Song Tea’s Peter Luong, photographed in Taiwan.

Song Tea & Ceramics sources and imports Taiwanese and Chinese teas of exceptional quality and character. I spend a month each year at origin sourcing for Song. My trip takes me to both China and Taiwan, visiting as many as five tea producing regions in each country. These trips are essential to build Song’s annual collection of teas. But sourcing—in so far as it involves traveling to origin and visiting producers—is not what distinguishes Song. It is not our differentiator. Most tea companies can source, we do so with a very specific perspective and blueprint. That perspective is what differentiates us.


The groundwork for our sourcing trip is laid months in advance of the trip. As a team, we determine what part of the collection we’d like to work on. In 2018, higher white tea prices affected pricing and availability of both white and green teas, so my priority was visiting both Sichuan, Zhejiang, and Fujian. This meant that the trip had to begin earlier in the Spring harvest season because the Sichuan green tea harvest typically begins in late March. I also was interested in oolongs, so I stayed through early May.

The goal of each sourcing trip is to assemble an annual collection of teas where each individual tea is special and has its unique place relative to other teas in the collection. This sometimes means upwards of 30-40% of the teas from the previous year are swapped out and replaced with new teas that I find. This dynamic menu is unusual for a tea company: many simply refresh rather than replace teas year to year.

To do this successfully requires relationship building, a whole lot of impromptu travel to follow constantly shifting harvest times, and a whole lot of hard work. Although I think the underlying mechanics of sourcing is straightforward, getting the execution right takes coordination and timing.

I used to think that this hard work is what differentiates Song. But over time, I’ve come to realize the process of sourcing is by no means a trade secret. In fact, we are quite open about when, and where we travel, and as many do, we too photograph our sourcing experiences and post updates on Instagram (it’s @hellosongtea, check it out!).

What I think ultimately distinguishes us is the perspective that we bring to our sourcing efforts. With each successive collection I’ve sourced (this upcoming 2019 collection will be our seventh since we launched in 2013), I’ve evolved that perspective.

To build Song’s inaugural collection in 2013, I spent nearly three months visiting tea producing regions in China and Taiwan. I followed what is, in retrospect, a laughably simple guideline: to source the best teas I could find. They had to be very high quality, interesting, complex, and unusually good. And they had to be better than any teas that I had sourced before.

That first collection had quite a few gems, but we were also heavily skewed towards green teas because… well… greens were harvested first and I was impatient. But once all 22 teas that made up the collection was assembled, I knew that the collection as a whole made a lot of sense: they were interesting and complex teas that were also so delicious!

advert but first coffee cookbook now available


And with each new collection of teas, I honed that perspective. I began looking for more interesting teas. We began working directly with some producers to craft one-of-a-kind teas: whites from cultivars traditionally intended to be made into oolongs. Finessing oxidation rates and roasting to yield sweetness and florality. With each year, Song was developing and refining its flavor profile, and its perspective on what a good tea ought to be. This whole process was ad hoc and very informal, but it became more and more a part of the discussion between myself and the team before each sourcing year.

Then, two years ago, we embarked on an internal project we called “Tea by Character.” The idea was to develop a better way for our customers to navigate the tea collection on To do this, we sat down as a team to develop a taxonomy to categorize the flavor and aromatic profiles of our teas. What we discovered was so much more.

To create the taxonomy, we parsed the individual flavor profiles of each tea, and grouped them into broader categories. After many iterations, we developed seven distinct “characters”: grass, sugar, fruit, floral, wood, earth, and texture. We then ranked the teas in each group by how “grassy” or “sugary” or “fruity” the tea was relative to others in the group. For example, Fragrant Leaf ranks higher under “grass” than Dragonwell. When we summarized our rankings, we discovered that the Song collection of teas skews heavily towards three primary characteristics: sugar, fruit, and floral.

This was the “aha!” moment. We had discovered in a roundabout way the blueprint of our palate, and the character profile of the teas that end up in the Song collection. In other words, we source teas that are sweet, with distinct fruitiness and florality. They also had to be complex, texturally interesting, and delicious.

I realized that everything that I’ve sourced for Song since we launched has been driven by this character profile. It’s the blueprint that I use to make almost all of our tea acquisitions, and some key decisions we’ve made over time.

The Song collection consists of traditional teas, unique teas, and experimental teas. Traditional teas, such as our Dragonwells, are recognizable. Unique teas have unusual flavor profiles, and are less common. Experimental teas are one-of-a-kind teas we are actively involved in commissioning from producers. Regardless of type, all teas we source are done intentionally with a mind towards Song’s character profile.

I will go out of my way, and pay more, for teas that fit the character profile. In Spring of 2018, I acquired 72 catties of an oolong we call “Four Seasons Gold” from a tea maker on Lishan, a lush and beautiful high elevation tea producing region in central Taiwan. About a half day drive from Taipei, Lishan is a range some of the highest elevation tea gardens in the world. Most teas from these high mountain oolongs are light oxidized, capitalizing on a preference for light floral teas. The Four Seasons Gold is entirely different: it’s mid-oxidized, using a cultivar that is unusual at that elevation. The tea is complex, rich, and delicious.

While on Lishan, I also visited a few other producers we do not source from. One told me that the Four Seasons Gold is the type of tea that he drinks, but does not craft. Instead he makes the type of light oxidized oolongs that have wider market appeal.

One last interaction with that producer cemented my own resolve: he said that someone from Michigan had stopped in the week before and mentioned Song. My only thought when I drove away was that if a tea company from Michigan could be there talking to this guy (no disrespect to Michigan, hey Madcap Coffee!), that I best find producers that are a little more out of the way, who could produce teas that better mesh with our character profile.

Every tea company ought to develop its own character profile.

Peter Luong is the founder of Song Tea, a purveyor of fine teas from China and Taiwan. He lives in San Francisco. This is Peter Luong’s first feature for Sprudge. 

Breville Logo

Sprudge Tea Week is Presented By Breville USA.

New Rules of Coffee banner advertising an illustrated guide to the essential rules for enjoying coffee