A language learning tool that uses a browser extension and a mobile app to turn unknown words from Netflix, YouTube and web content into flashcards in one click.
A tool company that grew out of the YouTube language-learning community "MIA (Mass Immersion Approach)": first a group of people wrote Japanese-learning add-ons for Anki and taught on YouTube how to "watch anime/Netflix and mine sentences into cards", accumulating a base of hardcore immersion learners; in 2020-2021 they turned the methodology and add-ons into the Migaku brand (migaku.io, starting with a Patreon-style "Support Us"), then into a browser extension + app subscription, converting the existing YouTube audience and Discord community directly into paying users; afterwards it kept acquiring users through YouTuber affiliate links, Netflix language-learning ads, a 20K-member Discord, and "Japanese learning + long-tail search" content.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is word of mouth in the immersion community: direct visits are 51%, second is organic search, and the top keywords are mostly branded ones like migaku, migaku extension and migaku anki.
The lubricant is daily review: flashcards must be reviewed every day, so existing users keep coming back; YouTuber affiliate links and language-learning ads keep bringing new people into this community.
By inference, about half of visits are existing users coming back; about 1.9 of every 100 visits click through to the checkout page.
Model: Lenny's racecar growth framework. Engine = word of mouth in the immersion community (YouTube methodology videos, affiliate links, and recommendations among Reddit and Discord users bring people who open the site directly or search the brand name); lubricant = flashcards must be reviewed every day, so existing users return daily; long-tail search and language-learning ads are supplements. It started from YouTube viewers of the MIA mass immersion approach and Anki extension users.
Real data: 780,100 visits, month-over-month -4.5%, biggest channel direct at 50.58%, top three channel ranking (direct, organic search, referrals), organic 95.16% and paid 4.84% of search, 15,800 search keywords, top search keywords (migaku, a Japanese phrase meaning "convert address to English", migaku extension, migaku anki, th sonido en ingles nombre), top referrals (eastbourne-college.co.uk 1.68%, simplycodes.com 1.6%, theresanaiforthat.com 0.37%), social sources (YouTube, Facebook, Reddit), countries (US 32.75%, Japan 6.56%, Italy 5.18%, Vietnam 4.05%, Australia 4.05%), bounce rate 42.92%, 3.29 pages per visit, average time on site 3 minutes 10 seconds, all from the Similarweb public page for last month; the shares of the second and third channels are not given on the public page. 14,888 visits to a Stripe checkout page (about 14,900: checkout.stripe.com 12,894 + invoice.stripe.com 1,994) are from quietmoney payment page data. Semrush desktop source shares (direct 50.4%, search 42.6%, social 7.1%) are used only to judge that the second circle has a lot of search and little social, not taken as the site's true overall shares.
Supporting evidence: The migaku.io page in 2021 sold an immersion learning toolkit plus "Support Us"; YouTuber video descriptions carry migaku.com/<creator name> affiliate links; WaniKani forum users say they saw a lot of Migaku Netflix extension ads; the lifetime price rose from about $400 to $500; a press release says it is entering the Japanese market to teach Japanese people English, and the search keywords already include a Japanese phrase meaning "convert address to English"; Discord has 22,000 people.
Inference: The two channel circles are real, but how each channel splits is inference: typed URL directly 394,600 = existing users 280,000 + community recommendations 60,000 + YouTubers and ads 54,600; other channels 385,500 = community recommendations and coupon sites 105,500 + existing users searching the brand name to log in 100,000 + YouTubers and ads 100,000 + long-tail search 80,000. Long-tail search is estimated at 80,000 on the basis that "non-branded keywords are a bit over 10% of search". The 3% who visit after seeing videos and ads, 10% who click through after finding an article and 20% who click through after seeing a community recommendation are guesses.
| Checkout | 4–6 months ago avg | 3 months ago | 2 months ago | Last month |
|---|---|---|---|---|
| checkout.stripe.com | 7,867 | 17,297 | 0 | 12,894 |
| invoice.stripe.com | 371 | 0 | 29 | 1,994 |
“'Join our active Discord community!' (21,781 members); 'our 10-day free trial. No credit card required'; 'we are a ~30 person team'”Quote from the site
New-user source: SEO (Non-brand search). Direct 50.4%, search 42.6%, social 7.1% (YouTube, Facebook, Reddit); non-brand search about 17%, from long-tail terms like '住所 英語 変換' and 'th sonido en ingles nombre'. There is also a Discord community, and referrals include simplycodes.com 1.6% (a coupon code site).
Semrush source breakdown (desktop)
| Attributable new users, last month | Share |
|---|---|
| SEO | 17% |
| 7% | |
| Other | 76% |
Direct + brand search 76%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.
| Country | Share of visits |
|---|---|
| United States | 32.75% |
| Japan | 6.56% |
| Italy | 5.18% |
| Vietnam | 4.05% |
| Australia | 4.05% |
Countries (Semrush): United States 36.4%, Japan 13.3%, United Kingdom 8.0%, Canada 5.8%, Australia 4.2%
Mostly English, plus Japanese and Spanish
YouTube, Facebook, Reddit
| Source | Share |
|---|---|
| ChatGPT | 0.14 |
| Claude | 0.12 |
| Gemini | 0.1 |
AI referrals (Semrush): ChatGPT 758 (0.14%), Claude 652 (0.12%), Gemini 512 (0.10%), Perplexity 95 (0.02%)
15.8K keywords · brand ~76% · organic 95.16% / paid 4.84%
| Similarweb top 5 keywords |
|---|
| migaku |
| 住所 英語 変換 |
| migaku extension |
| migaku anki |
| th sonido en ingles nombre |
Non-brand queries: 住所 英語 変換, th sonido en ingles nombre
8 referring sites
| Site | Share of referrals |
|---|---|
| eastbourne-college.co.uk | 1.68% |
| simplycodes.com | 1.6% |
| theresanaiforthat.com | 0.37% |
| Category | Share of referrals |
|---|---|
| Dictionaries and Encyclopedias | 20.37% |
| Programming and Developer Software | 5.12% |
| Computers Electronics and Technology - Other | 2.74% |
| Digital Marketing | 1.6% |
| Computer Security | 0.22% |
About 14,888 last month (Similarweb estimate), #207 of 606; the month before was 29. Opening checkout is not paying, so this is not revenue.
New-user source: SEO (Non-brand search). New users mostly come from need-based searches Direct 50.4%, search 42.6%, social 7.1% (YouTube, Facebook, Reddit); non-brand search about 17%, from long-tail terms like '住所 英語 変換' and 'th sonido en ingles nombre'. There is also a Discord community, and referrals include simplycodes.com 1.6% (a coupon code site).
USD 10/month; USD 15/month (Early Access); USD 499 one-time lifetime
Dossier confidence: high. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.
The number: checkout visits is Similarweb's estimate of visits from the site to a checkout payment page; opening checkout is not paying, so this is not revenue. Small sites land on fixed buckets.
Data through last month. Page updated 2026-10-31.