Essay · Executive Depth

The Choice Behind the Choice

Autumn 2026 · Reading time · 16 minutes

Every decision an owner defends in public was chosen twice. Once by the situation and once by the self-image she was quietly protecting. Telling the two apart is the whole game.

An owner spends her career defending decisions. She defends them to her team, her investors, her family, her lawyers, herself. What she rarely gets asked, and rarely asks herself, is which of those decisions were actually made by the situation and which of them were made by the version of herself she was trying to remain. The two look identical from the outside. Both arrive with reasons. Both come with a story about why this was the right call given the constraints. Both survive scrutiny for a while. Only one of them ages well. The other is the source of almost every quiet regret a serious operator carries by year ten.

The decision the situation asked for

A decision made from the situation is boring to describe. The owner looked at what was in front of her, considered the parties involved, weighed the reversible against the irreversible, made the move that best fit the constraints and moved on. There is no biography inside that decision. There is only the situation and the response. Decisions of this kind tend to be small enough that no one writes about them and consequential enough that the company would look very different without them. They accumulate quietly and they are the ballast of a healthy business.

An owner running mostly situation-based decisions is usually described as level-headed, unshowy, occasionally boring. Her meetings do not produce many quotable moments. Her calendar contains an unusual number of second and third conversations with the same person about the same topic, because she has not felt the need to force premature resolution to satisfy her own self-image as a decisive person. Her team, over years, comes to trust that the decision she announces on Thursday is not going to be quietly reversed on Monday, because it was made in response to the actual situation rather than in response to her mood about the situation.

The decision the self-image asked for

A decision made from self-image is more interesting to describe and more expensive to live with. Here the owner is not primarily responding to the situation. She is responding to a picture of herself she is quietly determined to keep intact. The picture might be that she is the tough founder who never backs down, or the generous boss who never disappoints, or the visionary who never hedges, or the professional who never gets emotional. The situation gets used as raw material for a scene the picture requires. The decision that emerges is technically about the business and functionally about the self.

These decisions are recognizable, in hindsight, by a specific texture. They are unusually emphatic. They tend to be announced with more finality than the situation required. They often produce a small internal glow in the owner after the announcement, a sense that she was, at that moment, the person she wanted to be seen as. That glow is diagnostic. A decision that made you feel more like yourself is not necessarily a bad decision but it is a suspicious one and worth revisiting when the glow has worn off.

The question I ask myself the morning after

The question I ask myself the morning after any decision I felt slightly too satisfied by is straightforward. Would I have made the same call if no one had been watching? The witnesses do not have to be many. They can be a single colleague whose respect I care about, a partner whose approval I want, an internal audience of one whose picture of me I do not want to disturb. When the answer is yes, the decision was situation-made. When the answer is no, or when the answer is a hesitant maybe, the decision was self-image made and I need to revisit it before it hardens into a policy the whole company inherits.

The point of the question is not to abolish self-image. Self-image is not the enemy of good judgment. It is a stabilizer that allows an owner to act coherently across thousands of small moments without renegotiating her identity every time. The point is to keep the stabilizer in the passenger seat and the situation in the driver's seat. When the arrangement flips, the company begins to serve the owner's picture of herself instead of the market it exists to serve. That inversion is subtle at first and disqualifying by year five.

Why this matters more now

The reason the distinction has become urgent is that owners now have access to a form of decision support that will happily flatter either mode. Ask an AI model to help you think through a business decision and it will produce a competent-looking analysis in either register. Ask it to help you defend a decision you have already made, and it will produce a competent-looking defense in either register. The model has no view on which of your two selves is driving the ask. It optimizes for whichever picture you have already loaded into the prompt and the ease of that optimization is precisely what makes self-image decisions so much cheaper to manufacture than they used to be.

Five years ago, defending a self-image decision required the owner to construct the defense in her own words, which produced enough friction to sometimes wake her up mid-defense. Today the defense arrives fully formed in twelve seconds. The friction that used to serve as a diagnostic has been engineered out of the loop. An owner who wants to keep her judgment honest in this environment has to reintroduce the friction on purpose. Otherwise, every self-image decision she is inclined to make comes prewrapped in a rationale that sounds like it was written by someone thoughtful, because it was, only that someone was not her.

The self-image that hides best in a crisis

A self-image that is nearly impossible to catch in the ordinary week becomes fully visible in a crisis, and this is one of the underappreciated uses of a hard situation. When a company is under real pressure, the picture the owner has been quietly protecting stops being a background influence and becomes a foreground actor. She reaches for it because the pressure has stripped away the smaller decisions that used to let her hide it. Every choice under stress is amplified, and the amplification is diagnostic in a way that a normal quarter never is.

An owner who wants to understand which self-image is driving her should not look at what she did last week. She should look at what she did during her last real crisis. The pattern that emerged there is the pattern that has been quietly running the smaller decisions all along. Once she can name it, she can start to work around it deliberately in ordinary weeks. Until she can name it, she will keep making the same shape of decision under normal conditions without noticing that the shape has an author and the author is not the situation.

The forty-minute rule I use

The rule I use, for any decision above a certain size, is that the first forty minutes of my thinking happen without a model in the room. Pen, paper, questions. I try to state the decision in one sentence, then the situation in three sentences, then the party who will bear the consequence in one sentence. Only after I have done that do I open a chat window and let a model interrogate what I wrote. The model, at that stage, is very useful. It catches my sloppy reasoning, it surfaces angles I missed, it stress-tests the parts I moved through too fast. What it does not do, because I did not let it, is choose the frame. The frame is mine.

The forty minutes are not a productivity habit. They are a discipline for keeping the choice behind the choice visible. The moment I let a model set the frame at minute one, I lose the ability to notice which version of me is answering the questions inside the frame. By minute five, the answers will already be shaped by whichever picture of myself I unconsciously loaded into the prompt and by minute twenty I will be defending that picture with a confidence I did not earn. The forty minutes buy me the honesty I need to catch myself before the confidence sets.

The tell in the announcement

The reliable tell of a self-image decision, once you know what to look for, is in the announcement rather than in the decision itself. A situation-made decision tends to be announced plainly. Here is what we are doing, here is the reason, here is what changes. A self-image decision tends to be announced with unnecessary architecture around it. There is a preamble. There is an appeal to values. There is a moment of scripted reflection. The extra architecture is not there to communicate the decision. It is there to frame the decision-maker and the framing is what gives the game away.

An owner who wants to catch herself in the act can watch her own announcements for this signature. Not the content, the shape. If the announcement is doing more work on the self-portrait than on the decision, the decision is downstream of the self-portrait and worth revisiting before it hardens into policy. The rule is not that architected announcements are always wrong. Some situations genuinely need them. The rule is that architecture is diagnostic and any pattern of it should be treated as a signal about the state of the announcer, not only as a stylistic choice.

The three self-images most often in the driver's seat

Three self-images are, in my experience, the most common drivers of otherwise inexplicable owner decisions. The first is the picture of the owner as the person who never flinches. Under her, hard calls get made a week earlier than the situation asked for, because delaying would risk being seen as tentative. Employees are let go before the case has fully formed. Partnerships are terminated before the relationship has been given its last honest conversation. The decisions look decisive and feel good in the moment. Most of them, on review, turn out to have been about the picture rather than about the situation.

The second is the picture of the owner as the person who is always generous. Under her, difficult conversations get softened past the point of communication, expensive accommodations get granted long past their useful life and the company's stated values get diluted one small exception at a time. The generosity feels correct at every individual decision point. It aggregates, over years, into a company whose real values are unspoken and whose stated ones are quietly known to be optional.

The third is the picture of the owner as the person who is always ahead. Under her, strategic pivots arrive slightly before the market has confirmed the need for them, tools get adopted slightly before the team has been trained to use them and the company runs a small permanent ambient chaos that everyone attributes to the industry when its actual source is one person's need to be visibly a step ahead. Each of the three self-images produces genuine strengths in the years it earns its place. Each of them turns into a liability the moment it starts making decisions that belong to the situation.

The witness I invented to keep myself honest

One of the more useful mental habits I have developed is a small imaginary witness whose only job is to notice when I am about to make a decision because it looks good rather than because it works. The witness does not vote on the decision. She only notices. In practice, this means that before I commit publicly to a course of action, I spend a minute imagining the exact same course of action being taken by someone I neither like nor dislike and asking whether the reasoning still holds when I strip out the aesthetic satisfaction of being the one who made it.

Most of the time the reasoning holds and I move forward. Occasionally the reasoning collapses under this small transfer and I catch a decision that would have cost me a year to unwind. The witness is not moral judgment. It is a professional instrument. It exists because I know, from experience, that my most confident decisions are the ones most likely to be doing double duty for a picture of myself I have not audited and the witness is what forces the audit while it is still cheap.

What owners do when they finally see it

The first time an owner catches herself making a self-image decision cleanly, in real time, is disorienting. She notices the small glow arriving before the situation has finished speaking. She notices that the decision she is about to announce is unusually emphatic. She notices that she is more concerned with how the announcement will land than with what the announcement is. That noticing, the first time, is uncomfortable enough that most owners look away from it. The ones who look toward it discover something that changes the trajectory of the next ten years of their practice.

What they discover is that most of their most consequential decisions have been made twice. Once by the situation, which produced a reasonable answer they were about to accept. Once by the self-image, which overrode the reasonable answer with a more flattering one and packaged the override as decisiveness. Naming the pattern does not eliminate it. Nothing eliminates it. But the naming shifts the ratio over time and the shift shows up, five years later, as a company that grew in a direction the owner is proud of because the direction was chosen by the market rather than by her mirror.

The small mercy of the second decision

One of the small mercies of a practice like this is that almost every self-image decision is followed, within a week or two, by an opportunity to make a second decision on the same topic. The situation rarely holds still. A colleague you let go too quickly comes back around in the form of a related decision about another colleague. A promise you made too grandly is tested by a smaller moment that lets you either recommit or quietly downscale. The second decision is the one that most owners have available and most of them do not use, because using it requires admitting that the first one was not clean.

The choice behind this essay

There is a version of this essay I could have written that would have made me look wiser than I am. It would have organized these ideas around a single elegant framework, offered a small model to memorize and closed with a stirring line about the courage of self-knowledge. It would have been read and shared and remembered. It also would have been an example of the pattern the essay is describing. The essay you have just read is longer, less symmetrical and less quotable on purpose, because the harder move here was to describe the practice honestly, including the parts where I have failed at it, rather than to package it in a shape that would have flattered the picture of myself I might otherwise have been protecting.