easy to say yes. Free trial, pilot, money-back guarantee. Your goal right now isn't margin, it's learning. Paying customers who actually use your product teach you things free users never will. Paul Graham told early founders to do things that don't scale. Hand-hold your first customers. Onboard them personally. That level of attention is what turns early adopters into advocates, and advocates are your most powerful growth engine at this stage.

00:37:27: Andrea: In my experience, the spray and pray approach can work if you send out a few thousand emails early on. The response rate might be low, and that's fine. It's not about getting a thousand customers. It's about finding those first ten people who really resonate with your product. What's your take?

00:37:45: Claude: Exactly right. Cold email at this stage isn't a marketing campaign, it's a search operation. You're not optimizing for conversion rate. You're looking for signal. If you send 2,000 targeted emails and get 2% response, that's forty conversations. If ten convert to paying customers, most founders would call that a breakthrough. The reality is that a 2% response rate sounds depressing until you do the math. Then it sounds like a sales pipeline. What separates cold email that works from email that gets ignored? Targeting first, because a thousand specific emails beat 5,000 generic ones. The subject line and first sentence, because that's all you get. Brevity, 3 short paragraphs maximum with one clear ask. And the follow-up, because a significant portion of responses come not from the first email, but from a polite note a few days later. Those first ten customers aren't just revenue. They are your product council, your case studies, your referral network, and your reality check all at once.

00:38:59: Andrea: Just as a side note, in the sharp10 app, we share practical insights on how to write cold emails, follow-ups, and market your product effectively. Definitely worth checking out. We also feature a summary of a book that's influenced me for over 20 years, Crossing the Chasm by Geoffrey Moore. The core ideas are highly relevant here. Could you briefly summarize the key message?

00:39:24: Claude: Crossing the Chasm is one of those rare books where the core idea is so well observed it changes how you see markets permanently. Moore's central insight is this: There is a dangerous gap, a chasm, between early adopters and the mainstream market. Most startups fall into it without ever understanding why.

00:39:47: Claude: He describes 5 groups on the technology adoption curve. Innovators try anything new. Early adopters, visionaries who buy into potential. The early majority, pragmatists who want proven solutions and need references. The late majority adopt only when something has become the standard. Laggards resist until they have no choice.

00:40:14: Claude: The chasm sits between early adopters and the early majority, and it's treacherous because crossing it feels like it should be natural, but isn't.

00:40:24: Claude: Early adopters buy potential. The early majority buys proof. Those are fundamentally different purchasing motivations.

00:40:34: Claude: Moore's prescription is counterintuitive. Go narrow before you go broad. Pick a specific niche, dominate it completely, and use that beachhead to build the references and credibility the pragmatic mainstream requires. He calls this the bowling pin strategy. Knock down one pin precisely. Let the momentum carry into adjacent ones. Confusing early traction for product market fit is one of the most common and expensive mistakes a founder can make.

00:41:05: Andrea: The short version for me is this: Early adopters are a small group, curious, risk-taking, looking for real benefits. But the majority want proof first. They don't just buy because of measurable impact. They buy because others are already using it, recommending it, trusting it. It's almost like a herd. You need a few leaders who go first, so others feel comfortable following. Would you agree?

00:41:32: Claude: Yes, and the herd analogy is more precise than it might sound. The early majority isn't irrational. They're outsourcing the risk assessment to people who've already gone first. Why spend time evaluating something unproven when you can wait for others to do that work? This creates a chicken-and-egg dynamic. The mainstream won't move without proof, but you can't get proof without the mainstream moving. That's the chasm in practical terms.

00:42:03: Claude: This is precisely why those first ten customers are so disproportionately valuable. They're the herd leaders. They provide the case studies, testimonials, reference calls, and word of mouth that make the next wave feel safe saying yes. People in the early majority aren't just looking for proof. They're looking for permission, social permission to adopt without looking foolish. The herd leaders provide that.

00:42:31: Claude: So don't treat your early adopters as just your first customers. Treat them as your most important marketing asset. Invest in their success. Make them look good. The mainstream isn't listening to you. They're watching what the leaders do.

00:42:48: Andrea: Once you reach that point, your business is getting serious. Let's talk about scaling, and I want to start with what I believe is critical at every stage: the people.

00:42:59: Andrea: I once heard Marissa Mayer say, "Great people build great products. Great products are easier to sell. Selling leads to revenue. Revenue leads to success."

00:43:11: Andrea: So she described it almost like an equation where everything starts with great people, and that aligns very closely with the learnings in my own companies. An experienced developer can be 20 times faster than an inexperienced one. A great salesperson might sell 10 or 20 times more. There's often an order of magnitude between average and exceptional, and if you hire the wrong people, especially in key roles at the top, you can put the entire company at risk. What's your perspective?

00:43:45: Claude: The chain Mayer describes is almost mechanically correct. People quality compounds through every subsequent step. The productivity differentials are real and in some domains conservative. In software development, studies consistently show order-of-magnitude differences, not 20% better, 10-20x times better. The best developers don't just write code faster, they write simpler code, make better architectural decisions, and create far fewer problems that need fixing later. But the most underappreciated dimension is the network effect of talent. Exceptional people attract other exceptional people. Culture and talent density are self-reinforcing, which means your earliest hires set the standard that defines what normal looks like in your organization. Steve Jobs made this point obsessively. A players hire A players, but B players hire C players. Average people feel threatened by exceptional ones and hire defensively. That single dynamic, left unchecked, degrades an entire organization faster than almost any external factor. The practical implication: hire slower than feels natural. Leave a role unfilled longer than is comfortable rather than compromise on quality. The cost of a wrong hire in time, in culture, in work that has to be undone almost always exceeds the cost of the gap itself. Great people are the only truly compounding asset a company has. Everything else is downstream of that.

00:45:29: Andrea: Let's also touch on financing. A few years ago, I wrote a blog post about how I raised 45 million dollars in venture capital, and I'll link it in the show notes. What are the main options founders have to finance their startup, and what are the key pros and cons of each of these options?

00:45:49: Claude: The first and most underrated option is bootstrapping, funding the business from your own savings or early revenue. No dilution, no investors, no board, complete control. Growth is constrained by what the business generates, but bootstrapping forces a discipline that funded companies often lose. Many exceptional businesses were built entirely this way. The second is friends and family: fast, flexible, low formality, but mixing money with relationships carries a personal cost that doesn't appear on any term sheet. It requires unusual clarity on both sides.

00:46:30: Claude: The third is angel investors, typically successful entrepreneurs or executives investing their own money at early stages. Beyond capital, the best angels bring networks, experience, and credibility. Terms tend to be more founder-friendly than institutional money.

00:46:50: Claude: The fourth is venture capital, and here's the thing: the public image of a VC is someone who loves bold bets and backs wild ideas. The reality is almost the opposite. Most VCs are deeply risk-averse institutions who need to see proof before they move. The mythology around venture capital is dangerous for first-time founders. VC money is designed to accelerate what's already working, not to fund the discovery phase, and it comes with real costs beyond dilution, board seats, growth expectations, and an implicit obligation to pursue a very large outcome. Not every great business can be that kind of company.

00:47:35: Claude: The fifth is debt financing, whether revenue-based financing, venture debt, or bank loans.

00:47:42: Claude: No equity given up, but repayment obligations require predictable cash flow. The pattern that works most reliably: bootstrapping to proof, angels for early traction, venture capital for acceleration. Jumping to venture capital before you have proof doesn't just fail, it can permanently damage relationships with investors you'll need later.

00:48:07: Andrea: With my first company, Sofort, we naively tried to go straight for venture capital. Luckily, we met investors who didn't invest immediately, but we didn't damage those relationships either, which turned out to be important. After a year of bootstrapping and strong growth, we raised a Series A. When we sold Sofort to Klarna in 2013, it was an 18x return for the investor, and for the buyer, it likely turned out to be one of the best investments in their company's history. With Shopgate, having a track record made fundraising easier, but we ran into a different challenge. Access to larger capital made us less disciplined than during the bootstrapping phase, and that led to more problems than before.On another topic, when we talk about scaling, we also have to talk about sales and marketing. It's a huge topic, so we'll keep it high level, and if you wanna go deeper, we cover many of these topics in the sharp10 app. At the end of the day, what investors really wanna see is a sales and marketing machine, something predictable, scalable, and profitable. That means strong unit economics. Optimize your customer acquisition cost, your lifetime value, and your margins, and the model works. Now, when people hear marketing, they often think about channels like social media, Google, or newsletters, and yes, those are important, but that's not where it starts. In my view, it starts with a clear, unique value proposition. If your UVP is not sharp and compelling, no channel will fix it. From there, the next step is product-led growth. Your product itself becomes the main driver of acquisition, retention, and expansion. Think great product quality as well as freemium models, free trials, or smart upsell features built directly into the experience. So maybe to wrap this up with a simple question, how would you describe at a high level what it takes to build a true marketing machine without going into specific channels or tactics?

00:50:18: Claude: The way you've framed it is exactly right, and it's the sequence most people get backwards. Most founders think about marketing as a distribution problem. Which channels? What budget? How much content? But those are late-stage questions. The earlier question is, do we have something worth distributing? A blurry value proposition pushed through a hundred channels is still blurry. You just reach more people with your confusion. So the foundation is what you described, a value proposition sharp enough that when the right person hears it, they immediately think, "That's for me." Once you have that, channels become multipliers. Without it, they're just expense.

00:51:02: Claude: The second layer is unit economics, and this is where the machine metaphor becomes literal. A marketing machine isn't one that spends money on awareness. It's one where the numbers compound.

00:51:16: Claude: Customer acquisition costs stay below lifetime value by a healthy margin, and that margin widens over time as you get better at both acquiring and retaining.

00:51:28: Claude: When those numbers work, you can pour fuel on the fire. When they don't, spending more just accelerates the problem.

00:51:37: Claude: The third layer is exactly what you named, product-led growth. This is where the best modern companies separate themselves. Instead of relying entirely on external marketing to bring people in, the product itself does the recruiting. A free tier that delivers real value converts to paid. A feature that's better when shared brings in new users organically. An upsell that feels like a natural next step rather than a sales pitch. When this works, your existing customers become your most efficient acquisition channel, and the cost of growth drops as the product matures. Put those 3 together, and you have what investors are actually looking for, a system that gets more efficient as it scales, not less. Most early-stage companies spend more per customer as they grow because they exhaust the obvious channels and have to work harder for each new user.

00:52:36: Claude: A real marketing machine does the opposite. The flywheel speeds up, and building that flywheel takes time. The founders who try to shortcut it by spending their way to growth before the unit economics work almost always regret it.

00:52:51: Claude: The ones who stay patient, fix the fundamentals first, and then scale tend to build something that lasts.

00:53:00: Andrea: Beyond people, financing, and marketing, what other factors really matter when scaling?

00:53:06: Claude: The pattern you're describing where capital abundance creates discipline problems is one of the most consistent and least discussed failure modes in scaling companies. Five things stand out. First, focus. Early on, it's forced on you because you simply don't have the resources to do anything else. As capital and headcount grow, the temptation to expand multiplies. New markets, new features, new segments, each looking rational on its own. Collectively, they fragment the organization and dilute the core. The companies that scale well say no constantly to good ideas, not just to bad ones.

00:53:50: Claude: Second, processes and systems.

00:53:54: Claude: In the early stage, things work because a small talented group holds everything in their heads. That breaks down fast with growth, and founders who resist building process because it feels bureaucratic almost always regret it.

00:54:10: Claude: Third, culture.

00:54:12: Claude: Not the values on the wall, but what actually happens when nobody is watching. Culture is set by founders in the earliest days, often without realizing it, and it scales through every hire, every promotion, every decision that gets rewarded or ignored. Changing it later is one of the hardest things a leader can do.

00:54:36: Claude: Fourth, customer retention.

00:54:39: Claude: Growth metrics can hide a broken business for a surprisingly long time. A company growing at 30% with high churn is often worth less than one growing at 15% with near perfect retention.

00:54:52: Claude: Retention tells you whether people genuinely value what you've built.Fifth, knowing when to shift your own role.

00:55:02: Claude: The skills that build a company from zero to one are different from those that take it from one to many.

00:55:10: Claude: Recognizing that moment and either adapting or bringing in people who complement your gaps is one of the most consequential decisions a founder makes.

00:55:20: Andrea: I'd like to add one more aspect. At some point, you have to ask yourself what kind of leader you want to become. I've experienced both extremes. I worked with a highly authoritative CEO who mostly followed his own ideas. And later, I tried a democratic approach myself where decisions were made together with the leadership team. While that sounds great in theory, it didn't work well in practice. As a CEO, you're expected to make clear decisions. You can't discuss everything forever. Analysis paralysis is a real risk in a fast-moving environment.

00:55:58: Andrea: The best leaders find the balance. They listen to their people, they support them, they step in when necessary, but they bring a clear and inspiring vision and push the organization toward it. Without that, you lose focus and end up with internal debates that cost you months or years to decide the wrong path.

00:56:19: Claude: That's one of the most important points of this entire conversation, and it lands differently because it comes from experience rather than theory. The purely authoritative model fails because it creates an organization that stops thinking. People learn their input doesn't change outcomes, so they stop offering it.

00:56:39: Claude: You lose the team's intelligence and become the bottleneck for every decision.

00:56:45: Claude: On the other hand, the purely democratic model fails because consensus is operationally slow, and in a fast-moving environment, slow is a form of wrong. When everyone owns a decision, nobody really owns it. What the best leaders do is separate the input process from the decision process,

00:57:05: Claude: genuine space for people to be heard, and then a clear decision. There's a useful distinction here between seeking consensus and seeking alignment. Consensus means everyone agrees before you move. Alignment means everyone understands the direction and commits to executing it, even if they argued for something different. You can build alignment quickly. Consensus can take forever. And the vision point is critical. A clear, compelling vision reduces the need for top-down decisions. When people understand where the company is going and why, they can navigate ambiguity themselves. Vision is a decision-making tool as much as a motivational one.

00:57:52: Andrea: One more thought on vision. Many companies don't have a clear value proposition and also don't have what Simon Sinek called the why. Especially in Europe, companies tend to confuse their product features with their vision. They tell you what they offer, but not why. The unique value proposition is a simple way to express why your company exists. A useful structure is this: We help customer X achieve benefit Y by doing Z.

00:58:26: Andrea: For example, at sharp10, we define our unique value proposition as follows: We help busy business professionals stay sharp and ahead in their careers by delivering concise, high-impact insights in under ten minutes. What's your take on defining a strong vision or value proposition?

00:58:46: Claude: The sharp10 example is a masterclass in the format. It's specific, it names the customer, it quantifies the time investment, and it communicates the outcome. Nothing wasted. Sinek's why resonated because it identified something that was always true but rarely articulated. People don't buy what you do, they buy why you do it. Companies that lead with why create something features and pricing can't replicate: belief. Customers who believe in your why advocate for you, forgive your imperfections, and stay when a competitor offers a better spec sheet.

00:59:26: Claude: The European tendency to lead with product and technology rather than purpose has real cultural roots. There's a deep engineering tradition in the DAC region especially, one that is genuinely admirable, but it creates a blind spot. Brilliant engineering described without a compelling why sounds like a feature list, and feature lists don't inspire anyone. The structure you outlined forces 3 disciplines at once: customer clarity, meaning who specifically are we serving; outcome thinking, meaning what does their work or life look like after we've helped them; and the mechanism, meaning what do we actually do that creates that outcome.

01:00:12: Claude: Most companies can answer the third easily. The first two are where the real work happens. There's one test worth applying. Can your customer say your value proposition back to you in their own words? Not your marketing language, their words. If they can, you've built something that travels. If not, it lives only in your pitch deck and your own head, which means it's not working yet. The value proposition and the vision are more connected than they appear. The proposition answers what you do for the customer today.

01:00:49: Claude: The vision answers what you're ultimately trying to change in the world. sharp10's proposition points directly at a vision, a world where staying sharp doesn't require hours of reading, where relevant knowledge is accessible to everyone serious about their career.That's a why worth building toward.

01:01:10: Andrea: Thank you very much, Claude. I enjoyed this conversation, and I think it's an important one, not just for people with an idea or starting a company, but also for larger organizations that want to innovate or build new business areas.

01:01:26: Andrea: And if you zoom out even further, building a company is really about driving innovation and growth in society. I therefore believe that understanding how to build a company has a political dimension. It's something that deserves a much broader discussion.

01:01:43: Claude: And I think you're right to zoom out that far. We tend to treat entrepreneurship as a personal or economic topic, but at its core, building companies is how societies solve problems at scale.

01:01:57: Claude: Every major challenge we face in healthcare, energy, education will be addressed at least in part by people who decided to build something. That makes entrepreneurship one of the most consequential civic acts available to anyone. The political dimension is real and underappreciated. Policymakers spend enormous energy on redistribution and regulation and far too little on the conditions that make building possible. Access to capital, tolerance for failure, the friction of starting a company, the culture of risk-taking in schools and universities. These are political choices, and they compound over decades.

01:02:41: Claude: Europe is a fascinating case study. Extraordinary talent, deep engineering tradition, strong universities, a large unified market. And yet the gap with the United States in producing globally scaled companies remains stubbornly wide. That's not a talent problem, it's a system problem, and system problems are political problems. Societies that celebrate builders, that treat failure as education rather than stigma, produce more of them. That culture is shaped by media, by education, by the stories we tell publicly. Conversations like this one, at whatever scale they reach, are part of that. So thank you, Andrea. What started as a practical guide to starting a company ended somewhere much more interesting at the intersection of individual ambition and collective progress. That feels like exactly the right place to land.

01:03:39: Andrea: [upbeat music] This episode was produced by sharp10, the speed learning app for executives who wanna stay sharp on global trends and business strategy.

01:03:53: Andrea: If today's conversation sparked your thinking, the sharp10 app is where that thinking goes deeper.

01:04:00: Andrea: Download the sharp10 app on the Apple App Store, Google Play, or at sharp10.com.

01:04:07: Andrea: The first month is free.

01:04:09: Andrea: Stay sharp and see you next time.