Selfmade Modal

public source · not yet founder-confirmed · 2024-05 · 🇸🇬 Singapore · SaaS / developer tools · solo founder

The image-generation API that won by staying put

A Singapore solo founder shipped seven products in eight months for $0, then built Bannerbear, a template-to-image API, to about US$616,000 a year without investors.

Business
Auto-generated images and videos from templates, sold as an API (Bannerbear)
Founder
Jon Yongfook (solo founder; remote team of seven by 2022)
Started
2019
Start-up capital
about US$200,000 of personal savings across the 12-startups-in-12-months run (founder's own accounting, Aug 2022 post)
Revenue
US$51,300 / month ≈ US$616,000 / year (Starter Story breakdown, May 2024; the founder's later public updates point higher)
Team
Remote team of seven serving 500+ customers by August 2022

Numbers are ranges or the figure the source published; we never publish a founder's documents.

First customers — the part nobody publishes

Channel: Shopify app store, then developer content and the founder's public build log

What didn't work: Seven products launched January–August 2019 with $0 in monthly recurring revenue; the founder's retrospective lists them as the tuition for the eighth.

“I was already ramen profitable and suddenly my whole mindset changed. I kind of just wanted to grow organically at that point.”
— Jon Yongfook (solo founder; remote team of seven by 2022), founder interview, 2024-05

Started as Previewmojo, a link-preview image tool that reached US$400 a month by December 2019 through a Shopify app-store listing; rebranded Bannerbear in January 2020, rebuilt as a developer API that spring and left the Shopify app store. US$3,771 a month by August 2020, when the founder decided to stay bootstrapped. A 50:50 code-to-marketing week, documented in his milestone posts, carried it to US$10,455 a month by January 2021 and past US$50,000 a month by July 2023.

Would it work in…

The same idea, other markets — our read from the database's country filters, not a forecast.

  • Ho Chi Minh City — likely — the product is location-free; the moat is the public build log, which takes years
  • Jakarta — likely — same reason; local payments are the only friction

Similar ideas in the database

Starting this? Build the brand kit

Five questions → name lockup, palette, type pair, social templates. Free.

Build my brand kit →

Verified with the founder where marked · numbers as ranges · founders anonymised on request · corrections to [email protected] · ← the database

From the paper