First Principles Thinking + Confirmation Bias + Network Effects

You asked customers what they wanted. That’s why you failed.

Smart founders don’t fail because they’re incompetent. They fail because they ask the wrong questions. They interview customers, hear what individuals want, build exactly that, launch with no traction, and wonder what went wrong.

Winners see systems. Losers see surveys.

Key Concepts

Mental Model: First Principles Thinking

Strip everything to foundational truths. Ignore what “everyone knows.” Ask: what’s physics-level true here? Elon doesn’t ask “how do rockets work today?” He asks “what’s the cost of atoms that make a rocket?” Then builds from there. First principles bypass common wisdom that’s usually just common failure.

Cognitive Bias: Confirmation Bias

Your brain is a lawyer, not a scientist. It defends what you already believe. You ask customers leading questions. You see patterns that confirm your hunches. You ignore evidence that goes against your idea. You’re not gathering truth. You’re gathering proof for the case you’ve already decided.

Systems Concept: Network Effects

Value multiplies with users. One phone = useless. Two phones = useful. A billion phones = civilization infrastructure. Individual users can’t tell you this. They don’t see the system; they only see their experience. Network effects are emergent properties invisible to the parts inside the system.

Interconnections

  • You interview customers to validate your idea (confirmation bias active)
  • Customers tell you what they want individually (they can’t see network effects)
  • You build what they asked for (confirmation bias satisfied)
  • Product launches, users don’t compound value for each other (no network effects)
  • You built exactly what people wanted and it still failed

The trap: Asking customers what they want is asking people inside the system to see the system. They can’t. You have to think from first principles to see what emerges when many users interact.

Founder Applications

Example 1: Facebook vs. Friendster

Friendster asked users what they wanted. Users said: “More features. More customization. More profile options.”

Friendster delivered. Added features. It became cluttered. Died.

Facebook used first principles: “What’s deeply true about social networks?” Answer: Value comes from real identity + network closeness. Not features.

They stripped everything down. Real names only. College networks only. Limited features. Pure network effects.

Confirmation bias would’ve killed them, college kids wanted more features too. First principles revealed: the network IS the product. Features are noise.

Lesson: Customers request features. First principles reveal systems.

Example 2: Google vs. AltaVista

1999. AltaVista was the best search engine. They asked users what they wanted. Users said: “Portal features. Email. News. Shopping. Weather.”

AltaVista became a portal. Cluttered interface. Search became secondary.

Google used first principles: “What’s the fundamental job of search?” Answer: Find the right result fast. Everything else is distraction.

Clean interface. No portal features. Pure search. But here’s the key, they saw network effects in links. More websites linking = better ranking algorithm = better results = more users = more websites wanting to be indexed.

The network effect was in the web itself, not the search features users requested.

Lesson: Users see their immediate needs. First principles see the underlying force multiplier.

Example 3: Dropbox vs. File Storage Solutions

Before Dropbox, enterprise file storage companies asked IT departments what they wanted. Answer: “Security. Permissions. Compliance. Enterprise controls.”

They built exactly that. Complex. Expensive. Painful.

Dropbox used first principles: “What’s fundamentally true about file storage?” Answer: Files need to exist everywhere you work, seamlessly.

They were built for simplicity. Then discovered the viral loop (network effect): Shared folders mean recipients need Dropbox too. Every power user creates multiple new users. User growth compounds geometrically.

Confirmation bias trap: If they’d asked users pre-launch, users would’ve requested features competitors already had. First principles revealed: Distribution IS the product innovation.

Lesson: The breakthrough isn’t in the feature set. It’s in how growth compounds through the system.

Mechanism: Why It Works

Your brain takes shortcuts. It has to, because handling every piece of info from scratch would stop you.
would paralyze you.

So it builds models. “Customers know what they want.” “If I build what people ask for, they’ll buy it.” These models feel like smart advice. They’re actually traps.

Confirmation bias hijacks customer research. You ask questions that confirm your model.

“Would you use this feature?” (Leading question, pushes for a yes.)

“How much would you pay?” (Hypothetical, meaningless.)

“What’s your biggest pain point?” (They tell you conscious problems, miss subconscious systems.)

First principles thinking breaks the model. It asks: “What if everything I believe is inherited from people who also failed?”

Strip to atoms. Physics. Math. Human nature. What’s irreducibly true?

Network effects are irreducibly true. Metcalfe’s Law: value of network = n². Facebook with 10 users = 100 value units. Facebook with 1,000 users = 1,000,000 value units. Facebook with 1 billion users = 1,000,000,000,000,000,000 value units.

But individual users can’t tell you this. They experience step-by-step value (“I like seeing my friends’ photos”). You have to see the skyrocketing system they’re creating together.

The math: 94% of billion-dollar companies have network effects. 6% of founder customer research asks about network effects.

The question “what do you want?” captures individual preference. The question “what’s simply true?” captures system dynamics. One builds features. One builds growing empires.

Philosophical Bridge

Socrates walked around Athens asking “why?” He didn’t ask people what they wanted. He questioned what they assumed.

“Justice is giving people what they deserve.”
“Why?”
“Because that’s what justice means.”
“But why does it mean that?”

Layer by layer, stripping away inherited beliefs until only basic truth remained. That’s first principles thinking from 400 BCE.

Founders inherit beliefs from other founders. “Talk to customers.” “Build what they ask for.” “Customer is always right.”

These feel true because everyone repeats them. Socratic method reveals: they’re just consensus, not physics.

Plato’s Cave: Prisoners see shadows on the wall, think shadows are reality. They’re not. Reality is outside the cave but prisoners can’t see it because they’re chained inside. Your customers are in the cave. They see their individual experience (shadows). Network effects exist outside (reality). Confirmation bias chains you in the cave with them. First principles thinking walks you outside.

Today’s Reframe

Customers tell you what they see. First principles reveal what creates what they see.

Reflection Prompt

What assumption are you defending because everyone in your industry believes it, not because you’ve questioned it from first principles?