Definition
Know which decisions you will not be able to take back
Most startup decisions are cheap to undo. A small number are not, and those are the ones that decide the company.
Companies rarely fail on the decisions they debate longest. They fail on commitments that quietly removed future options: a hire, a price, a term sheet, an architecture, a market. Naming irreversibility before committing is the highest-leverage habit in the model.
Direct answer
What is irreversible decisions?
An irreversible decision is a commitment that cannot be undone cheaply or quickly, because reversing it costs capital, time, credibility, relationships or options the company cannot replace.
Deeper explanation
Common false signals
- Treating a decision as reversible because it can be technically undone, ignoring credibility and relationship cost
- Assuming a hire is reversible because employment can be ended
- Believing pricing can be raised later without consequence for existing buyers
- Calling a financing term standard rather than reading what it removes
Founder implications
- Before committing, you state what reversing this would cost in money, months and trust.
- You spend deliberation time in proportion to irreversibility, not to how interesting the decision is.
- You buy optionality when the evidence is thin, instead of choosing faster.
Investor implications
- You look at which options a company has already spent.
- You judge process by whether irreversibility was named before commitment.
Board implications
- You separate the agenda into reversible and irreversible items and give them different scrutiny.
- You require the evidence standard to be met only where reversal is expensive.
Common irreversible commitments in startups
- Senior hires and the organisational structure built around them
- Pricing and packaging visible to the existing base
- Financing terms, control rights and cap-table structure
- Architecture and data model choices that later products depend on
- Public positioning and the market a company becomes known for
What it proves and does not prove
Naming irreversibility proves only that the cost of reversal was considered before the commitment. It does not prove the decision was correct, and it does not remove the risk.
Where this sits in the book and the OS ebooks
More Scars Than Trophies explains irreversible decisions as theory: why the distinction exists and how it changes judgement. The Proof Stack OS ebooks turn it into operating practice, and each OS ebook has its own separate role-specific AI chatbot.
- Proof Stack Daily OS: The founder's operating system for turning proof into daily decisions.
- Proof Stack Company OS: The company and board operating system for evidence-based governance.
Evidence boundaries
What this page does not claim
The limits of the model are stated on the page, not buried in a footnote.
- Irreversibility is a matter of degree and context, not a binary label.
- The model does not tell you which decision to take. It tells you how much evidence the decision deserves.
Audience implications
What changes for each reader
The same evidence standard, read from three different seats.
Founders
- Before committing, you state what reversing this would cost in money, months and trust.
- You spend deliberation time in proportion to irreversibility, not to how interesting the decision is.
- You buy optionality when the evidence is thin, instead of choosing faster.
Investors
- You look at which options a company has already spent.
- You judge process by whether irreversibility was named before commitment.
Boards
- You separate the agenda into reversible and irreversible items and give them different scrutiny.
- You require the evidence standard to be met only where reversal is expensive.
Author
Petri Lehmuskoski
Petri Lehmuskoski has founded, scaled, repaired and exited companies over more than four decades. He is Founding & General Partner of Gorilla Capital.
Disagree with something here, or want it covered from another angle? Send a short note.
Related
Continue through the model
- Stage-matched proof
Stage-matched proof is external behaviour strong enough to justify the next irreversible decision at the company's current stage.
- The Proof Stack
The Proof Stack is an ordered model of the evidence a company must hold before a decision that is difficult to reverse: a Discovery stage followed by four Proof layers (Buyer Proof, Value Proof, Scaling Proof and Exit Proof).
- Governance without theatre
Governance without theatre is board practice in which decisions are tested against a stated evidence standard before they become irreversible, rather than reviewed after the fact through presentations.
A plain-text version of this evidence layer, generated from the same page content, is available at /concepts.md.
Read the chapter that starts the argument
Chapter 1: Most Founders Do Not Fail Because They Make Bad Decisions. Free to read now, no email required.