Perk Coffee

Paul Berthelsen came to coffee the long way round. After years in law and finance, he spent over a decade farming in Kenya (at one point managing a 12,000-acre cereal farm) before moving to Singapore and founding Perk Coffee in 2016 with his wife Serena. The idea was simple and stubborn: coffee roasted to order in Singapore and delivered within 48 hours of roasting, because freshness is the thing supermarket beans can never give you.

The business model made the website the whole company. Perk Coffee had no cafe and no retail shelf. Every order, every recurring delivery and every payment would run through the store, and the founders wanted subscriptions at the centre (choose your coffee, your quantity and your delivery frequency, then change any of it whenever you like). That is a harder ecommerce problem than a cart and a checkout, and in 2016 subscription coffee barely existed in Southeast Asia.

Paul also had, in his own words, no tech background and no experience in ecommerce. He needed a partner who could turn the roast-to-order idea into working software, and then stay around while the business grew.

Success looked like recurring orders running themselves: customers managing their own subscriptions, payments renewing without drama, and two founders free to concentrate on sourcing and roasting rather than wrestling with a website.

The store carried the business through every stage of that growth: the first organic 500 customers, expansion into Malaysia in 2018 and Hong Kong in 2020, and the scaling of a model where roughly 80% of revenue comes from subscriptions. Perk Coffee describes itself as Southeast Asia's largest online coffee subscription startup, and it got there bootstrapped, with no funding rounds propping up the numbers. A decade on, perkcoffee.co is still trading on the same WordPress and WooCommerce foundation.

What we learned

When the website is the business, architecture decisions are business decisions. Choosing WooCommerce over a hosted platform meant Perk Coffee kept its margins and its customer relationships as it scaled, which matters enormously in a subscription model where the economics live in retention, not one-off sales.

The second lesson is that boring reliability compounds. There was no dramatic relaunch in this story, no rescue project. There was a sound build, a care plan, and a decade of a store quietly taking orders twice a week while the roaster ran. The best ecommerce engagements often look uneventful from the outside; the drama is in all the downtime that never happened.

And subscriptions reward simplicity. The customer promise (choose your coffee, quantity and frequency, change them whenever you want) survived ten years because the store made it true, not because the copy said it.