Angus Cheng built Bank Statement Converter in early 2021 after spending about a week in Kotlin testing whether he could reliably pull transaction data out of PDF bank statements. A friend warned him that working with PDFs was a nightmare, which only made him more curious: if the problem was that hard, a good solution probably did not exist yet. He started building with a second friend handling the front end, and the site went live just days after they first sat down together in April 2021.

The early model was pay per use. Cheng added the ability to buy credits to process PDFs in May, then waited a week with no sales and started to think the app was a dud. The first sale arrived as a support email while he was about to start a six hour hike: a user had bought $43.99 of credits and the app had crashed on them. He fixed it from his phone on the trailhead. Soon after, his co-founder dropped out, saying the project felt boring, and Cheng continued alone.

The one-off credit sales never covered the Google Search Ads bill. Over the first stretch he spent $858 in ads to bring in $248, then $3,050 more in ads for $740 of revenue. He switched from credits to a quarterly subscription, which converted far better, and kept improving the processor to handle scanned PDFs and documents that were not actually bank statements. By October 2021 he was demoralised enough to turn the ads off entirely and consider going back to a bank job.

The turn came from content. After reading an article that convinced him to start blogging, Cheng began publishing posts. Traffic showed up almost immediately, one post reached the front page of Hacker News, and even though none of those visitors signed up, the backlinks pushed his Google rankings higher. He raised prices and added monthly and yearly plans. By February 2022, roughly ten months in, recurring revenue passed $1,000 and his ad spend was zero. He later reported MRR climbing past $5,000 in October 2022 and beyond $7,000 by February 2023, almost entirely from organic search.