Manga Method
Language learning through webtoon comics with tap-to-translate. First software product — content and development costs led me to shelve it.
My first real attempt at building a software product. I’d seen tap-to-translate work well for reading comics in a foreign language, but only for a single language. I thought there was an opportunity to do it across multiple languages — though in retrospect, the product was way too complex for the resources I had.
The product
Webtoon-style comics formatted for phone. You read through panels in your target language. Tap on text to hear voice acting. Double-tap for a full sentence translation. Tap individual words for word-by-word translations and audio. The same comic could be available in multiple languages.
The ambitious part was user-generated content. Users could add their own translations and voice recordings, which enabled any language pair — Korean to Spanish, not just everything through English. We got thousands of lines translated into different languages. People did voice recordings too, just fewer.
The build
Hired an agency to build it. I acted as product manager — planning workflows, systems, screens, giving feedback. The cost: I don’t want to think about it. A lot of money and a lot of time. Working with an agency, everything was slow and didn’t turn out as good as you’d want.
Despite all that, we ended up with a functioning app. It did everything I wanted. Not the prettiest UI, but the comic art looked good. Had a CMS where I could add content, a user area for contributing translations and recordings. I sourced comics by buying from webtoon creators, reformatted them for the platform, and produced initial translations and voice acting to seed the content.
Why I shelved it
v1 was a viable product. It worked and got usage. I used ALR traffic to send users over, and people did read and stick around. But we had minimal content to start, and also a problem that impacts most language learning products — people don’t actually use them. They buy aspirationally but don’t follow through. We didn’t have any upfront payment either, so weren’t monetizing those types.
I probably overestimated how much traffic I could funnel from ALR and how sticky the product would be. The cost of content production and app improvements didn’t feel justified for the traction it was getting — especially self-funding with high dev costs. Selling ALR also meant losing the main traffic source.