A Thai Japanese teacher built an app of his own and posted a video: Otomo and a cold start with a built-in audience

Otomo is a Japanese-learning app for Thai users, with a Thai interface, going from kana to N1, and you can raise a little Japanese yokai. Its maker is Takdanai Khuntong, the host of the Thai Japanese channel "JPN with TAKI." He posted a video in Nov 1 titled "I made a free Japanese app myself," which got over 30k views in under a week. On the Thai App Store, its ratings rose from 1 to 21 during the same period. Volume is still small. What's worth noticing is how it got started: no paid ads, no media coverage, and all its users came from his own audience.

Nov 1, one video

On the morning of Nov 1 (Thai time), JPN with TAKI posted a video titled "พาดูแอปเรียนญี่ปุ่นฟรี ยัน N1 แปลไทยทั้งหมด" ("Take a look at a free Japanese learning app, up to N1, fully translated into Thai"). The channel has about 10k subscribers. TAKI himself holds N1, normally teaches Japanese conversation, and also takes on Japanese hosting and interpreting work.

The first line of the video description reads: "I built a Japanese-learning app for Thais myself, called Otomo (お供), meaning a companion who goes along with you." The description also gives his reasons: many Thais have studied Japanese for a long time, some have passed N1, and still can't open their mouths; many others never started at all. So he wanted an app that covers everything from hiragana to N1 and actually trains real speaking. The description ends by asking people to rate and review it on the App Store, and to tell him in the comments about any bugs.

As of 11-06, the video had 30,640 views, 1,221 likes and 95 comments (YouTube page). The top comment, with 42 likes, reads: "Thanks for making this app, letting Thais learn Japanese for free, and specifically for speaking practice." Others include: "I'm one of the beta testers!" "Finally no ads interrupting, I'll save up to buy Plus" "In 7 months I'm going to Japan alone, testing it now" "A good example of using AI well." One comment with 4 likes says: "Can't find it on Play Store 😢"

You can see these people on the App Store too. One 5-star review says, in Thai: "ตามมาจากยูทูปครับ," meaning "followed over from YouTube."

What it is

An iOS app of 603MB, listed in Oct 29, in the Education category. The developer has only this one app.

• Vocabulary cards: N5–N1 vocabulary with Thai pronunciation and definitions; spaced review pulls out words you're about to forget, shown in the interface as the "Forgetting Garden."
• Photo word lookup: point the camera at signs, menus or books, tap a word for its meaning and pronunciation, and save it straight into review.
• Speaking: "Dojo Tengu" gives topics by level. You record yourself, AI scores your pronunciation, and explains in Thai what went wrong.
• There is also N5–N1 grammar, 2,136 kanji, listening, reading, a Japanese-Thai bidirectional dictionary, and a kana introduction.
• Gamification: pick a Japanese yokai or small animal as your companion. The more often you review, the faster it grows and evolves. A map adventure across 9 regions lets you battle using the words you've learned and the yokai, and earn coins to buy outfits.

Where the money comes from

Learning is free every day, with no ads. There are two paid tiers: Plus at $5.99 a month or $34.99 a year, unlocking all speaking lessons and 20 AI scorings a month, with unlimited vocabulary and photo lookup; Premium at $49.99 a year, adding PRO courses and raising AI scoring to 60 a month (App Store in-app purchase list, 11-06).

This structure has an advantage: AI is the costly part, and it's capped by count, so costs are easy to calculate. However, it hasn't yet reached the top grossing charts in Thailand, so it's unclear how many people are paying.

How fast it grew

App Store Thailand, 11-02 to 11-06. On 11-06, the store page already showed 21 ratings, averaging 4.67 stars. It peaked at #31 in the Thailand Education Top Free chart.

Ratings started climbing the day after the video went up, which lines up with the timing. Don't read too much into "31 on the chart": the Thai education free chart has a low bar, and a few dozen downloads and a dozen-odd ratings a day is enough to reach the top 50. Going by the rating count, total downloads are probably a few thousand (inferred).

Why this one took off, and not the others

Search "Thai + Japanese" in the Thai store and you'll find plenty of small apps with Thai JLPT content, most with ratings of 0 or 1 (App Store Thailand, 11-06). The content is out there. What's missing is the person who can put the app in users' hands.

Several things line up for Otomo:

• The audience is the user base: people who click on Thai-language Japanese lesson videos already want to learn Japanese. A single video converts straight into downloads.
• The person is trusted: N1, years of teaching, saying "I built this, it's free." Viewers don't have to judge whether the app is legit. The comments are almost all "thank you for your hard work," with no one nitpicking features.
• The pain point is stated precisely: "can pass the exam but can't speak" is something Thai Japanese learners keep raising. So speaking and AI scoring sit at the center, and the paywall is on speaking.
• Mainstream products don't serve this: Duolingo ranks #1 in the Thai education chart with 220k ratings in total, but Thai learners often say it slows down by intermediate level and doesn't match the JLPT. Todaii and Migii are exam-prep tools from a Vietnamese team, with Thai as an add-on translation. MARU and Bunpo have English interfaces, so Thai users have to get past English first.

Two things I haven't verified: TAKI has 64k followers on TikTok and posted Japanese short videos daily before and after launch, but I found no specific post promoting Otomo. Some commenters call it "a good example of using AI well," but the developer hasn't said whether he used AI to write the code.

Who it's competing with

This slot isn't empty. Shima Bird took nearly the same path half a year earlier (Thai developer, free, Thai-language, recruiting Japanese YouTubers) and reached 880 ratings. Otomo is a latecomer on the same road. What it has over Shima Bird is that "the developer is the teacher and the channel."

Its weak spots

• No Android yet: the video description says the Android version is only open to beta testers, who sign up via his profile. Some commenters have already said they can't find it. Many people in Thailand use Android, and right now that whole group is missed.
• Overheating: within the first few days of launch, 3 five-star reviews said "the phone got hot after a short while." 11-02's 1.2 update added a high/low graphics toggle, saying "smoother and opens faster," which reads as a direct response.
• What comes after one video?: about 6 new ratings a day, riding the afterglow of one video. Where the next high-traffic content will come from isn't clear yet.

Hardly anyone knows about it: it doesn't show up in searches on X, there's no press coverage, no public revenue, and no copycats.

If it were you

In one line: what Otomo proves isn't "there's an opening for Thai Japanese apps," but "a teacher with an audience building their own app can cold-start for almost nothing." Copying the Thai-Japanese pairing is already late, and you don't have TAKI's audience. The combination is what's worth learning from.

Who it suits: a teacher who already runs a channel for some minor language pair, or a developer who can team up with one (revenue share or partnership both work). Able to use AI to write code and finish an app alone.

Who it doesn't suit: anyone with no audience who plans to cold-start through paid acquisition or store search; anyone trying to take on Duolingo head-on with a general language app.

How to do it: first find a pairing. Look for one where lots of people are learning, mainstream apps are shallow, and a local teaching creator has tens of thousands of followers. Examples: Vietnamese learning Korean, Indonesians learning Japanese, Thai people learning Chinese (inferred, you'll need to count these yourself in the stores and on YouTube). Then find the person: a channel with 10k–100k subscribers, a real teacher, ideally with paying students already. Follow Otomo's skeleton: native-language explanations + spaced repetition + photo-to-word lookup + AI speaking scores, free to learn, with AI uses paid. Before launch, recruit a hundred or two hundred beta users through the creator's Facebook group or Discord, and have them leave reviews on launch day.

How to get users: the first wave is the creator's own "I built a free app" video, explaining why they built it, with the link in the description and an ask for ratings. After that, weekly lesson videos should casually demo the app, not rely on that one video. Launch Android and iOS on the same day. Don't do what Otomo did and miss half the audience.

Why get in early: the sooner the better. Shima Bird launched half a year ago, Otomo a week ago, and two have already appeared in the same pairing. Once one app in a pairing reaches thousands of ratings, latecomers can only compete on content.

What would mean I got it wrong: (1) Otomo's daily Thai-store ratings fall below 1 a month later, and TAKI has posted promo videos again, which would show creator traffic is a one-off; (2) it cracks the top 100 of the Thai education Top Grossing chart, which would show this audience will pay for AI speaking and the path is worth more than it looks now; (3) Duolingo or a major company launches a Thai-language course organized by JLPT level with native-language speaking feedback, and the gap closes.