Pivot When Reality Disproves the Assumption, Not When Work Gets Hard
Alex Hormozi argues that setbacks are useful only if they are read at the right level: a failed result may mean a founding assumption has been disproven, or it may only expose weak execution. In a discussion with Chris Williamson, he frames the decision to pivot or persist as a question of whether reality has invalidated the premise behind the effort, rather than merely made the effort uncomfortable. The broader case is that people misread reality when they mistake labels for explanations, leave expectations undefined, or generalize too much from early losses.
Bad feedback is not always a reason to change course
Bad feelings and bad results are signals that something needs to be updated. They are not, by themselves, proof that the whole path is wrong. Alex Hormozi draws the practical line as a distinction between “push” and “pivot”: when to endure difficulty, and when to change the plan.
Chris Williamson presses the tension directly. If feeling bad is useful feedback, how do you tell the difference between discomfort that should be endured and discomfort that should change your approach? Williamson gives the gym version of the same problem: are you ending a set because you are being weak, or because you are about to injure yourself?
Hormozi’s answer is that pivoting makes sense when feedback has disproven one of the fundamental assumptions that justified the project in the first place.
If one of the fundamental assumptions that you began your quest with has been proven untrue based on the feedback, then that is where pivoting makes sense.
His example is deliberately simple: a “doggie skateboard” business built on the belief that dog owners will want to buy skateboards for their dogs. If the ambition depends on a large enough market and a high enough take rate, and after talking to 100 dog owners none of them want the product, Hormozi would not treat that as a “push” situation. The original premise has failed.
But if the response is not a rejection of the underlying desire, and instead people are confused about what the product is or how it is being presented — “I don’t know what you have in your hands” — then the assumption has not necessarily been disproven. The problem may be execution. In that case, Hormozi says, the lesson may be to get better and push through.
The practical test is whether reality has disproved the premise, or merely exposed a weak presentation, skill, or behavior. The danger is not that failure teaches nothing. Hormozi’s point is the opposite: “teaching will occur.” The risk is learning the wrong generalization.
Losing teaches you shit. And we just need to make sure that we learn the right thing.
His example is hiring one bad first employee and concluding, “therefore all employees suck.” The loss produced a lesson, but not necessarily a true one.
Early mistakes can become reinforced defensive models
Chris Williamson draws out a second implication: because bad lessons can become embedded, early decisions matter disproportionately. If the first three employees betray or disappoint you, you may build defensive operating habits around that experience. When a competent fourth employee arrives, those compensatory mechanisms remain: hyper-vigilance, restriction, mistrust, slower progress, a worse place to work.
Alex Hormozi agrees and puts it in behaviorist terms: changing a behavior with a long history of reinforcement is harder than changing one with no history of reinforcement. The point is not just that first experiences linger. It is that they can train a model of the world that later distorts judgment, even when the current situation no longer matches the conditions that produced the model.
Williamson calls him “Skinner-pilled”; Hormozi says behaviorism is “the only thing that’s made sense” to him. But the important claim is not the label. Hormozi says he has been reinforced for thinking in inputs and outputs because doing so has improved his predictions.
When people do not have what they want, he says, either the model through which they view the world is incorrect, or the model is directionally right but contains wrong or insufficient variables. The recurring problem, in his account, is being “constantly surprised by reality”: the business did not respond as expected, the conversation landed worse than expected, the life you built is not the one you wanted. For Hormozi, that surprise is evidence that the explanatory model needs work.
Description is not explanation
Alex Hormozi’s central distinction is between describing a pattern and explaining it. He uses the sentence “Johnny stole because he’s dishonest” to show the problem. At first, the sentence sounds explanatory. But “dishonest” is only a label applied to a cluster of behaviors, one of which is stealing. Restated plainly, the sentence becomes: “Johnny stole because he’s the type of person who steals.” That is circular.
A better explanation, in Hormozi’s model, is that Johnny has been reinforced for stealing in the past, or saw someone else get reinforced for stealing and modeled that behavior. Chris Williamson adds the companion possibility: Johnny may have been rewarded for stealing, or punished for doing the opposite.
Hormozi treats this as more than a semantic issue. He says many people use words to explain reality without understanding what those words actually refer to. They mistake labels for causes. That makes reality harder to predict.
The same distinction applies to competence. Telling a child to “get good at basketball” is too abstract to act on. Even breaking the game into dribbling, passing, and shooting may still be too vague. Hormozi says instruction becomes useful only when it resolves into observable behavior: step with the left foot toward the person, extend the elbows, finish with the thumbs down; if the ball reaches the target in the direction they are running and they catch it, the pass succeeded.
From there, in Hormozi’s behavioral unbundling, the abstraction can be rebuilt. Repeating that chain of behaviors successfully becomes “passing.” Passing, dribbling, and shooting eventually bundle into the description “good basketball player.”
Undefined words create avoidable conflict
Alex Hormozi applies the same unbundling to relationships and workplace conflict. A couple can fight indefinitely over whether one partner “loves” the other because each defines love through different behaviors. One person may point to paying bills and taking out the trash. The other may mean compliments, hugs, listening, and asking about their day.
The conflict persists because both sides are using the same word while attaching it to different observable actions. Hormozi’s proposed move is blunt: ask what the word means in behavior.
He gives an example from Acquisition.com, where two employees were arguing. One said they would be fine if the other person were “kind and polite.” Hormozi asked what that meant. After some hesitation, the request became concrete: ask more questions rather than making more statements. The other person agreed, and Hormozi says the issue resolved.
Chris Williamson connects this to a Neil Strauss line: “unspoken expectations are premeditated resentments.” He adds that the resentment may not be deliberate; people often create it unconsciously by expecting behavior they never articulated.
Williamson extends the same model to complaints. A complaint, in his formulation, often reveals a mismatch between someone’s expectation and what reality delivered. “Why is there all of this traffic on the way to work?” really means: I assumed there would not be traffic, and reality disagreed.
Reality, Williamson says, does not care. It will keep delivering what it delivers. Hormozi puts it more tersely: “It’s undefeated.”