Correlation Isn't Cause — And Confusing the Two Is Costing You
On distinguishing nuanced situations so you can understand what's actually driving an outcome — and respond to the right thing.
When something goes wrong, there is almost always an obvious explanation nearby. Like:
The launch underperformed
marketing didn't do enough
The team lost momentum
the project manager dropped the ball
The client churned
the product wasn't good enough
These explanations feel like causes. They're almost always symptoms. And treating a symptom as a cause produces an intervention that makes you feel like you're doing something while leaving the actual driver intact.
The most expensive mistake you can make is solving the wrong problem with complete conviction.
The Signal is the column for slowing down enough to ask: what is actually driving this? Not what it looks like from the outside. Not the most recent visible event. What is the structural or relational dynamic that would keep producing this result even if you changed everything on the surface?
A few distinctions that tend to clarify quickly:
Correlation vs. cause: two things happening together does not mean one caused the other. The question to ask is — if I changed only this variable, would the outcome actually shift?
Proximate vs. distal cause: the most recent event in a chain is usually not the most important one. What happened three steps earlier that made this event almost inevitable?
Person vs. system: when the same problem keeps appearing across different people or teams, the problem is almost certainly structural, not personal. Replacing individuals without changing the system produces the same result with new faces.
The work here is resisting the pull toward the obvious explanation. The obvious one is usually fast, feels satisfying, and is often wrong. The real cause tends to be quieter, less emotionally compelling, and harder to see — which is exactly why it keeps operating.
The most expensive mistake you can make is solving the wrong problem with complete conviction.
The Signal is the column for slowing down enough to ask: what is actually driving this? Not what it looks like from the outside. Not the most recent visible event. What is the structural or relational dynamic that would keep producing this result even if you changed everything on the surface?
A few distinctions that tend to clarify quickly:
Correlation vs. cause: two things happening together does not mean one caused the other. The question to ask is — if I changed only this variable, would the outcome actually shift?
Proximate vs. distal cause: the most recent event in a chain is usually not the most important one. What happened three steps earlier that made this event almost inevitable?
Person vs. system: when the same problem keeps appearing across different people or teams, the problem is almost certainly structural, not personal. Replacing individuals without changing the system produces the same result with new faces.
The work here is resisting the pull toward the obvious explanation. The obvious one is usually fast, feels satisfying, and is often wrong. The real cause tends to be quieter, less emotionally compelling, and harder to see — which is exactly why it keeps operating.
The Manipulation Index maps 27 of these patterns — what they look like, why they work, and what changes when you can name them.
Free download when you subscribe.