Taste is often dismissed because people confuse it with preference.
Preference says, “I like this.” Taste says, “This works here, for this customer, under these constraints, against this standard, for these reasons.”
That distinction is not semantic. It is the difference between private opinion and operational judgment.
A founder who says, “I hate this landing page,” may be expressing preference. A founder who says, “This landing page uses our internal category language instead of the buyer’s pain language; it asks for a demo before earning trust; and the proof points do not reduce the perceived switching risk,” is exercising taste.
A manager who says, “This update feels weak,” may be expressing a vibe. A manager who says, “This update reports activity but not decisions, hides the unresolved dependency, and gives no confidence level on the date,” is making quality legible.
A technical leader who says, “This architecture is ugly,” may be protecting a preference. A technical leader who says, “This creates a data ownership ambiguity, makes the next integration harder, and gives us no rollback path,” is naming consequences.
Preference can be arbitrary. Taste can be argued.
This matters because companies run on decisions, not moods. If taste is treated as preference, it becomes political. Senior people win because they are senior. The team learns to optimize for pleasing the reviewer. Work becomes a guessing game. People stop asking, “What is true about the problem?” and start asking, “What will get approved?”
That is how taste becomes elitist. Not because high standards are elitist, but because hidden standards are.
Real taste is accountable to context. It has to explain itself. It has to connect the artifact to the customer, strategy, risk, system, and constraint. It can still be debated. In fact, it should be debated. The point is not that the person with taste is always right. The point is that the judgment has substance.
A useful test: if the reasoning cannot survive being written down, it may be preference wearing a better jacket.
This is especially important in creative and product work, where language gets slippery. “Premium,” “simple,” “clear,” “sharp,” “delightful,” “enterprise-grade,” “high quality,” “on brand,” and “strategic” can all hide weak thinking. They sound like standards, but they are not standards until people can point to examples, tradeoffs, and consequences.
For operators, taste has to earn its authority.
It earns authority by being specific. What exactly is off? What standard is being violated? What customer reality is being missed? What risk is being ignored? What downstream cost will this create? What stronger example shows the bar? What compromise would be acceptable?
This is why taste is learnable. If taste were only preference, it could not be taught. You could only absorb someone’s likes and dislikes. But if taste is consequence-aware judgment, it can be trained through exposure, comparison, critique, and aftermath.
Consider hiring. Preference likes a candidate’s style. Taste notices whether the candidate’s strengths match the actual job. The charismatic candidate may be wrong for a role that needs patient operating discipline. The polished executive may be wrong for a messy build phase. The brilliant specialist may be wrong if the job is mostly cross-functional influence. Taste does not ask, “Do I like this person?” It asks, “What will happen if this person is inside this system?”
Consider strategy. Preference likes a bold narrative. Taste asks whether the strategy names the tradeoff. A strategy that says yes to everything is not strategic; it is a decorated backlog. Taste looks for the cut. What are we not doing? What customer are we not serving? What operational promise are we refusing to make? What metric could make this look good while the business gets worse?
A leadership team with preference debates whether the plan feels ambitious. A leadership team with taste asks whether the plan can survive Monday morning: who owns the decision spine, which functions must change behavior, what constraint will break first, and which attractive initiative should be killed so the real priority has oxygen.
Consider AI-assisted work. Preference likes the fluent answer. Taste asks whether the answer is grounded, useful, and safe to trust. It notices generic phrasing, missing source trails, unsupported confidence, and the strange smoothness of a draft that has not earned its conclusions.
The practical problem is that preference often feels faster. It is easier to say, “I like this one.” It is harder to explain the mechanism. But the cost of not explaining accumulates. The team repeats the same mistakes. Review cycles stretch. Leaders become bottlenecks. Quality depends on who happens to be in the room.
Taste becomes scalable when it produces shared language.
Instead of “make it better,” say “make the tradeoff visible.” Instead of “more premium,” say “reduce the buyer’s perceived risk.” Instead of “cleaner,” say “remove choices that are not decision-relevant.” Instead of “stronger candidate,” say “more evidence of operating through ambiguity.” Instead of “the AI output is bad,” say “the summary is plausible but not source-grounded enough for a customer claim.”
This is not pedantry. It is how judgment becomes transferable.
The best leaders are often picky, but not randomly picky. Their dissatisfaction points somewhere. They reject the work because it does not fit the promise, because it will create rework, because it hides a decision, because it lowers the bar, because it optimizes for the wrong audience, because it looks finished while being structurally weak.
Taste is the ability to be dissatisfied for the right reasons — and to explain why.
The explanation prevents taste from becoming ego. It also prevents teams from lowering the bar in the name of speed. Once the reason is visible, people can improve against it. They can disagree with it. They can refine it. They can turn it into a standard.
That is the operating progression: taste sees, standards protect, craft executes.
Preference stops at personal reaction. Taste moves from reaction to reason. And in a serious organization, reason is what lets quality scale.
