A Vietnamese team teaches Taiwanese people to learn Japanese with Crayon Shin-chan: Lingwa climbed the education chart via Threads check-in accounts, but the pool is still small

Lingwa is an app that teaches languages through videos, books, and news: tap a word for its meaning, AI shadowing with scoring, dictation, and flashcards, all in one phone app, with a full Traditional Chinese interface and one subscription for 14 languages. It's built by a Vietnamese team that has sold a TOEIC app in Taiwan since 2020. It launched in August, and with a circle of Threads accounts posting "Learn Japanese Day N" every day, plus some ads, it peaked at No. 10 on Taiwan's iOS education free chart in early November. But it still has only just over 10k installs on Android. Below: how it got its push, how it makes money, what it does poorly, and whether it's worth following.

"What app is this?"

Around Nov 1, under a Threads post, a user named @rifkymaaulaana (judging by the topics they follow, likely Indonesian) tagged Meta AI: "what app is this?" Meta AI replied:

"It's Lingwa — the app OP named in the post. The clip is Crayon Shin-chan playing inside Lingwa's video player, with Japanese + furigana + English subs you can tap line-by-line."

The original post has been deleted; what remains is the scene: Crayon Shin-chan playing in a player, with three lines of subtitles below it (Japanese, kana, English), and every line tappable. Getting people to stop and ask "what app is this" is exactly what posts like this are meant to do.

Threads has a whole stream of these 'daily study log' accounts from Taiwan. @celinelaurent0812 (44 followers, bio says Marketing) posted in Oct 12: 'Lazy Japanese learning Day 37... 37 days straight using Lingwa'. Then: 'I just found that Lingwa has English novels sorted by B1 and B2 difficulty, even The Little Prince.' @anan_yingwen (780 followers) has talked about it in nearly every post since Oct 16, including this one from Nov 6: 'I really love listening to news, reading articles or stories during my commute... just tap once and see the word explained instantly.'

What it is

It takes the LingQ and Language Reactor approach of 'learning languages from real content' and puts it in a phone app. Watch a video, read a graded novel or a news article, tap any word you don't know to see its meaning, and it goes into your personal vocabulary. Dictation: listen to a sentence, type it out, then check it against the original. Shadowing: repeat after the original audio, with AI scoring. Saved words are reviewed with flashcards and spaced repetition. You can also import your own YouTube videos. The latest version (1.0.8,11-05) adds Japanese news, podcasts, and a 'continuous learning' home-screen widget.

None of the individual features are new. What's new is putting them together: everything done on one phone, fully in Traditional Chinese, with one subscription covering 14 languages including Japanese, Korean, English, French, Spanish and German.

It's made by Lingoland. The company is registered in the Cayman Islands and Hong Kong, but lingolandedu.com is entirely in Vietnamese. The same developer has, on the Taiwan App Store, '2020-year TOEIC® Test Prep', '2022-year Learn English - Lingoland', Ligu AI English speaking (June this year), and ChineseUp (last November). Learn English - Lingoland on Google Play has 232k installs. They've been selling English apps in Taiwan for six years, and Lingwa carries that localization and monetization experience over to multiple languages.

How big it is

Its home market is iOS in Taiwan. Android: 10k installs 2.5 months after launch. iOS ratings grow by only 1 to 4 a day. It cracks the top 10 because the Taiwan education category charts have a low bar: Speak, the speaking-focused app, has 38k ratings in Taiwan, the local VoiceTube has 17k, and Lingwa has just 184.

Where users come from

Check-in accounts on Threads are the main channel. The playbook is simple: a few 'language diary' accounts with tens to hundreds of followers that post every day, labeled Day N, under the 'Learn Japanese' and 'Learn English' tags, plus a few lifestyle accounts with thousands of followers (@mittynomad 4,104 followers, @katelyn_0922 2,728 followers, recommended at the end of October). The material is instantly recognizable animation like Crayon Shin-chan, with three lines of subtitles, prompting people to ask the app's name in the comments. Taiwan Threads' recommendation feed pushes posts from low-follower accounts to strangers, so this costs almost nothing.

A clue to who runs these accounts is direct: @hoobi609 wrote in English, 'I've been using Lingwa to learn Chinese', and the same account posted 'My Chinese Learning Journey – Day 20... with ChineseUp'. ChineseUp is another app from the same company. One account pushing two products from one company in rotation is strong evidence it's the company's own operating account, or the same agency running them (inference, high confidence).

They also ran ads. On the Taiwan App Store in Oct 25, a three-star review opened with: 'I saw the ad saying I could practice grammar and pronunciation by shadowing, and got burned.' Where the ads ran and how much was spent isn't visible from outside. Inference: Instagram / Threads or Apple Search Ads.

Almost nothing elsewhere. No one is discussing it on X, Reddit, Dcard, or in the media. Also, lingwa.world is an unrelated open-source project with the same name. Don't confuse them.

Where the money comes from

iOS subscriptions: NT$290 a month, NT$590 for three months, NT$860 a year (about US$27). Gems are sold as in-app purchases too. Google Play in-app purchases run from $0.99 to $47.99. That's much cheaper than LingQ or Lingopie, and the interface is in Chinese.

That people pay is certain. 11-03 It reached #74 on the Taiwan iOS education Top Grossing chart. Reviews include 'Paid, works well' (10-17) and 'Already bought a year's subscription' (10-24). How much it earns can't be seen from outside. Top Grossing rank 74 isn't a high bar in Taiwan's education category, so monthly revenue is likely in the low thousands of US dollars at most (estimate).

What users are complaining about

The positive reviews all say the same thing: learn with content you like, tap to look up and save words, and it covers multiple languages. The negative reviews are also concentrated, almost all about content:

'At first I thought this way of learning was good, but then I found there were very few videos, and many of them were pretty boring.' (App Store Taiwan, 10-24, two stars)

'There are too few video options. I think they should add some animation, like SpongeBob or Adventure Time... right now I open it and don't know what to watch.' (App Store Taiwan, 09-12)

'When the shadowing part tries to make you follow the speed and style of the native speakers in the videos, the scores are often very low... You have to speak slowly and enunciate clearly to get a high score, but then you can't learn the linking sounds of native daily conversation... I feel like I spent money but have no motivation to open this program, and I can't get a refund anymore.' (App Store Taiwan, 10-25, three stars)

'Paid, works well~ but on today's content the YouTube keeps showing "sign in to confirm you're not a bot??" so I can't read or listen.' (App Store Taiwan, 10-17, four stars)

There are also 'I already bought the annual plan but limited-time offers keep popping up' (10-24) and 'Subtitles won't load, and login often throws errors' (10-16). Two more ask for Thai to be added to 'Learn Chinese' (09-07, 10-14), and one of them says they'd subscribe if it were added.

Taken together: its biggest weakness is content. The built-in videos are few and not appealing. The shows users want, like Crayon Shin-chan and SpongeBob, are copyrighted, so users can only import from YouTube themselves, and once YouTube blocks bots, that route closes. Shadowing scoring sells 'speak like a native', but the scoring rewards slow, careful reading. The Crayon Shin-chan posts that drive signups are a good hook, but once people open the app, there's not enough to watch.

Who it's competing with

The big players are all here, but nobody builds specifically for Traditional Chinese users with 'multiple languages + learning from content'. Lingwa occupies that gap. So far I haven't seen a new product doing the same thing.

If it were you

In short: you can learn something, but following it isn't a good bet. A team that has run language apps in Taiwan for six years, operates its own Threads accounts, and has run ads reached 10k Android installs and 184 iOS ratings in 2.5 months. That result itself shows the Traditional Chinese 'LingQ' market is not big. Two things are worth taking: the Threads check-in playbook, and the part they didn't get right, content.

Who it suits: people who already have a source of content rights or stable content, like graded readers, original short dramas, or a small team that can negotiate licenses with animation or podcast producers. Or native speakers of the target language who are already in the Threads / Instagram learning communities there. Reviews show people asking for Thai to be added to 'Learn Chinese', and the person asking under the Crayon Shin-chan post seems to be Indonesian, so learners of Japanese and Korean in Southeast Asia may be a segment (inference, no numbers behind it).

Who it doesn't suit: anyone wanting to build the same thing again in Taiwan. Lingwa, VoiceTube, and Langua are all here, and the education charts don't have much money to go around. Anyone who only builds features: tap-to-look-up, SRS, and AI shadowing scores can be built by anyone now, and Lingwa itself proves it. Even with the features complete, users still complain they 'don't know what to watch when they open it'. Anyone planning to use scraped YouTube as the content library: once YouTube blocks it, the product stops.

How to do it: don't build another shell with everything in it. Get the content right first. Pick one language and one audience (say, Taiwanese people learning Japanese who watch Japanese dramas and anime), and prepare several hundred leveled videos or audio clips that can be used legally: shoot your own everyday short videos, negotiate licenses with Japanese YouTubers, or make leveled stories with AI voiceovers. Each clip should have tap-to-look-up, dictation and shadowing. Split shadowing scoring into two tiers, 'clear' and 'sounds like a native', and give credit for linking sounds. Stick to its price range, around NT$800 for the annual plan, paid up front. Don't show limited-time offers to people who already paid for the year.

How to get users: copy Lingwa's Threads playbook: three to five 'Learn X Language Day N' accounts posting every day, with instantly recognizable visuals and clickable multi-line subtitles that make people ask 'which app is this' in the comments, plus a few lifestyle accounts with thousands of followers writing about real experiences. Don't push several products from one account; @hoobi609 is how it got exposed. For ads, the 'shadowing' hook works, provided the product actually delivers on shadowing.

How much time you have: no rush, this isn't a one-off window. But Lingwa is already adding Japanese news and podcasts, so it's filling in content. If you want to enter the same market, you need to see retention within six months.

What would show I got it wrong: ① if Lingwa holds a top-30 spot on the Taiwan iOS education Top Grossing chart for three months, or passes 100k Android installs, that means the multi-language Traditional Chinese market is bigger than it looks and worth competing with head-on; ② conversely, if its Threads accounts stop posting and rating growth drops to a few per week, it means organic volume depends on the seeding, and apps like this can't sustain themselves; ③ if you build a version with better content and less than a tenth of the users it brings in via seeding are still active in week two, the problem isn't the content, it's that 'learning languages from content' itself doesn't retain people.