Bagan IT billed unbanked Myanmar via telco bills and sold via repair shops (interview)

In a first-person account told to Rest of World in July 2020, Bagan IT founder and CEO Ricky Thet (Thet Lynn Han) described building consumer apps in a market missing every layer other countries take for granted. He had left Myanmar in 2000 to study in Singapore — no computer science was available at home — and returned in 2010 when the country began to open up, wanting to put Burmese-language books online: almost none were digitized.

The first product was a Burmese keyboard, hard because the dominant font, Zawgyi, is not Unicode-compatible and 40–50 percent of phones were Chinese domestic-market models. Google launched its own Burmese keyboard only years later; Thet says Bagan’s now has 10 million monthly active users. Next came an e-bookstore — years of trust-building with suspicious publishers yielded 10,000 books — and then the fortune-telling app Min Thein Kha, 50,000 daily active users per a France24 report, after the e-bookstore’s payments proved the concept to the famous fortune teller’s family.

Payments were the hardest wall: under 2 percent had credit cards, 75 percent were unbanked. The only entities reaching everyone were the telecoms, so Bagan persuaded them to bill app purchases onto phone bills — with guaranteed minimum revenue, and design that made clear what users were paying for. The price was steep: 45–60 percent in charges versus a credit card’s 5. And since few users stayed online — data was costly and switched off unused — Bagan used the technicians at local tuck shops as its support network, sending relationship teams nationwide to train them on each release.

Trust got its own mechanism: in the early years, anyone claiming an incorrect charge got their money back, no questions asked. 'If we didn't get it right, how would users ever trust something like mobile money?' Thet says. He frames it as an investment: even in 2020, only about 30 percent of Bagan's work was building the apps — the other 70 percent was infrastructure.

Use who already exists: phone-repair tuck shops became the human support network, and telecom billing became the payment rail.

Data was so expensive that users switched it off, making analytics and remote support nearly impossible — human networks filled the gap.

No bank cards and no banking meant no conventional payments, so the telcos' phone bills were the only rail covering everyone.

Guaranteed minimum revenue was the entry price for telco billing, which made product quality and support capacity existential.

First-time users of digital payments needed demonstrated trustworthiness, hence automatic no-questions refunds.

Where the infrastructure is missing, don't wait for it: recruit the street-level institutions people already trust and route payments through the pipes that reach everyone.

As of July 2020, Thet reported the keyboard at 10 million monthly active users, 10,000 Burmese books digitized, and Min Thein Kha at 50,000 daily active users; mobile money was growing in Myanmar, but about 70 percent of Bagan's work remained infrastructure — payments, support and trust — rather than app code.