Eric Ries — Lenny's Knowledge Sketch

Core Concept

A Startup Is an Experiment

"A startup is a human institution designed to create a new product under conditions of extreme uncertainty."

Every startup plan is a set of untested hypotheses
The goal is to learn what is true as fast as possible
Experimentation is not optional — it is the job
If you can't fail, you can't learn

Framework

MVP: Minimum Necessary to Test Your Riskiest Assumption

Write your feature list — cut it in half, then cut it in half again
MVP ≠ cheap: sometimes the right MVP takes 10 years (LTSE took exactly that)
High quality, wrong value prop = nothing to be proud of
Users taught Eric what IMVU was for — not the team

90%

of v1 work thrown away after a pivot

100×

more work than necessary — most teams' natural instinct

The IMVU teleport story

Eric's team couldn't afford proper avatar walking animation. They made avatars "teleport" with a janky cut. They felt embarrassed. Customers called it "more advanced than The Sims." The constraint became the feature. They skipped inverse kinematics for years.

Eric's rule Try being uncomfortably small in what you build. Most people are off by one or two orders of magnitude. If you're not embarrassed, you waited too long.

Deep Dive

The Pivot and the Zombie Company

What a pivot actually is:

Change in strategy, not in vision — one foot anchored in what you've learned
Founders who say they never pivoted almost always did; the psychological defense kicks in
Building a startup is self-discovery: the vision is found, not decreed at day one
Slack started as a game. Segment stumbled onto analytics. Facebook nearly got abandoned by Zuck himself while it was already working

"If you're asking whether you should pivot, you probably already know the answer. When it's working, there's no time for naval-gazing."

The 6-week pivot protocol

Give yourself a fixed window. Have everyone say: "If I could start over, I'd be building…" Often the whole team secretly wants the same new thing and nobody was willing to go first.

The zombie company trap:

A startup failure is not even in the top 10 worst outcomes
Far worse: a company that won't die, that you hate but can't leave
Building something that becomes a maligned force in the world — and having to pretend you weren't involved — is the real tragedy
The founder mental health crisis is real, and the so-called "successes" are often the worst cases

The key question "If you could wave a magic wand and start a new company today, would it be this one?" If the answer is not yes — it's time to move.

LTSE: the failure that unlocked everything

Two years of work on a partnership agreement — hundreds of pages, millions spent — all thrown away when the SEC withdrew approval. It was the best thing that ever happened. It forced the pivot to what actually worked.

Tactics

Build a Company That Can Be Trusted

Ask your lawyer: "If Philip Morris offered $1 more per share, do I have a fiduciary duty to say yes?" Most standard articles of incorporation say yes — and that should horrify you
Structural options: Public Benefit Corp, foundation-controlled entity, board mission pledge, LTSPV financing instrument
Encode values in governance, not just culture docs — so the company keeps its promise even if you're replaced
It's always "too early" until it's too late: the four-day window never arrives

The Toyota lesson Long-term governance is a competitive advantage, not charity. Companies insulated from quarterly activist pressure compound faster. Amazon's customer centricity and long-term thinking are structural, not cultural.

Eric's governance test File quarterly report on time: 100% certain. Avoid complicity in harm to users: "probably." That gap is the governance problem.

Contrarian

Lean Startup Myths — Still Wrong After 12 Years

✗ MVP means build something cheap and scrappy INSTEAD → ✓ MVP means the minimum needed to test your riskiest assumption. Sometimes that takes 10 years and $18M. "Minimum" is relative to the hypothesis, not the budget.

✗ Lean Startup is opposed to bold vision INSTEAD → ✓ Lean is the method for testing whether your bold vision is actually true. Experimentation doesn't replace vision — it's how you find out if the vision is real.

✗ Having your startup fail is the worst outcome INSTEAD → ✓ Failure isn't even top 10. A zombie company you hate but can't leave — or building something that becomes a maligned force in the world — is far, far worse.

✗ Doing the right thing costs profit INSTEAD → ✓ Companies built around human flourishing crush their competitors on compounding returns. We've defined profit wrong. Truly additive value creation and extractive rent-seeking are not the same thing — we just account for them identically.