How 7 People Test 400 Million Pounds Of Starbucks Coffee A Year

At Starbucks’s Seattle headquarters, seven people taste 75% of the beans that the company buys and roasts each year. Here’s how they get it done–without unduly influencing fellow testers.

How 7 People Test 400 Million Pounds Of Starbucks Coffee A Year

After any brainstorming session, you’re left with a pile of ideas. Some are good. Some are bad. And it’s up to you and your colleagues to parse which is which. There’s just one problem: The climate of the room, set by a single zealot or Debby Downer, may guide the otherwise objective thinking of the team. So how do you make sure that you’re sidestepping corporate politics and magnetic personalities to get everyone’s unfettered opinion?


Starbucks has a fascinating solution: Keep your mouth shut, and simply move a glass just two inches left or right.

Image: Courtesy of Starbucks

Inside Starbucks’s corporate headquarters in Seattle, a team of just seven people–a diverse group of men and women ranging from their 20s to 50s–tastes endless cups of coffee. They’re responsible for testing 75% of the 500 million-plus pounds of beans that the company will buy and roast each year, which is a process that has no shortcut. It requires brewing and sipping carefully curated samples collected from every shipping container that arrives in U.S. ports.

“When I first came in the tasting room 10 years ago, I remember my colleague said, ‘We’re really happy you’re here. We know you like to talk. Just don’t say anything,’” explains Major Cohen, senior project manager of global coffee engagement at Starbucks, who has sat in on the process. “We don’t want to telegraph anything to anybody.”

The tasting room feels like a surprisingly humble home for the tongue of the global coffee giant, akin to middle school science lab without the beakers, quietly set off to one end of a cubicle-laden floor in Starbucks HQ. The space is dominated by four long stainless steel tables. Around the periphery, you can spot restaurant-grade burners for boiling massive tea kettles you need two gloved hands to carry, and a hodgepodge of the best coffee grinders money can buy. In a room off the side, an industrial coffee roaster toasts a test batch of beans at a time. (Despite all of the coffee, this well-ventilated room smells less like coffee than your average Starbucks store during rush hour.)

After filling hundreds of juice glasses with carefully measured grounds and almost-boiling water, tasters form a line with spoons in-hand, smelling each cup, then slurping, swishing, and spitting a spoonful of coffee for each with mechanized efficiency. They’re confirming whether the beans will end their life as house blend, espresso, or even a Starbucks reserve. For the most part, these decisions have been predetermined by global operations and earlier tastings in Switzerland. But in this room, on this day, the team is the last line of defense.

Photo by Mark Wilson

They’re tasting with clear parameters for outliers–is something exceptionally good or bad? Should a bean be pulled from rotation or elevated to a reserve status? Might a particular shipment improve by being aged not three but five years? Does something just appear mislabeled?

And when these outliers are tasted, no one says a thing. Instead, a taster moves one glass in an endless line of glasses just a few inches to the right or to the left. They flag it, in essence, to tell the others “give this one special consideration.” It’s somehow notable.

Only once all of the tasters have tasted (and maybe retasted) that glass, forming their own, unfettered opinion, does a discussion begin. “Eventually, someone will go out on a limb,” Cohen explains. “They’ll say, the reason I put that there–did you notice?” And by that point, the room will have either noticed something special or not–never having the seeds of good or bad planted too early.

Now most of us don’t sell coffee for a living, but the Starbucks method could be equally applicable in a board room as a tasting room. So often, we need to evaluate and prioritize new designs, concepts, or amorphous ideas, and so we sit across from one another, politely offering one another the floor with an open mind, allowing ourselves to mold to the perception and opinions of our teammates.


But maybe it would be better to hand everyone a list of ideas before the meeting. And anyone could mark the things that he or she found notable–moving that glass an inch, so to speak–so that, by the time an idea is actually discussed, colleagues would be drawing from their best personal evaluations rather than a generic or misled group think.

About the author

Mark Wilson is a senior writer at Fast Company. He started, a simple way to give back every day.