For most of the history of running companies, the scarce thing was the answer. Finding out what the market wanted, what the numbers said, what the options were and what each would cost took time, people and money, and the founder who could get to a good answer faster had an edge. That scarcity is ending. Answers of a certain quality are now cheap and instant, available to every founder with the same tools, and the edge that came from having them has evaporated in the way that edges do when everyone has the same equipment. What has not become cheap, and is in fact becoming more valuable by the month, is judgment: the ability to look at a set of cheap answers and decide, and to own the decision afterwards.
This inversion catches founders off guard because it runs against the instinct that better tools should make the job easier. They make parts of it easier. They make the part that matters harder, because when the analysis is free the only thing left to distinguish one founder from another is what they do with it, and that is a skill most people have never deliberately trained. Judgment used to be exercised rarely, at the end of a long process of gathering. Now it is exercised constantly, on a stream of plausible options, and a founder who has not built the habit will find themselves either paralysed by choice or, worse, quietly letting the tool choose.
Stress-testing a decision before owning it
The first habit is to use the cheapness of answers against itself. When a decision matters, do not ask for one recommendation. Ask for several, from deliberately different angles: the cautious operator, the aggressive competitor, the customer who would be most annoyed, the investor who would ask the hardest question. Founders who lead technology companies have started describing this as running a council, a set of distinct perspectives that argue with each other before the founder decides. The point is not to average them. It is to see the decision from enough sides that the founder’s own view is formed rather than borrowed, and to find the objection that the first answer quietly skipped.
The perspectives should be genuinely different rather than variations on a theme, and the most useful one is almost always the least comfortable: the view that assumes the founder is wrong and asks what would have to be true for that to be the case. A council that only ever agrees with the founder is a mirror, and mirrors are expensive to consult when the decision is real.
This practice has a second effect that is easy to miss. It forces the founder to articulate what they actually think, because a council of perspectives only helps if there is someone at the head of the table with a view of their own. Reading five arguments and picking one is not judgment. Reading five arguments, noticing that none of them weighs the thing you care about most, and deciding accordingly, is. The tools supply the arguments. The founder supplies the weighting, and the weighting is the product.
The decision is not the outcome
The second habit is harder and more important: separating the quality of a decision from the quality of its outcome. Founders conflate the two constantly, judging a call as good because it worked and bad because it did not, and the conflation destroys the ability to learn. A sound decision can turn out badly because the world is uncertain. A reckless one can turn out well for the same reason. A founder who only ever looks at outcomes will learn to repeat their lucky recklessness and abandon their sound reasoning, which over a career is a reliable way to get worse.
Reading five arguments and picking one is not judgment. Reading five arguments, noticing that none of them weighs the thing you care about most, and deciding accordingly, is.
The discipline is to write the decision down at the moment it is made: what was known, what was assumed, what the options were, why this one. Then, when the outcome arrives, judge the reasoning against what was knowable at the time rather than against what happened. Performance psychologists who work with people whose careers depend on decisions under pressure teach exactly this separation, because it is the only way to improve the process rather than merely to feel good or bad about the result. It is uncomfortable, because it sometimes means admitting a success was luck. It is also the only way judgment compounds.
A daily review that compounds
Compounding needs a rhythm, and the rhythm that works is small and daily rather than large and annual. One sentence a day: the most important decision I made, and what I was thinking. It takes a minute, it feels pointless for the first month, and after a year it is a record of a founder’s own reasoning that no tool could produce, because the tool was never in the room when the reasoning happened. Read back, it shows patterns: the situations where the founder consistently decides well, the ones where they consistently do not, the kind of pressure that degrades their thinking. Those patterns are the raw material of better judgment, and they are invisible without the record.
The sentence should record the reasoning, not the feeling. I chose X because Y is a useful entry. I think this went well is not, because it records the outcome the founder was trying to separate the decision from.
Annual reviews have their place, but they are too far apart to catch the drift in how a person decides, and they rely on memory that is already reshaped by outcomes. The daily sentence catches the decision fresh, before the result has had a chance to edit the recollection of why it was made. Over years rather than quarters it becomes the founder’s own training data, which is a resource worth more than any amount of the borrowed kind.
What cannot be delegated
All of this points at the one thing that no tool takes off a founder’s plate, which is responsibility. The analysis can be delegated. The options can be generated. The arguments can be rehearsed by a council of perspectives that costs nothing to convene. What cannot be handed off is the act of choosing and the ownership of what follows, and that act is becoming the whole job. A founder who understands this stops competing on having answers, which is a race the tools have already won, and starts competing on the quality of their decisions, which is a race that has barely begun.
Judgment is trainable. It responds to deliberate practice the way any skill does, and the practices are not mysterious: seek the disagreement, separate the call from the result, write it down, read it back. What it does not respond to is being outsourced, and the temptation to outsource it grows with every improvement in the tools. The founders who resist that temptation will find that their judgment is the only thing about their company that cannot be copied, and that it was the product all along.