Here’s why you’re probably chasing the wrong idea

Youtube URL: https://youtu.be/VxAwUb86MUE?si=ia08mbYpKa7zh2gY

Host(s): Lenny Rachitsky

Guest(s): Andrew Wilkinson (Co-founder and CEO, Tiny)

Podcast Overview and Key Segments

Overall Summary

Andrew Wilkinson has started or been deeply involved in 75+ businesses and now runs Tiny, a holding company often called the Berkshire Hathaway of the internet. In this episode, he shares a practical playbook for finding startup ideas, avoiding crowded markets, and building businesses with moats. He argues for “fishing where the fish are,” favoring boring, profitable niches over flashy, overfished spaces. He explains how Tiny buys durable businesses with network effects (e.g., Letterboxd) and strong brands (e.g., AeroPress), and why people problems are the only real problems. Andrew also details his AI stack (Lindy, Replit, Limitless, Claude, ChatGPT, Gemini) and how agents already act like reliable $200/month employees. He explores AI’s impact on jobs, the likely timeline, and how to stay relevant. Finally, he discusses money and happiness, his shift to philanthropy, and how SSRIs and an ADHD diagnosis changed his life.

Reference

• Moat: A durable advantage that protects a business (brand, network effects, switching costs).
• Network effects: A product becomes more valuable as more people use it.
• Switching costs: Pain of moving to a new solution keeps customers in place.
• SSRIs: Antidepressants that can reduce anxiety and depression.
• ADHD: A neurodevelopmental disorder affecting focus and executive function.
• Palm Treo: Pre-iPhone smartphone; used here to describe “early, clunky” phase before a breakthrough.
• “iPhone moment”: A clear leap in usability that brings tech to the masses.
• Compute: Processing power needed to run advanced AI models.
• API: A way for software to connect and share data.
• CRM: Software to track customers and relationships.
• LLM (large language model): AI system trained on vast text to perform tasks.
• Agent: An AI that takes actions across tools using rules and context.
• Deep Research (ChatGPT): Automated research runs to synthesize sources.

Key Topics

Finding startup ideas: Fish where the fish are

Andrew urges founders to seek niches with demand and less competition. Avoid spaces where millions wake up wanting the same idea (e.g., cafes). Look for unsexy, under-served markets like funeral homes, pest control, or government form-filling tools. Start with a domain you care about and where you have an unfair edge. Then go deeper to find the profitable corner within it. Andrew backed Letterboxd because it matched his passion for film and had a strong network effect. He bought AeroPress after years as a barista. His rule: start simple, get an early win, and build confidence. Don’t deadlift 300 lbs on day one. Choose baby weights, and compound.

Avoid creating a job you hate: Scale away from the task

A job is you doing the work. A business is systems and people doing the work. Many founders trap themselves in jobs (e.g., running a cafe) with no escape. Andrew focuses on achieving scale fast. He calls this “lazy leadership” and being “Teflon for tasks.” If the model supports 10+ leads a day, you can hire and step back from the wrench-turning. He started a pressure washing business with a student and designed it so the founder could focus on sales or marketing, not hoses. The key is to build a path to delegation from day one. If it cannot scale beyond you, don’t start it.

Software vs. non-software: Profit hides in the supply chain

You can be a digital native and still win in physical businesses. The trick is to find where the money pools. Andrew’s restaurateur friend noticed vendors—grease trap cleaners, vent cleaners—made more money than many restaurants. Use your skills (e.g., sales, design taste, distribution) to find the profitable wedge inside your passion. Andrew loves film and coffee; he invested where moats exist (Letterboxd’s network effects, AeroPress’ brand). Start with your Venn diagram of skills and unfair advantages. Then pivot toward the buyer who can and will pay more (e.g., realtors vs. restaurants for social media).

Crowded markets vs. boring riches

Andrew lost $10M trying to compete with Asana by bootstrapping Flow. Lesson: don’t invade the U.S. with Fiji’s army. Crowded, sexy categories compress margins and demand massive capital. Meanwhile, boring can be beautiful. He cites a company making tens of millions helping families fill government forms and charging a slice of the grant. No one dreams of building it, which is the point. Look for cash-rich pain, low glamor, and real urgency. A heuristic: if many great teams tried and died, ask why you would beat the odds now. Instacart beat Webvan, but tech shifts like that are rare.

Bootstrapped vs. venture-backed: Both can be big

Tiny bootstrapped to hundreds of millions in revenue. Andrew sees no hard cap if you pick moaty, non-commodity businesses. Venture capital is a tool for ideas that need burn and speed (e.g., regulated, frontier tech). Otherwise, bootstrapping lets compounding work without dilution or fire drills. He contrasts Asana (VC) with Things (bootstrapped) to show both paths can “win,” depending on the founder’s goals. Things stayed small, focused, and beloved. Define your game: market share at all costs, or an incredible life, product, and customer base.

Buying moaty businesses: The Tiny playbook

Andrew read Buffett, then shifted from starting to buying. He looks for businesses that are hard to mess up and can be held forever. Moats matter most: brand and network effects. He cites Letterboxd (film social network) and Serato (DJ software with deep hardware ties). Switching costs can be a moat too (e.g., Salesforce), though it’s less consumer-friendly. Tiny’s integration style is “do no harm.” Keep teams, culture, and playbooks intact. If a founder leaves, hire a CEO very aligned with how the business actually grows.

People and hiring: There are only people problems

Andrew’s rule: if you wonder even once whether to fire, you probably should. Hire for “is,” not potential. Superstars make firing unthinkable. For CEOs, alignment is everything. To a person with a hammer, everything looks like a nail. If you need organic marketing and the candidate only sees enterprise sales, don’t hire them. “They’re an elephant, you’re the rider.” You cannot drag them where they don’t want to go. Also, avoid psychos and narcissists. Most disasters are people disasters. Filter hard up front, be direct, and cut fast.

Andrew’s AI stack and agents: A $200/month super-employee

Andrew runs his day on AI agents, led by Lindy (workflows + agents), plus Replit (ship apps), Limitless (life recorder), Claude (writing), ChatGPT (general + Deep Research), Gemini (large context, medical). Key Lindy workflows: email triage (auto-archive, urgency labels, quick replies), calendar rules, emoji tagging, CRM enrichment, and travel-triggered meetups. A pre-meeting agent scrapes context from email and runs Perplexity research, then texts a briefing. Replit lets him build and deploy websites and Python apps solo. Limitless records life, enables “what did I promise today?” and even relationship coaching prompts.

AI and jobs: Palm Treo era to iPhone moment

We are early. Building agents is possible, but clunky. The “iPhone moment” will be when anyone can say, “Help me run my business,” and an AI sets up data access, tools, and workflows. Some jobs are already changing (researchers, translators, assistants). If models reach “smarter than all PhDs” by 2027 (as Anthropic’s Dario Amodei suggested), most knowledge work will shift. Advice: get excellent with the tools, build wealth while advantage exists, and diversify into compute and energy. Expect many new service roles in a world of abundance. Don’t bet on robots raking your leaves soon, but do prepare.

Money and happiness: Decoupled, for real

Andrew chased numbers to billionaire status and still felt anxious. The “loop” followed him, not the money. He reframed wealth toward philanthropy, reduced spending and complexity, and chose a lower-visibility lifestyle. The largest change came from addressing mental health: SSRIs turned down the volume on anxiety; ADHD diagnosis and meds improved focus and relationships. He now sees many driven founders as ADHD outliers who can build systems at work but struggle at home. Fix the internal state first. No car or house can mute an unquiet mind.

Key Themes

1) Boring > Sexy (if you want profits)

• Explanation: Sexy markets attract elite competition, raise CAC, and compress margins. Boring markets hide real pain and stable willingness to pay. Find niches where urgency is high and glamour is low. The fewer dreamers, the better the unit economics.
• Quotes:
• • “Fish where the fish are.”
  • “Nobody wakes up and goes, ‘I want to make form-filling software’—until they see $20M a year.”

2) Start with unfair advantage

• Explanation: Map your Venn diagram of skills, passions, and access. Then find the profit wedge in that space. Andrew loved film and coffee but invested where moats lived (Letterboxd network effects, AeroPress brand). Sales is portable across industries. Your edge determines your market.
• Quotes:
• • “Follow your passions, then sniff around for the profitable niche within.”
  • “Luck comes to the prepared mind.”

3) Moats, not heroics

• Explanation: Tiny buys businesses you can’t easily screw up. Brand, network effects, and switching costs defend cash flows. Integration should be minimal. Great CEOs already see the path you want. Your job is capital, culture, and not breaking it.
• Quotes:
• • “I’m looking for a business so good it’s hard to mess up.”
  • “We generally just leave them alone.”

4) People problems are the only problems

• Explanation: Most failures trace to misaligned or toxic people. Hire for what the job needs today. If you waver on someone, cut fast. For CEOs, alignment beats pedigree. Elephants go where they want. Your role is selection, not persuasion.
• Quotes:
• • “There are no problems—only people problems.”
  • “If I think once about firing someone, I should.”

5) Agents are already real leverage

• Explanation: With Lindy + core LLMs, a solo operator can run email, calendar, CRM, research, and scheduling with near-zero human help. Agents do work, not just drafts. This is a preview. When set-up vanishes, everyone will have a digital employee.
• Quotes:
• • “It’s like having the world’s most reliable employee who costs $200 a month.”
  • “We’re in the Palm Treo phase.”

6) Wealth ≠ peace; treat the mind

• Explanation: More money doesn’t fix anxiety loops. Reduce complexity. Reframe wealth to serve others. Consider meds and diagnosis if anxiety or ADHD are present. It can change relationships and work quality. Shame keeps many from relief.
• Quotes:
• • “No amount of money did what this little yellow pill did.”
  • “I always felt a bit broken—now I have empathy for myself.”

Key Actionable Advise

Key Problem

You keep picking crowded, sexy markets.

• Solution Target boring, under-served niches with real urgency.
• How to Implement List 10 unsexy industries. Interview buyers. Price the pain. Build the simplest solution. Sell first.
• Risks to be aware of Regulatory hurdles. Low-tech adoption. Underestimating ops complexity.

Key Problem

You created a job, not a business.

• Solution Design for scale and delegation from day one.
• How to Implement Set a lead target (e.g., 10/day). Document SOPs. Hire frontline roles early. Track time and offload.
• Risks to be aware of Hiring too fast. Margin compression. Founder reluctance to let go.

Key Problem

Your hiring bets keep missing.

• Solution Hire for “is,” not potential. Align CEO’s default playbook to the business.
• How to Implement Pressure test candidates on their chosen growth motion. Use work samples. Check deep references.
• Risks to be aware of Bias toward charisma. Ignoring culture fit. Slow to cut.

Key Problem

You drown in email, meetings, and context switching.

• Solution Adopt agents to handle admin, research, and routine communication.
• How to Implement Start with Lindy for email triage, replies, calendar rules. Add a pre-meeting research agent. Use Replit to ship small tools.
• Risks to be aware of Privacy and data access. Over-automation without review. Broken edge cases.

Key Problem

Unsure whether to bootstrap or raise.

• Solution Match funding to market structure and moat needs.
• How to Implement Ask: Is this winner-take-most? Regulated? Requires speed? If yes, consider VC. Else, bootstrap with compounding.
• Risks to be aware of Raising for status. Starving a moaty product. Misreading competition intensity.

Key Problem

Money is not moving the happiness needle.

• Solution Reduce complexity. Reframe goals. Treat the mind.
• How to Implement Simplify assets. Commit a % to philanthropy. Talk to your doctor about SSRIs. Consider ADHD testing.
• Risks to be aware of Medication side effects. Stigma. Over-correction into under-ambition.

Noteworthy Observations and Unique Perspective

• “Don’t deadlift 300 lbs on day one.” Start small, win early, build confidence.
• • Quote: “You don’t want to walk into the gym on day one and try and deadlift 300 lb.”
• A great founder fit is a precise Venn diagram, not a hot market.
• • Quote: “Follow your passions and find the profitable niche within.”
• Buying can beat building if you optimize for peace and compounding.
• • Quote: “Why am I not just buying businesses and letting them run?”
• CEO hiring is selection, not coaching.
• • Quote: “They’re an elephant—you’re the rider.”
• AI agents already replace a full-time assistant for $200/month.
• • Quote: “It’s like having the world’s most reliable employee.”
• SSRIs and ADHD meds can unlock peace and performance.
• • Quote: “For the first time in my life, I felt relief.”

Companies, Tools and Entities Mentioned

• Tiny, Metalab, Dribbble, WeCommerce, AeroPress, Letterboxd, Serato
• Flow, Asana, Things (by Cultured Code), Salesforce, Instacart, Webvan, Amazon, Coupang
• Lindy, Replit, Lovable, Bolt, v0
• Limitless
• Claude, ChatGPT (Deep Research, o3), Gemini 2.5, Perplexity
• Airtable, HubSpot, Gmail API
• Buffett, Munger, Apple, Steve Jobs
• Sponsors mentioned: Sauce, Enterpret, Miro

Linkedin Ideas

1. Title: Stop Chasing Sexy Ideas—Fish Where the Fish Are
2. • Main point: Boring niches with real urgency beat crowded, cool markets.
   • Core argument: Competition crushes margins; “unsexy” solves expensive pain.
   • Quotes: “Fish where the fish are.” “Nobody wakes up wanting to build form-filling software—until they see $20M a year.”
2. Title: From Job to Business: How to Escape the Founder Trap
3. • Main point: Build for delegation. If it can’t scale beyond you, don’t start.
   • Core argument: Design for leads, SOPs, and early hiring. Be “Teflon for tasks.”
   • Quotes: “There’s a big difference between a business and a job.” “Lazy leadership.”
3. Title: Bootstrapped vs VC: Pick the Game You Actually Want to Play
4. • Main point: Both paths can win. Align capital with market structure and moat.
   • Core argument: Moaty businesses can compound without burn; some spaces need VC.
   • Quotes: “We bootstrapped to hundreds of millions.” “Things vs Asana shows two valid wins.”
4. Title: My AI Agents Are a $200/Month Super-Employee
5. • Main point: Agents can now run email, calendar, CRM, and research reliably.
   • Core argument: Use Lindy + LLMs to automate knowledge work today.
   • Quotes: “We’re in the Palm Treo phase.” “It’s like having the world’s most reliable employee.”
5. Title: Money Didn’t Fix My Anxiety—Treatment Did
6. • Main point: Wealth ≠ happiness. Treat the mind first.
   • Core argument: SSRIs and ADHD meds can change life and work.
   • Quotes: “No amount of money did what this little yellow pill did.”

Blog Ideas

1. Title: The Moat Playbook: How Tiny Buys Businesses You Can’t Mess Up
2. • Main point: Brand, network effects, and switching costs protect cash flows.
   • Core argument: Buy and hold with minimal meddling works when moats are real.
   • Quotes: “I’m looking for a business so good it’s hard to mess up.”
2. Title: The Case for Boring: Why Unsexy Markets Print Cash
3. • Main point: Boring categories hide the best margins.
   • Core argument: Less competition + urgent pain = pricing power.
   • Quotes: “Competition equals lower margin.”
3. Title: Agents at Work: The Exact AI Workflows That Replaced My Assistant
4. • Main point: Practical agent stack with Lindy, ChatGPT, Perplexity, and Replit.
   • Core argument: Show email triage, pre-meeting briefs, CRM enrichment, travel nudges.
   • Quotes: “Before meetings, it texts me who they are, why we’re meeting, and what we share.”
4. Title: Fire Fast, Hire for ‘Is’: Wilkinson’s People Rules for Operators
5. • Main point: People problems drive most failures. Alignment > potential.
   • Core argument: CEOs are elephants; they go where they want. Select, don’t steer.
   • Quotes: “If I think once about firing someone, I should.”
5. Title: Billionaire, Anxious: What Finally Moved the Needle on Happiness
6. • Main point: Lifestyle changes help, but medical treatment can be decisive.
   • Core argument: Reframing wealth, reducing complexity, and SSRIs/ADHD meds.
   • Quotes: “I felt relief for the first time in my life.”