Over four days, 117k installs and zero ratings: how Birdgo-style "Duolingo mashups" rocket up

Birdgo is a Duolingo-style learning app that packs three things into one: learning English, writing Japanese, Korean and Chinese characters stroke by stroke, and learning chess. It was listed on Google Play in Oct 19, under a developer account belonging to a Hanoi company that makes queue machines and self-service kiosks. In Nov 6 its Play installs hit 126,180, and about 117k of that came in during 10-31 to 11-03, while its rating count stayed at zero. Below: where these installs probably came from, what the three words in the title are doing, and what you could actually learn from it if you want to build something similar.

by the time it hit 220 installs, it was already #2 on the New Apps chart in ten countries

Around Oct 29, an AppBrain snapshot showed Birdgo at 220 installs, in the "100+" bracket. At the same time it ranked #2 on the Top New Free Education chart in 10 storefronts including Austria, Switzerland, the Netherlands, Norway, Mexico and Iran. In these small stores, the bar to chart on the new-apps list is only a couple hundred installs.

By Oct 31 it had accumulated 3,524 and placed in the top 200 of the Education Top Free chart in 20 storefronts: #20 in India, #39 in Indonesia, #44 in Nigeria (Google Play category chart, 10-31).

On Nov 2 installs jumped by 99,706, to a cumulative 103,230. That same day its India rank fell from 20 to 24, Indonesia from 39 to 59, and Nigeria from 44 to 84. Installs were up roughly 30 times while rankings went down. Installs rose another 17.7k on 11-03, were flat for the next two days, then gained 5,231 on 11-06. By 11-06 it held only #114 in India, #183 in Nigeria and #200 in Mexico (Google Play installs and category charts, 10-30 to 11-06).

(Play updates install counts in batches every few days; the days not listed show the same increment as 0.)

What it is

The store page's mascot is a myna bird called Leo. The Play subtitle reads "Write Japanese, Korean & Chinese stroke by stroke, learn English, play chess":

• Writing: hiragana, katakana, 431 kanji from JLPT N5, Hangul letters and syllables, and 300 HSK 1 characters. It checks each stroke's order and direction as you lift your finger, and the mnemonics are written for English speakers.
• English: story lessons from A0 to B2, set in Hoi An and street food stalls; pronunciation scored phoneme by phoneme; you can talk to Leo by voice or text.
• Chess: 7 chapters of lessons aiming at 1500 Elo, 6 bots with different playing styles, you can challenge friends, and Leo reviews your game afterward.

The monetization is all Duolingo's playbook: hearts, streaks, gems, weekly leagues, treasure chests; watch ads for gems; the Birdgo Plus subscription unlocks unlimited hearts, no ads and more pronunciation checks, with a free trial for new users. In-app purchases run $0.99 to $22.99 (Play store page, 11-06).

The same content gets a different pitch depending on the market. Its Vietnamese website headline is "Learn foreign languages through stories set in Vietnam", and the selling point is "specifically pointing out the pronunciation mistakes Vietnamese people commonly make", like reading /θ/ as /t/ or dropping final consonants; the Chinese course is pitched as "HSK 1 made for Vietnamese learners". The iOS version listed in Nov 4 is titled "Birdgo: Languages & Chess", leading with English and chess. Only the Play version changed the title to "English, Kanji & Chess", aimed at people who speak English and want to learn to write Chinese characters.

Where these 117k installs came from

Most likely concentrated paid acquisition or incentivized installs. This is an inference; I couldn't find specific channels. Three reasons:

1. Six-figure installs, zero ratings. Users who arrive through search or word of mouth usually leave hundreds to thousands of ratings at that scale. Birdgo's Play page still showed "--" for the rating through 11-06, with zero reviews.
2. Rankings fell on the day installs spiked. One possibility is that the new installs came from regions that aren't being tracked. 11-03 pushed the 1.0.7 update that added Persian, Arabic, Russian, Turkish, Hindi and Indonesian interfaces all at once, and the app had charted in Iran's new apps list too. The other possibility is that Play discounted this batch as low-quality traffic when calculating rankings. Either way, it doesn't look like organic traffic.
3. Another app on the same account follows the same curve. The wallpaper app EdgeMuse, under the same developer account, launched in Oct 2; it gained +45k in a single day on 10-24 and +85k in a single day on 11-06, flat in between, with ratings also empty. This account's pattern is to release a batch of installs every so often.

Searching for birdgo, anatoforge and smartsol in the public ad libraries of Meta, Google and TikTok turned up nothing, but these libraries have incomplete coverage, so no results doesn't mean no spend. Searching birdgo on TikTok returned 0 posts, and X had 0 posts in the last 30 days; there's no press coverage. The only thing online is a Hindi YouTube video, "Birdgo App Kaise Use Kare" (how to use Birdgo), recorded when the app was still in the "1,000+ downloads" bracket. It's the template content of an Indian channel that makes a tutorial for every new app, and the video came after Birdgo had volume, not before; it isn't the source of the volume.

The three words in the title drive impressions, not installs. English, Kanji and Chess are three keywords that each have search volume; packing them into one title lets a single package show up in three keyword pools at once. Add the 9 interface languages it had already set up, and within 12 days of launch it was on the education charts in over twenty storefronts at once. New-apps chart exposure is limited in itself, but it makes a just-launched package look like a real product in each country's charts and search results, so the paid traffic that follows has something to land on.

Shipped 6 versions in two weeks. 1.0.2 (10-22) to 1.0.7 (11-03), roughly one every two or three days, with the last one adding languages for the incoming traffic. It looks like the team was adjusting based on the data their spend was returning (inference).

Who it's competing with

(Google Play installs, around 11-06)

This table shows two things. First, none of Birdgo's features are unique: even the "languages plus chess" combination was done first by Duolingo. Stroke-order checking is indeed weak in Duolingo, but small tools like Skritter have done it for years. Second, an education app reaching 100k installs within two or three weeks isn't unusual on Play this autumn; several did it in the same period, and Birdgo just had the largest one-week gain among them.

Almost nobody knows about it yet: no press, no discussion, revenue isn't public, it doesn't appear on the grossing charts, and from the outside there's no way to tell whether the bought traffic paid back.

If it were you

C: you can learn a cold-start playbook, but cloning a Duolingo bundle isn't worth it. Birdgo's installs were bought and the product is a Duolingo knockoff. The only thing you can really replicate is the process: keyword-stuffed title, languages added first, pushing into small-market new-apps charts, then batched volume. That process only pays off for a team that knows paid acquisition and has the budget. In this category Duolingo is free, has everything, and spends heavily on ads, and whether the ROI works out is something Birdgo itself hasn't proven.

Who it suits: a small team that has already run Play paid campaigns and has a learning-app shell, looking for a way to get a new package ranked and visible in its first two weeks.

Who it doesn't suit: indie devs who haven't run ads and are relying on organic store traffic. Birdgo itself shows that stuffing three big keywords into a title and landing on twenty storefront new-apps charts only brought in 3,000+ installs before the big batches (inference, 10-31 cumulative 3,524). It's also not for anyone who wants to head-on build "another Duolingo".

How to do it: if you follow this, build the piece that Birdgo didn't do well and that Duolingo doesn't do well either, not another bundle. Two directions are clearest:

• Pronunciation correction tailored to one native-language group. Birdgo's "catching Vietnamese mistakes" pitch is the right direction, it just got buried in a bundle. Switch to Indonesian, Thai or Brazilian learners of English, or English speakers learning Chinese tones, and design lessons around the mistakes that group makes most. ELSA is the benchmark, but it only does English and doesn't split by native language.
• Stroke-order writing for kanji and kana only. English speakers learning Japanese get stuck on writing in the first few months, and Duolingo is weak here. Build a small, focused tool; subscription pricing can run higher than for a bundle.

The keyword-stuffing trick is reusable, but the words should be ones the same type of person searches for (e.g. Kanji, Hiragana, JLPT N5), not three unrelated subjects.

How to acquire users: use keywords in the title and multiple interface languages from day one, aim for the education new-apps charts in small markets in the first week, and 200 installs can get you a top-few spot and some cheap exposure. Then run a small Google App campaign in cheap markets like India, Indonesia, Nigeria and Mexico to test conversion, and check retention and the trial-to-subscription rate first; only increase spend once the numbers hold. Also, trigger the rating prompt: six figures of installs with zero ratings makes later organic users suspicious at a glance.

Timing: there's no seasonal deadline. To chart on the new-apps lists you have to concentrate effort in the first two weeks after launch; once that window closes, it's much harder to climb.

Signs you've read this wrong: (1) Birdgo charts on any storefront's education grossing chart within one or two months, or its rating count climbs into the thousands with an average above 4: that would show the bought installs are real users and the bundle playbook pays back, so it's worth taking seriously; (2) conversely, if installs stall around the low hundreds of thousands for weeks and updates stop, the batches didn't pay back, and even the publisher has given up on the playbook; (3) if Duolingo ships stroke-order writing or native-language-specific pronunciation correction, the two gaps described above close.