Scoring Taste: Can You Measure the Unmeasurable?

There is a conversation I have had dozens of times, in different forms, with different people. It goes something like this: I describe Taste OS, a system that scores brands on five dimensions of taste, each worth twenty points, for a total score out of one hundred. Or I describe The Relevance Index, which tracks over twelve hundred brands on cultural relevance using a hybrid of real data and AI judgment. And the response is always some version of the same objection: "You can't score taste. It's subjective. You're trying to measure the unmeasurable."

They are right. And they are also completely wrong.

Why you would try

The instinct to quantify taste sounds absurd if you think of taste as purely subjective — a matter of personal preference with no external referent. If taste is just "I like what I like," then scoring it is as pointless as scoring someone's favourite colour.

But taste is not just preference. Taste, in the professional sense, is a form of judgment that can be evaluated against outcomes. A brand strategist with good taste makes decisions that lead to stronger brands. A product designer with good taste makes products people want to use. A creative director with good taste produces work that resonates with audiences. The taste is not random. It correlates with results.

If taste correlates with outcomes, then it can — in principle — be measured. Not perfectly. Not objectively in the way you can measure revenue or page views. But measured in the sense that you can build a framework that captures some of its dimensions and makes them visible, comparable, and debatable.

That last word is the key one. Debatable. The purpose of scoring taste is not to arrive at a definitive number. It is to create a structured conversation about quality.

The Taste OS approach

Taste OS scores brands across five dimensions, each worth twenty points. The dimensions are: Vision (clarity of purpose and point of view), Craft (quality of execution and attention to detail), Cultural Resonance (relevance and connection to the cultural moment), Coherence (consistency across all touchpoints), and Courage (willingness to make bold, distinctive choices).

Twenty brands are pre-scored in the system. You can also score any brand yourself using the interactive scorer. The scoring is deliberately subjective — it is your judgment, expressed through a structured framework.

The interesting thing about Taste OS is what happens when people use it. They almost always disagree with at least one dimension. Someone will score Apple highly on Vision and Craft but argue that Coherence is too high because the recent product line has lost focus. Someone else will score Supreme highly on Courage but question its Cultural Resonance now that it has been acquired by a larger corporation.

These disagreements are not bugs. They are the entire point. The framework's value is not in the scores it produces but in the conversations it provokes. By breaking "taste" into five specific dimensions, it gives people a language for arguments they were already having intuitively. It makes the implicit explicit.

The Relevance Index approach

The Relevance Index takes a different approach. Instead of subjective scoring, it uses a hybrid model that combines real data with AI judgment to score over twelve hundred brands on cultural relevance.

The system pulls Wikipedia page views (a proxy for public interest), Reddit mentions (a proxy for community discussion), and then sends each brand through multiple independent AI calls that assess five domains: Attention (are people paying attention?), Conversation (are people talking about it?), Creation (is the brand creating things worth noticing?), Desire (do people want to be associated with it?), and Zeitgeist (does it feel of the moment?).

The scores update weekly. The data is real. The AI assessments are calibrated against the data to avoid pure hallucination. And the result is a ranked list of over twelve hundred brands that, week to week, tells a story about which brands are culturally ascending and which are fading.

Is it perfect? No. Is it useful? Very. The Relevance Index catches things that pure data misses (a brand that is culturally relevant despite low search volume) and things that pure intuition misses (a brand that feels omnipresent but is actually mentioned less than you would expect). The hybrid model — data plus judgment — captures something that neither could capture alone.

The tension is the feature

What I have learned from building both systems is that the tension between data and intuition is not a problem to be solved. It is the most valuable feature of any taste measurement system.

When The Relevance Index says a brand is more culturally relevant than you thought, it challenges your intuition. You have to ask: am I wrong, or is the model wrong? Sometimes the answer is that the model is picking up signals you missed — a collaboration, a viral moment, a shift in how people talk about the brand. Other times the answer is that the model is overweighting one signal and your intuition is right. Either way, you learn something.

When Taste OS asks you to score a brand on Courage and you hesitate — is this brand genuinely bold, or just loud? — that hesitation is productive. It forces you to distinguish between different kinds of visibility. Between courage and attention-seeking. Between genuine conviction and performance. The framework does not answer the question. It makes you ask a better one.

This is what the "you can't score taste" objection misses. The point of scoring taste is not to produce a definitive answer. It is to create a rigorous conversation. It is to take something that usually lives in vague assertions — "that brand has good taste" or "this feels relevant" — and give it enough structure that you can argue about it productively.

What the data cannot tell you

There are real limits to what quantification can capture about taste, and I want to be honest about them.

Data cannot tell you whether something is beautiful. It cannot tell you whether a brand's visual identity makes you feel something. It cannot tell you whether a product's user experience has that ineffable quality that separates "this works" from "this is delightful." These are the dimensions of taste that remain stubbornly, beautifully subjective.

Data also cannot capture taste that is ahead of its time. The most interesting taste decisions are the ones that seem wrong to everyone else — the brand that zigs when the industry zags, the product that strips out features when competitors are adding them, the creative work that breaks conventions that everyone assumed were sacred. These decisions look like mistakes in the data until enough time passes for the world to catch up.

And data cannot measure the absence of things. Some of the best taste decisions are about what not to do — the feature not added, the product not launched, the campaign not run. Restraint is invisible to data. You can measure what exists. You cannot measure what was wisely left unmade.

Taste is greater than data

I hold a strong view on this: taste is greater than data. Not because data is useless — it is incredibly useful, and I use it extensively. But because taste operates at a level that data cannot reach. Data tells you what happened. Taste tells you what should happen. Data is retrospective. Taste is predictive. Data measures the market as it is. Taste imagines the market as it could be.

The best creative decisions in history were not data-driven. They were taste-driven, validated by data after the fact. Nobody A/B tested the iPod. Nobody focus-grouped the Nike swoosh. Nobody ran sentiment analysis before launching Supreme. These were conviction decisions — bets placed on taste, by people who had spent years building the judgment to make those bets.

That does not mean data is irrelevant. It means data is an input to taste, not a replacement for it. The best strategists I know are data-informed and taste-driven. They use data to sharpen their intuition, not to override it. They use it to test their instincts, not to avoid having instincts.

The best framework makes you argue

If I had to distil everything I have learned from building taste measurement systems into one sentence, it would be this: the best framework is one that makes you argue with it.

A framework you agree with completely is too loose. It is confirming what you already believe. A framework you disagree with completely is too rigid. It is imposing a worldview that does not match reality. The sweet spot is a framework that mostly makes sense but occasionally produces a result that surprises you, challenges you, makes you reconsider.

Taste OS does this by forcing you to score dimensions separately. You might believe a brand has great taste overall, but when you have to assign specific scores to Vision, Craft, Cultural Resonance, Coherence, and Courage, you discover that your overall impression was being carried by one or two dimensions while the others were weaker than you thought.

The Relevance Index does this by combining data with AI in ways that occasionally contradict conventional wisdom. When the index says a legacy brand is more culturally relevant than a direct-to-consumer darling, the instinct is to dismiss it. But sitting with the discomfort — asking "could this be right?" — often reveals something genuine.

Can you measure taste? Not definitively. But you can build frameworks that make your taste visible, debatable, and improvable. And that is worth more than a perfect number.

The score is not the point. The argument is the point.