Food, cooking & wine · 5 min read + interactive

The wine was already good. The label was getting in the way.

Stag’s Leap is one of my favourite Napa Cabernets. I knew I liked the wine before I knew that one of its bottles had taken part in perhaps the most consequential blind tasting in modern wine history.

In 1976, California wine was not considered an equal to the great wines of France. Cabernet Sauvignon was already central to prestigious Bordeaux blends, but Napa was still the outsider—interesting, perhaps, but not world-class.

Then someone covered the labels.

British wine merchant Steven Spurrier organized a tasting in Paris that placed California wines beside celebrated French bottles. The judges were respected members of the French wine establishment, and the wines were served blind.

Among the reds were renowned Bordeaux wines and six California Cabernet Sauvignons. When the scores were counted, the winner was the 1973 S.L.V. Cabernet Sauvignon from Stag’s Leap Wine Cellars.

The wine hadn’t changed during the tasting. Its reputation had.

The event became known as the Judgment of Paris, and it helped establish Napa Valley as a serious wine region. But I also think it contains a useful lesson about how we evaluate products.

A label can become part of the experience

Imagine tasting two glasses of wine while looking at the bottles.

One comes from a famous French château. The other comes from a young California winery most people in the room have never heard of.

Even before the first sip, the labels have supplied information: history, geography, price, status and expectation. Those signals may be relevant, but they can also shape what we believe we are experiencing.

Now hide the bottles.

The wine stays exactly the same, but part of the decision environment disappears.

That was the disruptive part of the Paris tasting. It did not make California wine better. It created a setting in which the judges could encounter its quality without first filtering it through the existing hierarchy.

Was it an A/B test?

Not exactly.

It was a blind comparative tasting, not a modern randomized product experiment. There was a small panel of judges, a limited selection of wines and one afternoon of evaluation. There was no representative sample of wine drinkers, and the result did not prove that every Napa Cabernet was better than every Bordeaux.

But the tasting shares some important ideas with a well-designed A/B test:

The real lesson is not that blind tests are automatically fair. It is that a useful experiment must separate the thing being evaluated from everything else that could influence the result.

In product work, those hidden influences can be surprisingly ordinary.

A new experience might be shown to more engaged customers. One version may load faster. A redesigned feature may receive more prominent placement. A team might change the message and the interaction at the same time, then attribute the result to the interaction alone.

Like a visible wine label, context can quietly become part of the treatment.

Test the product—not the reputation around it

A strong experiment starts by asking what must remain consistent for the comparison to mean anything.

• Are both versions shown to comparable people?
• Is eligibility defined the same way?
• Are timing and placement consistent?
• Is the success metric chosen before the results arrive?
• Could familiarity or brand recognition affect behaviour?
• What guardrails would reveal a win that damages something else?

The Judgment of Paris also demonstrates the value of a credible benchmark.

California wine did not compete against an easier standard created for newcomers. It was placed beside the wines considered the best in the world and evaluated by people who respected that standard.

That made the result harder to dismiss.

For a challenger product, this can be more powerful than claiming to be innovative. Choose a meaningful incumbent, create a fair comparison and let the experience produce the evidence.

A surprising result should widen the question

It is tempting to turn an unexpected experiment into a sweeping conclusion:

California Cabernet beat Bordeaux. Therefore, California Cabernet is better.

That would claim more than the tasting established.

The useful conclusion was narrower and more important:

An excellent wine did not need a famous French label to be excellent.

That result created permission to look again. Buyers, collectors and critics began taking California more seriously. Other emerging wine regions could also imagine competing on quality rather than inherited status.

A good experiment often works in the same way. It doesn’t finish the conversation. It shows that an assumption is no longer safe and gives us a reason to investigate further.

The 1973 Stag’s Leap Cabernet was already good before it arrived in Paris.

The blind tasting simply created a moment in which the label could no longer speak first.