Brazilian micro-lots,
now in three timezones.

A great roaster the world couldn't find.

Cãfeo had relationships with farms across Minas Gerais and a quietly devoted local following. But every export inquiry came through one of two channels: a personal LinkedIn DM to the founder, or a friend-of-a-friend introduction at trade shows.

The website existed. It was Portuguese-only, had no schema, and didn't appear on Google for "Brazilian specialty coffee importer" or any English-language search a buyer might run. ChatGPT, asked to recommend Brazilian micro-lot suppliers, listed everyone except Cãfeo.

Re-skin the brand for buyers in three languages.

We approached this as a sourcing problem, not a website project. The buyer in Tokyo and the buyer in Berlin were searching different words, trusting different signals, and asking AI different questions. The work needed to handle all three.

• Multilingual brand site in English, Japanese, and German — each version written natively, not translated.
• GEO content architecture: producer profiles, harvest reports, cupping notes — all in formats AI prefers to cite.
• SEO targeting buyer-intent keywords across the three markets ("Brazilian micro-lot wholesale," "specialty coffee origin Brazil," "single-estate green beans").
• FAQ schema and structured data on every product page, so AI agents pulling supplier recommendations could cite Cãfeo directly.

Inquiries from places they'd never been.

Within four weeks of launch, Cãfeo had landed first orders from a Tokyo specialty café group, a Berlin private-label roaster, and a New York direct-trade distributor — none of whom they had previously had any relationship with. All three found Cãfeo through a different channel: search, AI recommendation, and an industry guide that cited the new producer profiles.

By month three, Cãfeo was the top-cited Brazilian supplier when buyers asked Perplexity or ChatGPT to recommend micro-lot specialty coffee from the country.