A counterfactual is a "what if" — a scenario that did not happen but could have. "If Kennedy had lived, would the Vietnam War have ended differently?" "If Blockbuster had acquired Netflix in 2000, would streaming have developed the same way?" Counterfactuals are essential for learning, accountability, and strategy. They are also systematically misused. The human mind is drawn to vivid, plausible alternatives — and often confuses plausibility with probability. We ask "what if?" but we rarely ask "how likely was that alternative?" or "what would have had to change for it to occur?"
In causal inference, counterfactuals are the gold standard. To say that X caused Y, you need to compare what happened (Y occurred when X occurred) with what would have happened if X had not occurred. The difference is the causal effect. The problem: we cannot observe the counterfactual. We cannot re-run history without Kennedy's assassination or with Blockbuster owning Netflix. We can only estimate — using randomised experiments, natural experiments, or statistical models. The quality of our conclusions depends on the quality of our counterfactual reasoning.
For decision-makers, counterfactuals serve two functions. First, they improve learning: "What would we have done differently?" surfaces actionable lessons. Second, they improve planning: "What if our assumption is wrong?" generates contingency scenarios. The trap is counterfactual drift — imagining alternatives that were never realistic, or attributing outcomes to factors that had negligible causal weight. The discipline is to make counterfactuals explicit, testable, and grounded in what was actually knowable at the time.
Section 2
How to See It
The pattern appears wherever people reason about cause and effect, assign blame or credit, or plan for uncertainty. The diagnostic: are they comparing what happened to a plausible alternative?
Business
You're seeing Counterfactuals when a board evaluates a CEO's performance by asking "would the company have done better under different leadership?" The question is inherently counterfactual — you cannot observe the alternative. The board's job is to construct a plausible counterfactual (e.g., industry benchmarks, peer performance) and compare. The mistake: assuming the CEO caused the outcome without specifying what the counterfactual would have been. The company might have done worse under anyone else — or better. You cannot know without a counterfactual.
Investing
You're seeing Counterfactuals when an investor asks "what if the Fed had cut rates sooner?" The question is counterfactual — it imagines a different policy path. The value is in understanding how sensitive outcomes are to the variable. The trap: overconfidence in the counterfactual. The Fed might have cut rates and inflation might have stayed high for different reasons. The counterfactual is a thought experiment, not a prediction. Use it to assess sensitivity, not to claim certainty.
Personal life
You're seeing Counterfactuals when someone regrets a decision and imagines "if I had taken that job, I would be happier now." The counterfactual is a story — plausible but unverifiable. The same person might have taken the job and been miserable for different reasons. Counterfactual thinking can motivate change ("I should have taken more risks") or trap you in rumination ("I could have had a better life"). The discipline is to extract the lesson without treating the imagined alternative as fact.
Strategy
You're seeing Counterfactuals when a team runs scenario planning: "What if our key supplier goes bankrupt? What if our main competitor launches a price war?" Each scenario is a counterfactual — a world that might occur. The value is in preparing for contingencies. The trap: spending too much time on scenarios that are vivid but unlikely, or too little on scenarios that are likely but boring. Good scenario planning weights counterfactuals by probability and impact.
Section 3
How to Use It
Use counterfactuals for learning and planning — but make them explicit, testable, and grounded in what was knowable at the time. Avoid counterfactual drift into imagined alternatives that were never realistic.
Decision filter
"Before attributing cause or drawing lessons, ask: what is the counterfactual? What would have happened if X had not occurred? Is that counterfactual plausible? What would have had to change for it to occur? If you cannot answer, your causal claim is weak."
As a founder
Build counterfactual thinking into post-mortems. When a project fails, ask "what would have had to be different for it to succeed?" — not "who is to blame?" The counterfactual surfaces actionable lessons: "If we had had a technical co-founder, we might have shipped faster." The next step is to test whether that counterfactual is true — would a technical co-founder have helped? — or whether it is a convenient attribution. The same logic applies to successes: "We won because we moved fast" — but what if the market had been different? The counterfactual keeps you honest.
As an investor
Evaluate investment theses through counterfactual reasoning. "This company will win because of X" — what if X does not materialise? What is the counterfactual world where the company fails? The exercise forces you to identify the key assumptions and their sensitivity. The best investors explicitly model counterfactuals: "Base case: X grows 20%. Bear case: X grows 5%. What would have to happen for each?" The counterfactual is not a prediction; it is a way to structure uncertainty.
As a decision-maker
When assigning credit or blame, specify the counterfactual. "She caused the failure" implies "if she had acted differently, we would have succeeded." Is that true? What would she have had to do differently? Could anyone have succeeded in that situation? The counterfactual forces you to distinguish between (a) outcomes that were caused by the person's choices and (b) outcomes that would have occurred regardless. The distinction matters for accountability and learning.
Common misapplication: Treating the counterfactual as fact. "If we had launched earlier, we would have won" — maybe. The counterfactual is a hypothesis. It might be valid, but it is not proven. The discipline is to treat counterfactuals as scenarios to test, not as conclusions to defend.
When Grove faced the decision to pivot Intel from memory chips to microprocessors in the mid-1980s, he used counterfactual reasoning. "If we got kicked out and the board brought in a new CEO, what do you think he would do?" he asked Gordon Moore. Moore's answer: "He would get us out of memories." Grove's conclusion: "Why shouldn't we walk out the door, come back, and do it ourselves?" The counterfactual — an imaginary new CEO with no emotional attachment to the memory business — revealed the right decision. Grove used the counterfactual to strip away the sunk cost and identity that made the pivot painful. He could not observe what a new CEO would do, but he could construct a plausible counterfactual and act on it.
Daniel KahnemanPsychologist, Nobel Prize in Economics 2002
Kahneman's work on counterfactual thinking revealed how "what if" scenarios shape emotion and judgment. In studies of Olympic medalists, bronze medalists (who imagined "what if I had come fourth?") were often happier than silver medalists (who imagined "what if I had won gold?"). The counterfactual — the alternative that was closest to reality — determined satisfaction more than the actual outcome. Kahneman applied this to decision-making: we evaluate outcomes by comparing them to counterfactuals, and the counterfactual we choose (the "reference point") shapes whether we feel gain or loss. Understanding this helps decision-makers structure the counterfactuals that will drive their own and others' evaluations.
Section 6
Visual Explanation
Counterfactuals — The unobservable comparison that defines causation. We cannot see what would have happened, but we can reason about it.
Section 7
Connected Models
Reinforces
Second-Order Thinking
Second-order thinking asks "and then what?" — a form of counterfactual reasoning. You imagine what would happen if your initial assumption were wrong, and you reason through the consequences. Counterfactuals and second-order thinking both use imagined scenarios to improve decisions.
Reinforces
[Inversion](/mental-models/inversion)
Inversion asks "what would cause failure?" — the counterfactual of success. By imagining the world where you fail, you identify what to avoid. Inversion is counterfactual reasoning applied to risk management.
Tension
Hindsight Bias
Hindsight bias makes the actual outcome seem inevitable — "we should have known." Counterfactual reasoning corrects for this by forcing us to imagine alternatives that were plausible at the time. The tension: hindsight bias suppresses counterfactuals; counterfactual discipline suppresses hindsight bias.
Tension
Causation vs Correlation
Causation requires a counterfactual — X caused Y only if Y would not have occurred without X. Correlation does not. The tension: we often infer causation from correlation without specifying the counterfactual. Proper causal inference requires explicit counterfactual reasoning.
Section 8
One Key Quote
"If we got kicked out and the board brought in a new CEO, what do you think he would do? He would get us out of memories. Why shouldn't we walk out the door, come back, and do it ourselves?"
— Andy Grove, Only the Paranoid Survive
Section 9
Analyst's Take
Faster Than Normal — Editorial View
The best post-mortems use counterfactuals explicitly. Instead of "we failed because of X," ask "what would have had to be different for us to succeed?" The counterfactual forces you to distinguish between necessary causes (we would have succeeded if we had had X) and sufficient causes (X alone would have been enough). Most failures have multiple causes. The counterfactual helps you identify which ones to address.
The trap: counterfactual drift. We imagine alternatives that are vivid and plausible but were never realistic. "If we had launched six months earlier" — but could we have? What would we have had to sacrifice? The counterfactual must be grounded in what was actually knowable and feasible at the time. Otherwise, it is a story, not a lesson.
Section 10
Test Yourself
Is this counterfactual reasoning done well?
Scenario 1
A team failed to hit its revenue target. In the post-mortem, someone says: 'We would have hit the target if we had hired two more salespeople.'
Scenario 2
An investor evaluates a startup: 'We passed on this deal because the market was too small. But what if the market had been larger? Would we have invested?'
Scenario 3
A CEO attributes a successful product launch to 'our great team.' The board asks: 'Would the launch have succeeded with a different team?'
Pearl formalises counterfactuals as the foundation of causal inference. His "do-calculus" provides a framework for reasoning about what would have happened under different interventions. Essential for understanding the logic of causation.
Grove's use of the "new CEO" counterfactual to decide Intel's pivot from memory to microprocessors is a classic example of counterfactual reasoning in strategy. The book shows how to use imagined alternatives to cut through emotional attachment and sunk cost.
Kahneman discusses counterfactual thinking, regret, and the psychology of "what if" scenarios. His work on reference points and loss aversion shows how counterfactuals shape our evaluation of outcomes.
The psychological literature on counterfactual thinking — how we generate "what if" scenarios, when they help learning, and when they produce regret or rumination. Foundational for understanding the biases in counterfactual reasoning.
A more accessible treatment of Pearl's causal framework. Pearl explains counterfactuals, the ladder of causation, and why "what if" reasoning is essential for understanding cause and effect.
Leads-to
Scenario Planning
Scenario planning is counterfactual reasoning applied to the future. Instead of "what would have happened if X had not occurred?" we ask "what will happen if X occurs?" Both use imagined alternatives to improve decision quality. Scenario planning is forward-looking counterfactual analysis.
Leads-to
Base Rate
Base rates provide the counterfactual for single-case reasoning. "What would we expect to happen in similar situations?" is a counterfactual. When we ignore base rates, we overconfidently attribute outcomes to the specific case. Base rates anchor our counterfactual to reality.
