What Happens After the Bottle Is Sold?
A winery can sell you a bottle in June that you don’t open until Christmas.
You spend an afternoon somewhere beautiful. There’s a great tasting room, an incredible view, maybe lunch, and someone walking you through the wines and telling you the story. You buy three bottles and take them home.
Months later, most of what helped sell that bottle has disappeared. You’re no longer at the winery. The view is gone. The person who explained the wine isn’t sitting beside you. You may remember the story, or look up the score, but the experience that led to the purchase and the experience of drinking the wine now belong to two different moments.
Eventually, it’s just you and the wine.
It actually reminds me of the whole thing about trading cards. Yes, one person can have more than one hobby. Bear with me.
With sealed trading cards, you’re buying expectation and possibility. You know the range of things you might get, but the reveal comes later. Once you open the pack, there’s a shared system for understanding what happened. You know the card and its rarity. You can look up its value. There’s grading. Other collectors recognize the same system.
The reveal becomes legible.
Wine’s reveal is less tidy.
You open the bottle at home and maybe it’s fantastic. But what made this wine different from the other fantastic wines you’ve had? What are you going to remember about it?
That question led me to look at how wine is communicated.
I analyzed 20 long-form wine videos containing about 6.5 hours of speech. The sample included educators, reviewers, blind tastings, buying guides and general wine content. I used the transcripts to identify what the speakers were doing: explaining production, providing story and context, discussing buying and prestige, describing sensory attributes, judging the wine or helping viewers perceive something in the glass.
The sample is exploratory. It isn’t a census of wine communication, and the percentages shouldn’t be treated as industry-wide estimates. The full dataset, coding definitions and methodological caveats are available in the accompanying research report at the bottom of this post.
I used AI to process the transcripts because doing this volume of repetitive coding manually would have taken far longer. I checked the arithmetic, caught bad data and discarded runs that couldn’t be trusted. The research also forced me to change my original assumption.
I expected scores and prestige to take up a major portion of the results.
They accounted for only 4.7% of the coded speech.
That surprised me because scores and rankings still seem to carry so much power in retail. If they barely appear in these videos, why do they remain one of the most visible ways to sell wine in a shop?
Then again, how much time does it take to say, “This wine scored 100 points, won Double Gold and took Best in Show”? Prestige may occupy very little speaking time while doing a lot of work. A future pass should probably count how often it appears, not only how many seconds it occupies.
What occupied more time was the broader body of information surrounding the immediate experience of drinking: context and story, production, buying advice and external validation. I combined those categories to ask how much communication explains or positions a wine without necessarily translating what happens in the glass.
I didn’t set out to use that grouping. I created it after the prestige hypothesis failed, so I treat it as a pattern to investigate, not a final result.
In the median video, that combination accounted for 52.6% of the coded speech.
The categories inside that number aren’t interchangeable. Limestone soils are not the same thing as a 100-point score. A producer’s history is not the same thing as buying advice. The combined figure shows how much attention goes toward explaining the wine from the outside, but the individual categories still matter.
Traditional tasting descriptions accounted for 16.1% of the core sample. Overall judgments such as balanced, elegant, powerful, concentrated or long added another 10.8%.
The question is what kind of access that language gives the person drinking.
Someone with more exposure to wine may understand what limestone can do to a wine compared with clay or sand. But simply saying, “This wine is grown on well-drained volcanic soils over sandy-loam subsoils,” does very little to tell the general public what the wine will taste like.
The same goes for, “This wine was fermented in stainless steel, then matured for 18 months in a mix of new and used Austrian oak barriques, followed by another six months of bottle aging.”
We hear information like that all the time, presented as if it somehow tells us what the experience of drinking the wine will be.
I’m not saying that information isn’t interesting. I’m a nerd. I like knowing those things. I’m asking in addition to that, whether there’s an opportunity to guide perception and attention more than we do now.
Perception guidance accounted for about 4% of the total coded speech, but the median video contained none. It appeared most strongly in formal educational content, where the purpose was to help viewers look, smell, compare and assess. Outside those instructional formats, it was much less common.
One experience has stuck with me whenever I think about this.
More than ten years ago, my wife and I went to a coffee shop in Taiwan. I ordered a siphon coffee made with Yirgachefe beans they roasted themselves. The staff explained how the beans had been processed, but I can’t remember any of that now.
What I remember is how the person serving us walked us through the coffee. He even did a short breathing exercise with us and explained how the taste would change as the temperature dropped.
At the time, I remember thinking, “Was all of this necessary?” Part of it felt performative. (He did make a very audible deep breath during his demonstration.)
But more than ten years later, I still remember the coffee as acidic and floral, with some earthiness. I didn’t create a metaphor for it. I didn’t memorize the production details. I remember what I was guided to notice.
One coffee doesn’t prove anything, but it probably explains why I started trying to measure language describing how a wine moves or changes over time: acidity arriving first, tannin gripping through the middle, weight building, structure opening or the finish slowly releasing.
The coding method in my research couldn’t measure that reliably.
Put simply, a sentence can do two things at once. If someone says, “The wine has high acidity that builds and spreads across the palate,” the phrase “high acidity” is a conventional sensory description, while “builds and spreads” describes movement over time.
Because I gave each section of transcript only one primary category, that sentence might be filed under acidity and the movement would disappear from the count.
That means I can’t use this method to give a credible percentage for dynamic sensory language. I would need a different approach that allows the same sentence to contain both attributes and movement.
Even without that number, the question doesn’t go away. We talk about wine a lot. We explain it a lot. We describe it a lot.
Yet many wines end up sounding similar once they’re reduced to citrus, stone fruit, red fruit, black fruit, oak, acidity, tannin, complexity, balance and length.
Those descriptions can be accurate and useful. But do they help someone understand why this particular wine should remain distinct from all the other excellent wines described with much the same vocabulary?
The market may be behaving perfectly rationally.
A producer has wine to sell. Sell this vintage, make the next one and sell that too. Maybe hold some back for a library release—or, you know, let the bottles that didn’t sell age into Reserva and sell those too. Scores, distribution, packaging, storytelling, hospitality and education all support that work.
Most consumers aren’t walking into a shop asking for a particular movement of acidity and tannin. They want to know whether they’ll like the wine, whether it’s good and whether it’s worth the money.
If the bottle sells and they enjoy it, the system may be doing exactly what it was built to do.
But the industry also puts enormous effort into making everything around the wine memorable. Wineries invest in architecture, restaurants, tasting rooms, vineyard tours, events and service. Building that entire hospitality ecosystem is no small feat. When it works, people remember the visit and want to come back.
The return is also easier to see. A tasting room or restaurant brings people through the door. An event sells seats. It is much harder to measure the value of helping someone understand what the wine actually tastes like after they leave.
So I keep coming back to this: are we getting better at making the experience around wine memorable without getting better at making the wine itself memorable?
The more brutal question is whether making the wine itself memorable is even important to the business. Most of us would probably say yes. But when I look at where the industry puts its time, money and attention, I’m not sure the two are treated as equally important.
People remember the beautiful winery, the dinner and the view. They may even remember that the wine was delicious.
Ask them later which wine it was, what made it different or what it actually tasted like. That is harder.
Perception and memory are separate questions.
The research so far only examines how people communicate and guide perception in the moment. It doesn’t tell us whether one kind of communication makes a wine easier to remember. It doesn’t show whether a different approach improves enjoyment or helps sell another bottle.
The next step is to test whether communicating how a wine develops and holds together helps people distinguish it from another wine and recall it later.
Once the bottle is home and everything that helped sell it has disappeared, what helps the wine itself survive?
