Every bottle we sell started as a sample vial in a small padded envelope. Before that, it was a conversation with a broker or a distillery. Before that, it was years of oak and spirit doing something in the dark that nobody could fully predict. By the time it reaches us, we are the last filter before it reaches you. We take that seriously.
We go through somewhere between thirty and forty samples before we find one worth buying. That ratio is not a complaint. It is the job.
Where the samples come from
We work with brokers and distilleries directly. The relationships matter. Good brokers know what we are looking for, which means the samples we receive are already a pre-selection, already filtered against a general profile. Even so, most of them do not make it through.
When a sample arrives, we know two things: the distillery and the cask number. That is it. No age statement, no suggested retail price, no tasting notes from the seller. We want to form our own opinion before we know anything else about what we are tasting, because the moment you know a cask is twenty years old and priced accordingly, your palate starts to cooperate with the price tag rather than the liquid.
Three people, three scorecards, no conferring
There are three of us making these decisions. Each of us tastes every sample independently and fills out the same spreadsheet: distillery, cask number, nose, palate, finish, and a score from zero to one hundred. We do not share notes until everyone has finished. We do not see each other's scores until all three are in.
The comparison matters as much as the scores. We are not just looking for high numbers. We are looking for consistency. If one of us describes a cask as earthy with a hint of beets, another writes earthy and dry, and a third says it reminds them of an old baseball mitt, those three descriptions are telling the same story. That alignment, even across different vocabulary, is a signal that the cask has a genuine, legible character worth putting a label on.
We are not looking for the easy drinking whisky. We are looking for something that will make you question everything and nothing at the same time.
The rule: 93 or above, from everyone
We have one hard rule. To buy a cask, every one of the three founders needs to score it 93 or above. Not an average. Not a majority. Everyone.
A 96, a 94, and an 88 does not clear the bar. That 88 is a vote against, and we respect it. The person who scored it 88 tasted something the other two did not notice, or weighted something differently, or simply could not get behind the finish in the way the others could. That disagreement is information. We do not override it.
Price enters the conversation only after the scores are in. If a cask clears 93 across the board and the numbers work, we move forward. If it clears 93 and the price does not work, we let it go. There will be another sample. There always is.
What kills a cask immediately
There are no bad whiskies. We genuinely believe that. But there are casks that are not right for what we do, and the fastest way to identify them is the finish.
A funky finish is a no. Not funky in the interesting sense, where a cask surprises you with something unusual that grows on you. Funky in the sense that it lingers in a way you did not ask for, that leaves you reaching for water, that makes the experience end on a note that undermines everything that came before it. Nobody opens a bottle of whisky hoping to be left with a bad taste. We filter for that before the bottle ever gets made.
What we are actually looking for
Character. Specifically, the kind of character that rewards attention. A whisky that opens up the more you sit with it, that reveals something different on the third sip than it did on the first, that gives you something to think about without making you work for it.
We are not chasing approachability for its own sake. The bottles we have bought and released share a common quality: they have a point of view. They are not trying to be everything to everyone. They are trying to be themselves, fully and without apology. That is what thirty to forty rejected samples is in service of. Finding the one that has something real to say.