Lenny RAG MCP Server

{ "episode": { "guest": "Andrew Wilkinson", "expertise_tags": [ "Startup Strategy", "Business Acquisition", "Holding Company Management", "Bootstrapping", "AI Automation", "Entrepreneurship", "Mental Health", "ADHD", "Founder Psychology" ], "summary": "Andrew Wilkinson, co-founder and CEO of Tiny (a Berkshire Hathaway-style holding company), discusses strategies for identifying great startup ideas, avoiding common pitfalls, and building profitable businesses. He emphasizes finding underserved niches ('fish where the fish are'), the importance of unfair advantages, and when to bootstrap versus raise venture capital. Wilkinson shares his extensive experience with AI automation tools, his journey from anxiety and depression despite wealth accumulation, and his recent ADHD diagnosis. He reveals how SSRIs and ADHD medication transformed his mental health more than his billion-dollar net worth ever did, and discusses the implications of AI on future employment and job displacement.", "key_frameworks": [ "Fish Where the Fish Are (Munger/Buffett)", "Lazy Leadership", "Business vs Job distinction", "Moats (Network Effects, Brand, Switching Costs)", "The Joneses Effect (Social Comparison)", "Palm Treo Phase of AI", "Easy Choices Hard Life framework" ] }, "topics": [ { "id": "topic_1", "title": "Coming Up with Great Startup Ideas", "summary": "Discussion of frameworks for identifying viable startup ideas, including Charlie Munger's 'fish where the fish are' philosophy, finding niches with less competition, and the importance of matching ideas to personal unfair advantages rather than pursuing universal appeal.", "timestamp_start": "00:00:00", "timestamp_end": "00:07:19", "line_start": 1, "line_end": 56 }, { "id": "topic_2", "title": "Why Starting Small Creates Competitive Advantage", "summary": "Andrew explains the gym analogy of not trying to deadlift 300 pounds on day one. He discusses building experience through simple businesses first, getting immediate positive feedback, and compounding success rather than competing in saturated markets like cafes or project management software.", "timestamp_start": "00:07:30", "timestamp_end": "00:09:59", "line_start": 58, "line_end": 67 }, { "id": "topic_3", "title": "Avoiding Jobs Disguised as Businesses", "summary": "The distinction between a scalable business and a job for yourself. Andrew shares how operating a business requires delegation and scale to avoid being trapped doing unfulfilling work forever, using examples like the pressure washing business and cafes.", "timestamp_start": "00:09:59", "timestamp_end": "00:12:00", "line_start": 67, "line_end": 75 }, { "id": "topic_4", "title": "Physical vs Software Business Trade-offs", "summary": "Exploring when to pursue non-software businesses like pressure washing or restaurants versus tech businesses, and how your unfair advantages (sales ability, taste, capital) should guide the choice rather than industry type.", "timestamp_start": "00:12:00", "timestamp_end": "00:15:19", "line_start": 75, "line_end": 87 }, { "id": "topic_5", "title": "Finding Problems Worth Solving with Market Demand", "summary": "How to identify whether a market actually has paying customers and understanding customer willingness to pay. Andrew discusses studying industries to understand economics, using realtors and government assistance programs as examples.", "timestamp_start": "00:15:19", "timestamp_end": "00:22:29", "line_start": 87, "line_end": 124 }, { "id": "topic_6", "title": "Business Models to Avoid Despite Popular Appeal", "summary": "Andrew identifies business ideas that consistently fail despite seeming attractive: bars, newspapers, restaurants, and other capital-intensive or operationally complex models. The lesson that business model matters more than talent or management.", "timestamp_start": "00:22:29", "timestamp_end": "00:25:32", "line_start": 124, "line_end": 144 }, { "id": "topic_7", "title": "Bootstrap vs Venture-Backed Business Models", "summary": "Discussion of the false dichotomy between lifestyle/bootstrap businesses and venture-backed companies. Tiny grew to $300M revenue bootstrapped. The key is choosing the right market: low-competition, defensible niches work fine without VC, while highly competitive spaces require massive funding.", "timestamp_start": "00:25:32", "timestamp_end": "00:31:04", "line_start": 144, "line_end": 175 }, { "id": "topic_8", "title": "Business Moats and Competitive Advantages", "summary": "Deep dive into moat types: brand loyalty, network effects, high switching costs. Examples include Letterboxd (network effect), Coca-Cola (brand), Salesforce (switching costs). Why moats matter for long-term business sustainability and resisting competition.", "timestamp_start": "00:31:38", "timestamp_end": "00:36:35", "line_start": 177, "line_end": 208 }, { "id": "topic_9", "title": "Hiring Philosophy and People Management", "summary": "Core principle: people problems are business problems. Hire fully-formed people for what you need now, not potential. Fire quickly if you doubt someone. Don't try to change people or coach them into different personalities. CEOs are like elephants you can't steer.", "timestamp_start": "00:36:35", "timestamp_end": "00:41:37", "line_start": 208, "line_end": 234 }, { "id": "topic_10", "title": "AI Automation Stack and Tools", "summary": "Andrew's primary automation tool is Lindy.ai for building email agents, calendar management, and workflow automation. Other tools include Replit for web development, Limitless for life recording, ChatGPT, Gemini, and Perplexity for various tasks. He's built dozens of agents to automate routine work.", "timestamp_start": "00:42:41", "timestamp_end": "00:49:32", "line_start": 241, "line_end": 311 }, { "id": "topic_11", "title": "Practical AI Workflow Examples", "summary": "Detailed examples of AI agents in action: email triage reducing inbox by 20%, simple decision automation, calendar agents with emojis, location-based contact suggestions. The level of personalization and automation replacing what used to require a full-time assistant.", "timestamp_start": "00:49:32", "timestamp_end": "00:54:22", "line_start": 311, "line_end": 357 }, { "id": "topic_12", "title": "Job Displacement and Future of Work in AI Era", "summary": "Andrew discusses the Palm Treo phase of AI, where tools exist but aren't accessible to everyone yet. Predictions on job displacement for researchers, translators, and admin roles. Discussion of Dario Amodei's 2027 prediction about AI surpassing PhDs. Uncertainty about long-term employment landscape.", "timestamp_start": "00:54:22", "timestamp_end": "01:02:06", "line_start": 357, "line_end": 411 }, { "id": "topic_13", "title": "Advice for Young People in an AI World", "summary": "Get good with AI tools early. Build wealth using AI to enable diversification into compute and energy. Expect a long window before robotics replaces all work. New jobs will emerge that don't exist today. The fundamental question: do all jobs become a single prompt?", "timestamp_start": "01:02:06", "timestamp_end": "01:04:37", "line_start": 411, "line_end": 428 }, { "id": "topic_14", "title": "The Happiness Paradox and Wealth", "summary": "Andrew's journey from earning $60K/year to being worth over $1 billion and discovering that money doesn't solve unhappiness. The realization that anxiety, depression, and comparison loops (the Joneses effect) persist regardless of wealth. Success in business doesn't translate to life satisfaction.", "timestamp_start": "01:04:37", "timestamp_end": "01:10:05", "line_start": 428, "line_end": 445 }, { "id": "topic_15", "title": "Mental Health Interventions That Actually Worked", "summary": "Three major changes: reframing money toward philanthropy, reducing possessions and visible wealth, and medication for anxiety/depression. SSRIs provided more relief than a billion dollars. ADHD medication transformed his focus and executive function. Mental health treatment more impactful than external success.", "timestamp_start": "01:10:05", "timestamp_end": "01:14:05", "line_start": 445, "line_end": 460 }, { "id": "topic_16", "title": "The Case for SSRI and Psychiatric Medication", "summary": "Andrew advocates for destigmatizing SSRIs by comparing them to Tylenol or antihistamines. People avoid mental health medication due to fear but take other drugs that affect the brain without hesitation. Understanding genetics through 23andMe helps minimize side effects. Medication addresses the root cause of anxiety loops.", "timestamp_start": "01:14:05", "timestamp_end": "01:16:40", "line_start": 460, "line_end": 472 }, { "id": "topic_17", "title": "ADHD Diagnosis and Impact", "summary": "Andrew was diagnosed with ADHD through neurological testing despite initial skepticism. 30% of entrepreneurs have ADHD vs 5% population rate. Taking stimulant medication dramatically improved focus and executive function. ADHD also explains relationship friction. Early diagnosis and treatment beneficial.", "timestamp_start": "01:16:56", "timestamp_end": "01:21:05", "line_start": 474, "line_end": 492 }, { "id": "topic_18", "title": "Lightning Round: Books, Media, and Philosophy", "summary": "Recommended books: Laws of Human Nature by Robert Greene (personality types and psychology), How to Get Rich by Felix Dennis (cautionary tale about wealth). Favorite movie: Challengers. Product: Matic robot vacuum. Life motto: 'Easy choices, hard life; hard choices, easy life.'", "timestamp_start": "01:21:05", "timestamp_end": "01:24:07", "line_start": 493, "line_end": 525 }, { "id": "topic_19", "title": "Steve Jobs Encounter and Early Entrepreneurship", "summary": "Andrew met Steve Jobs at age 17 at an Apple Store tour in 2003. Jobs shook his hand and Andrew stayed at his side asking questions about iMacs. The lesson: ask big things and maybe you'll get something great. Demonstrates the power of initiative and being first.", "timestamp_start": "01:24:07", "timestamp_end": "01:26:26", "line_start": 527, "line_end": 548 }, { "id": "topic_20", "title": "Tiny's Business Model and Founder Acquisition", "summary": "Tiny buys profitable, high-quality businesses with competitive advantages. They try not to change anything, letting existing management run the company. Focus on businesses with happy customers, happy employees, and something positive for the world. Examples include AeroPress, Letterboxd, Dribble.", "timestamp_start": "01:26:26", "timestamp_end": "01:28:03", "line_start": 548, "line_end": 570 } ], "insights": [ { "id": "i1", "text": "Fish where the fish are. The biggest mistakes I've made have been going into business models where other people have repeatedly failed and thinking I can do this better.", "context": "Charlie Munger's core advice on finding underserved markets", "topic_id": "topic_1", "line_start": 8, "line_end": 14 }, { "id": "i2", "text": "Competition equals lower margins. The more competitors there are, the lower your prices have to be and the more competitive the business is ultimately.", "context": "Why niche markets are economically superior", "topic_id": "topic_1", "line_start": 53, "line_end": 54 }, { "id": "i3", "text": "You don't want to walk into the gym on day one and try and deadlift 300 pounds. Starting a highly regulated, complicated, competitive business as a first-time founder is like trying to lift too much weight.", "context": "Why bootstrapping simple ideas matters", "topic_id": "topic_2", "line_start": 1, "line_end": 2 }, { "id": "i4", "text": "I got immediate positive feedback and I built my own narrative. I'm good at business, I can do this, keep going.", "context": "How early wins compound into entrepreneurial confidence", "topic_id": "topic_2", "line_start": 62, "line_end": 62 }, { "id": "i5", "text": "There's a big difference between a business and a job. If you can get to a scale where you can drive 10 leads a day, then you don't have to do any of the pressure washing and you just do what you love.", "context": "How to distinguish scalable businesses from self-employment", "topic_id": "topic_3", "line_start": 71, "line_end": 72 }, { "id": "i6", "text": "Lazy leadership: How do I get away from the things I hate as quickly as humanly possible? How do I be Teflon for tasks?", "context": "Delegation philosophy for founder success", "topic_id": "topic_3", "line_start": 74, "line_end": 74 }, { "id": "i7", "text": "What's your unfair advantage and what are you great at? Look around and see where your passion is and then sniff around within it, probably somewhere within your passion there's an opportunity.", "context": "How to find profitable niches within areas you care about", "topic_id": "topic_4", "line_start": 79, "line_end": 82 }, { "id": "i8", "text": "I follow my passions and spend a long time learning different industries and then I find the profitable niche within.", "context": "Strategy for discovering hidden opportunities in familiar domains", "topic_id": "topic_4", "line_start": 86, "line_end": 86 }, { "id": "i9", "text": "Being able to know what people would pay to solve the problem realistically is incredibly valuable. I can intuit that a realtor would be willing to spend $5,000 a lead because they make $20-50K per house sale.", "context": "How to estimate market value of solutions", "topic_id": "topic_5", "line_start": 101, "line_end": 104 }, { "id": "i10", "text": "I think it's mostly gut. I've trained on all this data of what works and what doesn't and what's a good moat. When I see it, I just immediately know.", "context": "Intuition develops from studying thousands of business models", "topic_id": "topic_5", "line_start": 109, "line_end": 110 }, { "id": "i11", "text": "You should be able to use AI to break down any business model. Ask Claude or ChatGPT: 'I'm thinking about starting a Botox clinic. Can you break down the numbers? What's hard about it? What is the regulatory moat?'", "context": "How to leverage AI for rapid business model validation", "topic_id": "topic_5", "line_start": 110, "line_end": 110 }, { "id": "i12", "text": "Boring is good. A business that makes $30M a year helping people fill out government forms for assistance—nobody wakes up and goes 'I want to make form filling software' but they would if they could make $20M a year.", "context": "Why unsexy businesses outperform sexy ones", "topic_id": "topic_6", "line_start": 120, "line_end": 122 }, { "id": "i13", "text": "You can't take a brilliant management team and change a bad business model. Ultimately the business model wins.", "context": "Business model determines success more than team quality", "topic_id": "topic_6", "line_start": 131, "line_end": 131 }, { "id": "i14", "text": "Enough technology has changed that Instacart can succeed where Webvan failed. But would it be easier to start Instacart, Amazon, or Coupang, or would it be easier to start an enterprise SaaS software company? Definitely the enterprise SaaS software company.", "context": "Context-dependent business feasibility", "topic_id": "topic_6", "line_start": 137, "line_end": 137 }, { "id": "i15", "text": "The only difference between what we do and what a venture capitalist does is the level of tolerance of burning money on fire.", "context": "Why bootstrapping profitable businesses can scale without VC", "topic_id": "topic_7", "line_start": 149, "line_end": 149 }, { "id": "i16", "text": "The decision ultimately comes down to how hairy do you want your big hairy audacious goal to be? If you want to build a satellite company, you need VC. If you want to solve a non-hyper-competitive problem, you can scale to hundreds of millions without raising money.", "context": "Framework for choosing between bootstrap and VC routes", "topic_id": "topic_7", "line_start": 149, "line_end": 152 }, { "id": "i17", "text": "Things has succeeded by being intentional: no AI stuff, no API, not on Android. Very focused. The founder has an incredible life with recurring revenue and gets to work with headphones on building beautiful software people love.", "context": "Success doesn't have to mean maximum market share", "topic_id": "topic_7", "line_start": 164, "line_end": 167 }, { "id": "i18", "text": "A moat is basically a brand, network effect, or high switching costs. Something that gives you pricing power or makes competition unlikely because switching to alternatives is costly or undesirable.", "context": "Defining competitive moats for long-term business value", "topic_id": "topic_8", "line_start": 185, "line_end": 189 }, { "id": "i19", "text": "Network effects: if someone else wants to compete with us, why would someone go to their social network that has a small number of users when all their friends are already on Letterboxd?", "context": "How network effects create defensibility", "topic_id": "topic_8", "line_start": 187, "line_end": 188 }, { "id": "i20", "text": "There are no problems. There is only people problems. When we have a bad actor, a psycho, or a narcissist, all bets are off and life becomes incredibly stressful.", "context": "Chris (business partner) insight on root cause of business failure", "topic_id": "topic_9", "line_start": 212, "line_end": 212 }, { "id": "i21", "text": "If I ever think 'should I fire this person' even once, I should fire them immediately. It has always been a signal that something is wrong.", "context": "Heuristic for personnel decisions", "topic_id": "topic_9", "line_start": 215, "line_end": 215 }, { "id": "i22", "text": "Hire for what you need now, not for potential. You need to hire someone who already is fully formed and can do what you need. You can never mentor someone into being a good employee.", "context": "Contrarian hiring philosophy against potential-based hiring", "topic_id": "topic_9", "line_start": 218, "line_end": 218 }, { "id": "i23", "text": "To a man with a hammer, everything looks like a nail. A CEO will pursue their background expertise even if the business needs something different. They're an elephant and you're the rider—they're going to go wherever they want to go.", "context": "Why CEOs execute on their beliefs, not your directions", "topic_id": "topic_9", "line_start": 224, "line_end": 227 }, { "id": "i24", "text": "People will shoot themselves in the foot. If you tell them an idea in a board meeting they don't really want, they usually sandbag it and don't put their heart into it. They want their idea to work.", "context": "Why top-down management directives fail", "topic_id": "topic_9", "line_start": 233, "line_end": 233 }, { "id": "i25", "text": "It's like having the world's most reliable employee who costs $200 a month and works 24/7. So many knowledge work jobs are going to change massively.", "context": "The economic advantage of AI automation", "topic_id": "topic_10", "line_start": 20, "line_end": 20 }, { "id": "i26", "text": "I try to take every single thing a human could do in my inbox and automate it with Lindy. I've replaced my full-time assistant with AI agents.", "context": "Comprehensive automation approach to knowledge work", "topic_id": "topic_11", "line_start": 245, "line_end": 278 }, { "id": "i27", "text": "If someone emails me and says 'Do you want to get lunch?' the agent emails me privately with multiple choice options. I just say a number and it emails you back as me with a thoughtful email.", "context": "How AI handles simple social decisions", "topic_id": "topic_11", "line_start": 275, "line_end": 278 }, { "id": "i28", "text": "When I look at my calendar 30 minutes before a meeting, an agent goes on Perplexity and does a deep dive on the person, gets the meeting context from email, and texts me their background and our commonalities.", "context": "AI research automation for better meetings", "topic_id": "topic_11", "line_start": 356, "line_end": 356 }, { "id": "i29", "text": "We are in the Palm Treo phase. Right now people like me who have the time and skillset can build agents, but it's not accessible to everybody. In five years, ChatGPT will be like Her from the movie—a digital employee.", "context": "Current state of AI accessibility and near-term evolution", "topic_id": "topic_12", "line_start": 380, "line_end": 383 }, { "id": "i30", "text": "Dario Amodei said by 2027, AI models will be smarter than all PhDs in any subject. He's been a very conservative voice, so when he says that, it's worth listening to.", "context": "Conservative AI researcher's bold prediction on timeline", "topic_id": "topic_12", "line_start": 386, "line_end": 386 }, { "id": "i31", "text": "People generally overestimate things in the short term and underestimate them in the long term. There's going to be a long window where robotics is nowhere near good enough.", "context": "Reality check on AI/automation timelines", "topic_id": "topic_13", "line_start": 404, "line_end": 404 }, { "id": "i32", "text": "The fundamental question is: do all jobs just become a single prompt? Does a CEO just say 'grow the business while making customers happy and turning a profit' and the AI runs the whole company?", "context": "The ultimate question about job displacement", "topic_id": "topic_13", "line_start": 400, "line_end": 400 }, { "id": "i33", "text": "Whatever's in your brain, whatever anxiety loop you have, doesn't go away just because you have a bigger house or more money in the bank account or you're in Bali. That's a chemical reaction inside you related to your past and DNA.", "context": "The fundamental truth about wealth and happiness", "topic_id": "topic_14", "line_start": 434, "line_end": 434 }, { "id": "i34", "text": "I went from $15-20M revenue to almost $300M revenue and I'm still stressed about the same things. At one point being worth over a billion dollars, and I'm still just as anxious as ever.", "context": "Personal experience with the failure of wealth to bring happiness", "topic_id": "topic_14", "line_start": 437, "line_end": 437 }, { "id": "i35", "text": "The Joneses effect: we're all comparing ourselves to our peers. A single-digit multibillionaire said 'Jeff Bezos is so fucking rich because he can buy a super yacht' while being worth $5 billion himself.", "context": "Social comparison persists at all wealth levels", "topic_id": "topic_14", "line_start": 440, "line_end": 441 }, { "id": "i36", "text": "I reframed money so that 90% of my money goes into philanthropic foundation and gets given away to great causes. The win isn't for me, it's about doing something good.", "context": "How reframing money toward impact improved life satisfaction", "topic_id": "topic_15", "line_start": 449, "line_end": 450 }, { "id": "i37", "text": "The more stuff I had, the more houses, the more people managing it, the less happy I was. I'm owned by my stuff. I stopped spending as much money.", "context": "Possessions as sources of stress and unhappiness", "topic_id": "topic_15", "line_start": 453, "line_end": 453 }, { "id": "i38", "text": "No amount of money or success or attention has done what this little tiny yellow pill could do for my mental state. For the first time in my life I felt relief.", "context": "SSRI effectiveness compared to all external achievements", "topic_id": "topic_15", "line_start": 458, "line_end": 458 }, { "id": "i39", "text": "My brain went from Times Square to a quiet library. ADHD medication transformed my focus and executive function.", "context": "Vivid description of medication impact on ADHD", "topic_id": "topic_15", "line_start": 458, "line_end": 458 }, { "id": "i40", "text": "You can actually do a 23andMe DNA test and find out how you metabolize different SSRIs. Then you can choose one that you metabolize well and have no side effects.", "context": "How to optimize medication choice for minimal side effects", "topic_id": "topic_16", "line_start": 470, "line_end": 470 }, { "id": "i41", "text": "If you take Tylenol or antihistamine, you're changing your brain chemistry. Yet people are scared of SSRIs. It doesn't make sense. Anxiety was ruining my life. Every minute of every day I was convinced something terrible would happen.", "context": "Destigmatizing psychiatric medication through logic", "topic_id": "topic_16", "line_start": 464, "line_end": 465 }, { "id": "i42", "text": "About 30% of entrepreneurs have ADHD versus 5% of the general population. Entrepreneurs love new things, jump between topics, describe themselves as unemployable.", "context": "ADHD as entrepreneur trait, not deficit", "topic_id": "topic_17", "line_start": 479, "line_end": 480 }, { "id": "i43", "text": "At work you can delegate around ADHD, but at home it shows up. My girlfriend asks me to take garbage out, I forget three times. She sees it as hurtful when it's actually a neurological issue.", "context": "ADHD impact on personal relationships", "topic_id": "topic_17", "line_start": 482, "line_end": 483 }, { "id": "i44", "text": "Getting diagnosed made me feel not broken. I always felt a bit broken but couldn't put my finger on why. Now I understand I have poor executive function as a neurological trait.", "context": "Psychological relief of diagnosis and understanding", "topic_id": "topic_17", "line_start": 485, "line_end": 486 }, { "id": "i45", "text": "Easy choices, hard life. Hard choices, easy life. When I make the hard choice to shut down a business, fire someone, or say bye to a project, it's almost always right and makes life easier.", "context": "Jerzy Gregorek's life philosophy", "topic_id": "topic_18", "line_start": 524, "line_end": 525 }, { "id": "i46", "text": "Ask big and maybe you'll get something great. I emailed Apple PR asking to interview Steve Jobs. They said no, but offered a tour, and I ended up spending time with him.", "context": "Power of bold requests and door-in-the-face technique", "topic_id": "topic_19", "line_start": 539, "line_end": 540 } ], "examples": [ { "id": "ex1", "explicit_text": "At Airbnb, we did X... (referenced implicitly through comparison)", "inferred_identity": "Not explicitly mentioned but Andrew references common business patterns founders encounter", "confidence": "low", "tags": [ "marketplaces", "business models", "community platforms" ], "lesson": "How competitive advantage works in marketplace businesses", "topic_id": "topic_1", "line_start": 1, "line_end": 50 }, { "id": "ex2", "explicit_text": "I started a pizzeria and I lost all my money. I started a designer cat furniture business, a online DJ school, a skin cream business, all of these things.", "inferred_identity": "Andrew Wilkinson's personal ventures", "confidence": "high", "tags": [ "failure stories", "bootstrapping", "learning from mistakes", "business selection", "first-time founder" ], "lesson": "Starting in competitive, capital-intensive businesses leads to failure; need simple wins first to build momentum", "topic_id": "topic_2", "line_start": 62, "line_end": 62 }, { "id": "ex3", "explicit_text": "My web design agency, which became Metalab, was so easy and it worked immediately. All I had to do was know how to build websites and be able to talk to potential customers.", "inferred_identity": "Andrew Wilkinson's first successful business (Metalab)", "confidence": "high", "tags": [ "services business", "first success", "founder origin story", "bootstrapping", "simplicity" ], "lesson": "Simple businesses with immediate feedback loops create confidence and momentum for future ventures", "topic_id": "topic_2", "line_start": 59, "line_end": 62 }, { "id": "ex4", "explicit_text": "I just started a pressure washing business. I was speaking at a local business school and after I spoke, one of the kids walked up to me and he said 'Hey I'm an entrepreneur. I've started two or three businesses in the past doing landscaping'", "inferred_identity": "Andrew Wilkinson's pressure washing business with a young entrepreneur", "confidence": "high", "tags": [ "non-software business", "scaling services", "delegation", "student founder", "mentorship" ], "lesson": "Service businesses can scale when you remove yourself from doing the work and focus on sales and marketing; media ownership creates unfair advantage", "topic_id": "topic_3", "line_start": 68, "line_end": 69 }, { "id": "ex5", "explicit_text": "A friend of mine owns a restaurant and he said it's this stunning beautiful restaurant in my hometown, but it's really just a job for a few different people and we can't make any money at it. But he noticed vendors like grease trap cleaning and exhaust vent cleaning for kitchens are making a killing.", "inferred_identity": "Andrew's friend who owns a restaurant (unnamed)", "confidence": "medium", "tags": [ "restaurant business", "profitability analysis", "adjacent opportunities", "passion vs profit", "B2B services" ], "lesson": "Look within industries you love for profitable niches; B2B service businesses to industries are often more profitable than consumer-facing businesses", "topic_id": "topic_4", "line_start": 80, "line_end": 83 }, { "id": "ex6", "explicit_text": "I met the founder of Letterboxd and I realized this is a business with a moat, a network effect, a huge social network for film reviewers. I could invest in film now.", "inferred_identity": "Letterboxd founder (implicit: Tom Holloway)", "confidence": "medium", "tags": [ "acquisition", "passion alignment", "network effects", "Tiny portfolio", "film industry" ], "lesson": "Following passion and domain expertise over time allows you to recognize opportunities with true competitive advantages", "topic_id": "topic_4", "line_start": 86, "line_end": 87 }, { "id": "ex7", "explicit_text": "We ended up buying the AeroPress coffee maker company. I used to be a barista and we ended up buying it.", "inferred_identity": "AeroPress company (Tiny acquisition)", "confidence": "high", "tags": [ "product acquisition", "passion-driven investment", "Tiny portfolio", "consumer products", "barista background" ], "lesson": "Passion and background knowledge can identify acquisition opportunities missed by others; AeroPress fit his coffee interests", "topic_id": "topic_4", "line_start": 86, "line_end": 86 }, { "id": "ex8", "explicit_text": "A UVic student interested in marketing found local clients doing social media for small restaurants, making a thousand bucks a month with lots of work. I said if you pivot to realtors or wealth managers with large budgets, you can charge $5,000 a month.", "inferred_identity": "University of Victoria student (Andrew's acquaintance)", "confidence": "medium", "tags": [ "social media services", "client profiling", "pricing strategy", "B2B vs B2C", "market positioning" ], "lesson": "Same skill set can command 5x pricing when applied to higher-value customer segments; choose customers, not just niches", "topic_id": "topic_4", "line_start": 91, "line_end": 93 }, { "id": "ex9", "explicit_text": "When I started my web design firm, I found a job board in San Francisco where startups would share projects. I could charge five times the amount for five times less work.", "inferred_identity": "Andrew Wilkinson's early web design business (pre-Metalab)", "confidence": "high", "tags": [ "market segmentation", "pricing discovery", "founder insight", "B2B", "geographic arbitrage" ], "lesson": "Premium markets (San Francisco startups) pay significantly more than local markets for the same service quality", "topic_id": "topic_4", "line_start": 95, "line_end": 95 }, { "id": "ex10", "explicit_text": "A realtor can make $20-50,000 selling a single house, so I can intuit they'd be willing to spend $5,000 a lead if I have a unique way of getting them clients that convert at a high rate.", "inferred_identity": "Real estate market analysis (industry study)", "confidence": "high", "tags": [ "real estate", "economics analysis", "customer willingness to pay", "market sizing", "lead generation" ], "lesson": "Understanding customer economics deeply tells you what they can afford to pay for solutions", "topic_id": "topic_5", "line_start": 101, "line_end": 104 }, { "id": "ex11", "explicit_text": "Flow was basically Asana before Asana came out. We made the mistake of going after an industry everybody goes after. We lost $10 million trying to compete with venture-backed businesses while bootstrapping.", "inferred_identity": "Andrew Wilkinson's failed product management software (Flow)", "confidence": "high", "tags": [ "project management software", "failed startup", "venture competition", "loss learning", "competitive markets" ], "lesson": "Competing with well-funded founders in sexy categories is essentially unwinnable for bootstrapped founders; avoid these markets", "topic_id": "topic_6", "line_start": 116, "line_end": 117 }, { "id": "ex12", "explicit_text": "Asana was run by the co-founder of Facebook and had raised hundreds of millions of dollars. Me competing with them was like Fiji deciding to invade the United States—completely silly in retrospect.", "inferred_identity": "Asana founded by Dustin Moskovitz (Facebook co-founder)", "confidence": "high", "tags": [ "venture-backed competition", "founder pedigree", "capital advantage", "market dynamics", "lessons from failure" ], "lesson": "Well-connected founders with massive capital and startup pedigree are nearly impossible to compete against directly", "topic_id": "topic_6", "line_start": 119, "line_end": 119 }, { "id": "ex13", "explicit_text": "There was a business making $30 million a year helping people fill out forms to get government assistance. Software that fills out forms for you—if you're disabled and need government funding, pay us $1000 and get a $20,000 grant.", "inferred_identity": "Anonymous government assistance form-filling SaaS (industry example)", "confidence": "low", "tags": [ "government services", "boring business", "high margin", "low competition", "profitable niche" ], "lesson": "Unsexy businesses with real customer pain and willingness to pay vastly outperform cool-sounding businesses", "topic_id": "topic_6", "line_start": 120, "line_end": 122 }, { "id": "ex14", "explicit_text": "Me and one of my best friends about 10 years ago wanted to start a bar. We thought we're tech guys, we know how to build systems, we're good at business. We were utterly humbled.", "inferred_identity": "Andrew Wilkinson and unnamed best friend (personal story)", "confidence": "medium", "tags": [ "bar business", "failure story", "operational complexity", "arrogance lesson", "business model difficulty" ], "lesson": "Operational complexity and logistical execution beat management theory; hospitality businesses have hidden complexity tech founders underestimate", "topic_id": "topic_6", "line_start": 128, "line_end": 128 }, { "id": "ex15", "explicit_text": "I ended up buying an old paper in Vancouver in the local news business. It's the exact same thing—you can't take a brilliant management team and change a bad business model. Ultimately the business model wins.", "inferred_identity": "Vancouver local newspaper (Tiny or related venture)", "confidence": "medium", "tags": [ "news business", "publishing", "failed business model", "organizational structure limits", "market structure" ], "lesson": "Even with great management, structurally broken business models cannot succeed; better to avoid them entirely", "topic_id": "topic_6", "line_start": 131, "line_end": 131 }, { "id": "ex16", "explicit_text": "Webvan failed, but Instacart succeeded. Yet Instacart was still harder to build than enterprise SaaS. Technology changed enough that it could work.", "inferred_identity": "Webvan (failed grocery delivery), Instacart (successful), Amazon, Coupang (international variants)", "confidence": "high", "tags": [ "grocery delivery", "logistics business", "technology requirements", "timing", "business model viability" ], "lesson": "Context and technology change can make previously failed business models viable; but structurally complex businesses still remain harder than knowledge work", "topic_id": "topic_6", "line_start": 137, "line_end": 137 }, { "id": "ex17", "explicit_text": "We bootstrapped the entire business and now across all our companies we do almost $300 million in revenue. Tiny owns over 40 businesses including Dribble, WeCommerce, AeroPress.", "inferred_identity": "Tiny (Andrew's holding company)", "confidence": "high", "tags": [ "holding company", "bootstrap growth", "acquisition model", "portfolio companies", "founder exit acquirer" ], "lesson": "Bootstrapped businesses can scale to hundreds of millions in revenue without VC if they pick profitable niches with moats", "topic_id": "topic_7", "line_start": 149, "line_end": 150 }, { "id": "ex18", "explicit_text": "Things is a task management app run by 1-3 people, bootstrapped, intentionally focused. Founder has 10,000 true fans and an incredible life. Asana built a multi-billion company. Both are success, but for different reasons.", "inferred_identity": "Things (task management app by Cultured Code)", "confidence": "high", "tags": [ "task management", "bootstrap business", "niche product", "founder lifestyle", "market dominance vs lifestyle" ], "lesson": "Success has multiple definitions; founder happiness and autonomy may matter more than market dominance", "topic_id": "topic_7", "line_start": 158, "line_end": 167 }, { "id": "ex19", "explicit_text": "Things probably makes between $5-25M revenue, which to a VC is not even worth considering. VCs need a story where it's worth $300M to $1B.", "inferred_identity": "Things (bootstrapped task management app)", "confidence": "high", "tags": [ "bootstrap business", "market sizing", "VC hurdle rate", "sustainable business", "profitable niche" ], "lesson": "Markets too small for VC ($5-25M revenue) are often perfect for bootstrap founders seeking lifestyle and autonomy", "topic_id": "topic_7", "line_start": 172, "line_end": 173 }, { "id": "ex20", "explicit_text": "We bought Serato. Serato is the largest DJ software company in the world. If you see a DJ with a laptop, probably using Serato.", "inferred_identity": "Serato (DJ software company, Tiny acquisition)", "confidence": "high", "tags": [ "DJ software", "market leadership", "acquisition", "Tiny portfolio", "passion-aligned business" ], "lesson": "Market-leading niche products with passionate user bases are excellent acquisition targets with built-in competitive advantage", "topic_id": "topic_8", "line_start": 200, "line_end": 200 }, { "id": "ex21", "explicit_text": "I used to DJ so I was aware of them and it was one of those moments where I understood the business. They had a really interesting moat of this huge passionate user base and deep hardware integration.", "inferred_identity": "Andrew Wilkinson's DJ background and Serato acquisition", "confidence": "high", "tags": [ "domain expertise", "founder advantage", "user passion", "integration moat", "acquisition rationale" ], "lesson": "Personal domain expertise allows you to recognize competitive advantages invisible to generalist investors", "topic_id": "topic_8", "line_start": 206, "line_end": 206 }, { "id": "ex22", "explicit_text": "I've probably started or been involved with 75 different projects or businesses where I've been a primary contributor. I've been an inch deep and a mile wide.", "inferred_identity": "Andrew Wilkinson's entrepreneurial journey", "confidence": "high", "tags": [ "serial entrepreneur", "experimentation", "business patterns", "learning through diversity", "founder style" ], "lesson": "Broad exposure to different business models teaches pattern recognition and failure avoidance better than depth in one domain", "topic_id": "topic_9", "line_start": 44, "line_end": 44 }, { "id": "ex23", "explicit_text": "I interviewed Steve Jobs at age 17 at an Apple Store tour in New York at Macworld in 2003. He shook my hand and said 'Hi, I'm Steve.' I stayed at his side and picked his brain about the new 17-inch iMac.", "inferred_identity": "Steve Jobs (Apple founder, interviewed by young Andrew Wilkinson)", "confidence": "high", "tags": [ "Steve Jobs", "founder hero", "Apple history", "serendipity", "initiative", "early entrepreneurship" ], "lesson": "Being first, asking bold questions, and showing initiative can create unexpected opportunities with influential people", "topic_id": "topic_19", "line_start": 530, "line_end": 533 }, { "id": "ex24", "explicit_text": "Lindy is a tool that lets you build AI workflows and agents. When I get an email about my kid's school, the agent automatically puts it on my calendar with notes for things I need to prepare.", "inferred_identity": "Lindy.ai (AI workflow automation tool)", "confidence": "high", "tags": [ "AI automation", "workflow tools", "email management", "productivity", "API-based automation" ], "lesson": "API-based AI tools enable non-technical founders to build sophisticated automation reducing time on administrative work", "topic_id": "topic_10", "line_start": 244, "line_end": 245 }, { "id": "ex25", "explicit_text": "Replit is a vibe coding platform using Claude 4. I can say 'build a website for my sound software business' or 'build a Python web app that does XYZ' and style it like Stripe or in the tone of Malcolm Gladwell.", "inferred_identity": "Replit (AI coding platform)", "confidence": "high", "tags": [ "AI coding", "web development", "design system", "tone matching", "rapid prototyping" ], "lesson": "AI-powered code generation enables non-engineers to build functional web apps and websites, eliminating dependency on developers", "topic_id": "topic_10", "line_start": 302, "line_end": 303 }, { "id": "ex26", "explicit_text": "Limitless clips to your shirt and records everything I do. I can ask 'what did I promise to people today?' or use it in a fight with my girlfriend asking it to play couples counselor and analyze the conflict.", "inferred_identity": "Limitless (AI wearable recording device)", "confidence": "high", "tags": [ "wearable AI", "life recording", "memory augmentation", "relationship tool", "personal data" ], "lesson": "Recording and AI analysis of personal interactions provides insights for self-improvement in relationships and commitments", "topic_id": "topic_11", "line_start": 319, "line_end": 323 }, { "id": "ex27", "explicit_text": "I've trained Gemini 2.5 on all my medical records and I can say 'I'm tired and stressed, what supplement should I take?' and it reminds me of all my medications to avoid conflicts.", "inferred_identity": "Andrew Wilkinson's personal health data with Gemini", "confidence": "high", "tags": [ "medical data", "AI health advisors", "personalization", "drug interactions", "long-context models" ], "lesson": "Large context windows in modern models enable personalized health advice by integrating individual medical history", "topic_id": "topic_11", "line_start": 337, "line_end": 339 }, { "id": "ex28", "explicit_text": "I used to have a full-time researcher doing prep for podcast interviews at $400-500 bucks per guest. Now I use ChatGPT deep research which is better and costs almost nothing.", "inferred_identity": "Lenny Rachitsky's former research contractor", "confidence": "medium", "tags": [ "research automation", "job displacement", "cost savings", "AI replacement", "podcast preparation" ], "lesson": "AI deep research capabilities can replace expensive consultant research while providing better results", "topic_id": "topic_12", "line_start": 353, "line_end": 353 }, { "id": "ex29", "explicit_text": "I went through my entire medicine cabinet, took photos of all the medications, and now Claude or ChatGPT can reference them whenever I ask a health question.", "inferred_identity": "Andrew Wilkinson's personal health documentation", "confidence": "high", "tags": [ "health data", "medication tracking", "personalization", "AI context", "memory augmentation" ], "lesson": "Simple documentation of personal information enables AI tools to provide highly personalized advice", "topic_id": "topic_11", "line_start": 338, "line_end": 339 } ] }