A San Francisco chocolatier mixes science, low-tech machinery and pure democracy in its quest to make the perfect candy bar.
Author Jane Black Photography Courtesy of Scott Beale
I’VE MADE SERIOUS SACRIFICES for a free box of chocolate (don’t ask). But in this case, all that’s required is my opinion of the product. The goods—two bars wrapped in silver foil—arrive via FedEx directly from the chocolatier, a tiny San Francisco outfit called TCHO. I break off a piece of Sample A and let it melt on my tongue.
It has a hint of peach and bright currant notes. Then I log onto TCHO’s website and open a confidential survey. Question number one: “Please rate the sweetness.” I enter three out of five, which means just right. After several more questions, I move on to Sample B. Wow—way too sweet. I mark it a five.
It turns out the two chocolates are all but identical. They are made from the same Peruvian beans, with the same amount of sugar. The only diff erence is a touch of vanilla. “A minor tweak can have a huge effect on flavor,” says Timothy Childs, TCHO’s cofounder and chief chocolate officer. “If the majority of the people say they like the one with vanilla, the next iteration will have vanilla. Then we’ll send out more and dial in the exact amount of vanilla until it’s perfect.”
Modern food manufacturers have long used surveys and focus groups— not to mention the blunt edge of industrial science—to figure out what people like. How else would we have ended up with Combos or Lunchables? But it’s rare indeed for a boutique chocolatemaker like TCHO (pronounced cho) to employ the same principles of science and polling.
Small producers, especially those who cater to foodies, pride themselves on their decidedly anti-empirical, artisanal methods. A good rule of thumb for a company that wants to stand out in the foodie marketplace: less science, more craft.
TCHO’s approach makes more sense given the backgrounds of its two founders: Childs started out as a researcher at NASA and went on to develop 3-D online worlds in Silicon Valley, and chief executive Louis Rossetto was the founder of Wired magazine. And the beta testing is only the beginning of TCHO’s tech-heavy approach. For instance, there’s a special iPhone application that, among other things, lets Childs remotely control roasting times and temperatures at the company’s flavor lab on Pier 17 along San Francisco’s Embarcadero.
“We have reexamined every part of the process and asked, Where does it make sense to do things differently?” he says.
To start, TCHO is reinventing the way sweet-tooth consumers classify chocolate. Until recently, there were only two kinds: milk chocolate and dark. And most people, Childs jokes, believed that the stuff came from a factory somewhere in Switzerland. In the 1990s, that began to change. Artisanal producers started to advertise single-origin chocolates from Ghana, say, or Venezuela. They slapped percentages on labels to indicate how much pure cacao each bar contained. The higher the percentage, the thinking went, the more real chocolate flavor.
A worthy effort, and one that makes sense. But how important is that 70 percent–pure designation in the taste? Actually not very. And does saying a chocolate is from Venezuela guarantee a quality product? Nope. While there are certainly some great chocolates made from Venezuelan beans (particularly the unrivaled El Rey), the country of origin never assures consistency, or even a product superior to, say, the average chocolate bunny. Even Venezuela can produce substandard beans.
The truth is, most of this is marketing. To cut through the jargon, Childs has coupled his cutting-edge innovations with a simplified classification system and some surprisingly low-tech machinery. When his beans arrive from Ecuador at his Embarcadero lab, he roasts a small test batch in a highly customized Ronco turkey roaster, controlled by that iPhone app. The bean grinder is an off -the-shelf Indian curry maker, which is attached to a $20 hairdryer from Walgreens. The finished product is stamped with a “flavor wheel” made up of straightforward descriptions of the cacao beans used in the bar: Chocolatey, Fruity, Nutty, Citrus, Earthy and Floral.
Rossetto and Childs are replicating TCHO’s high-tech-meets-low-tech process closer to the source, in Peru and, later this year, in Ecuador and the Dominican Republic. Along with the roasters and grinders, TCHO equips its farmers with digital weather stations to help make harvest decisions and temperature probes to monitor the beans’ fermentation.
They also encourage the farmers to become more involved. Unlike their counterparts working for other companies, TCHO’s farmers actually sample the final product. “Most cacao growers have never tasted the chocolate that is made with their beans,” Childs says. “I’m trying to democratize the process of making chocolate. Giving farmers the chance to actually taste it helps increase their understanding and improve quality.”
The same goes for tasters here at home. Which is why for the company’s first chocolate bars—TCHO version 1.0—it embraced the principles of open-source software development: The more people involved, the better the product. In December 2007, TCHO started selling $5 samples of chocolate on its website and invited buyers to send feedback. It received more than 4,000 responses, which it used to create TCHO’s first products, called “Fruity,” “Nutty,” “Citrus” and “Chocolatey,” which is a sultry, silky and profoundly dark bar.
The broad sample is a key to Chocolatey’s success. But for version 2.0, TCHO has limited its tasting circle. That’s how the FedExed box ended up on my doorstep. The company invited 120 chocolate experts, food writers and bloggers to participate in a “longitudinal” tasting, which means testers taste one version (1.9A and B, in my case), file comments, then receive a second round that incorporates their feedback.
“Reading the comments for Chocolatey, it became clear that some people had better palates and were better tasters,” Rossetto says. For example, one person might say, “The hazelnut notes were overwhelming,” while another might wonder, “Where’s the hazelnut?” “Now we can track what a particular taster thinks of the changes from generation to generation. And we can tell if we are satisfying the feedback they’ve given us.”
The results may not please everyone, of course. In the samples of Fruity and Nutty that I tried, I preferred the bar without any vanilla. Most testers, it turns out, go the other way. Indeed, the final bars don’t always satisfy Childs’ palate, either. “The process has helped me to dial in my palate to those of our customers,” he says.
Fruity and Nutty 2.0 will be released early this year, which is when TCHO’s Pier 17 factory opens to the public. Visitors can learn how chocolate is made, then tour the shiny grinders, roasters and refining and tempering machines. Finally, they enter the tasting room, where they’ll be invited to help develop the next TCHO chocolate.
If you go, tell them exactly what you think. Believe me, TCHO wants to know.
Washington Post food writer JANE BLACK used to prefer fruit to chocolate. Not anymore.