A reading app that wasn't selling, then "AI Talk" got added to the name, and it hit 59k installs in a week: Lingtutor and a small Ho Chi Minh City studio that lives on paid acquisition

Lingtutor was originally "Lingtutor: English Reading," a graded English reading app listed in late July that got only a few dozen installs a day all of September. Then it was renamed "Lingtutor: AI English," and a line was added to the top of the store description: "🆕 NEW FEATURE: AI Talk." Starting 8 in October, installs suddenly spiked. By November the cumulative total hit 115k 6, half of it from the last 7 days, and it entered the education Top Free charts in Mexico, the US, and Argentina. The product barely changed; what changed was the sentence the ads could say. Below: how it took off, where the money came from, who it's competing with, and what you should take from it.

A Spanish-language review

On Lingtutor's Google Play reviews, the top one is from Oct 1, written in Spanish by a user named Naty Venegas:

ninguna estoy en sero vi la aplicación y kiero aprender inglés gratis

It means "I know nothing at all, zero background, saw this app, want to learn English for free." The spelling is wrong (sero should be cero, kiero should be quiero). 5 people marked it helpful.

This review is basically the user profile of this app: Spanish speakers, starting from zero, wanting to learn English for free, who saw it somewhere rather than searching the store themselves. People like this install a lot but rarely leave ratings. 115k installs produced only 78 ratings.

What it is

The store page still leads with reading: each short article is tagged by difficulty, from beginner to upper-intermediate. You can listen to the audio while reading, tap any unfamiliar word to save it to a flashcard deck with spaced-repetition reminders, answer a few comprehension questions after reading, and track progress with daily streaks. The UI is pitched as "calm, distraction-free."

The AI piece was added later and is called AI Talk: you chat about everyday topics with an AI tutor, which points out grammar and word-choice mistakes on the spot and asks you to say it again. The description puts it first and marks it with 🆕, but in the feature list it's fourth.

Free to download, contains ads, one in-app purchase at $12.99 to $36.99 (Google Play, 11-06).

Who's behind it

The developer is listed as MTD TECHNOLOGY COMPANY LIMITED, and the brand is tarunow. The website's first line reads "tarunow is a small studio in Ho Chi Minh City," and the homepage lists its five apps on Google Play: three English-learning apps (Saya for business English, Lingtutor for reading plus AI conversation, Wiss AI for speaking practice with shadowing), a photo editor called Fliq, and a blood pressure and blood sugar tracker called Cardia, plus a game site with 164 browser games. It also offers game publishing, co-development, and outsourcing to outside clients.

The careers page is blunt about how it makes money. The website lists just one job, "Mobile Growth Intern (UA & Operations)," which opens with:

MTD Technology is a mobile app publisher (Android & iOS) for global markets, growing our user base through paid advertising on Meta, TikTok, and Google Ads.

The intern's job is to set up campaigns on Meta, TikTok, and Google, set budgets, and watch the data every day; report Spend, Installs, CPI, ROAS, LTV, and retention; write briefs for video creatives, run A/B tests, and summarize which creative won; and also manage store titles, descriptions, and screenshots. The stipend is 3M to 5M Vietnamese dong per month, roughly $120–200. No experience required, "we'll teach you from scratch."

In other words, this is a small publisher whose main business is paid user acquisition, and the app is the vehicle for the ads.

Why it took off in October

The curve is a switch, not a slope. In August, cumulative installs reached about 700. Early September, about 10k. Then September barely moved: single digits to a few hundred a day. October 8 had 2,702 in a single day; October 13 had 4,398 in a single day; cumulative hit 56k by end of October. November 2 brought 14k in one day; November 6 brought 10.5k in one day (third-party estimates plus public Google Play install counts). The data also comes in bursts: 10-30, 11-01, and 11-04 were 0 days, then the next day jumped by thousands to 28k. This is the shape of a paid budget switched on and off. Organic word of mouth never built up.

The rename and the scale-up happened together. The software download site Softonic still stored the old name "Lingtutor: English Reading," and its description is all about leveled reading, tap-to-save flashcards, and post-reading quizzes, without a single mention of "AI." The current name is "AI English," and the first line of the description is AI Talk. Version 0.3.4 came out in Sep 17 and 0.3.5 in Oct 15, and scaling started in Oct 8. From the outside, we can't see exactly which version added AI Talk or whether the two are causally linked (inference: "AI real-time correction" makes far better ad creative than "leveled reading," so the repackaging was likely done to support paid acquisition).

It picked chart spots with low barriers. In Nov 6 it appeared only on the education free charts: #13 in Mexico, #20 in the US, #36 in Argentina, #88 in Spain, #92 in Turkey; it never charted on any overall or top-grossing chart (Google Play, 11-06). Installs are cheap in Mexico and Argentina, and many Latin Americans want to learn English, so a few thousand new installs a day is enough to land in the top twenty of the education free chart. The #20 spot in the US was likely driven in part by Spanish-speaking users (inference).

Almost nobody found it on their own. 78 ratings against 115k installs is about 0.07%; similar apps that grew through organic traffic usually sit above 0.5%, and EngLearn Speak, which also relies on paid ads, is at 0.06%. The distribution is lopsided: 5 one-star, 52 1-star, and 13 2-star ratings 0, the pattern you see when in-app prompts push for ratings (unverified). On X, there were 0 posts over 90 days; searching "Lingtutor" on TikTok turns up only tutorials for a mobile game hero named Ling in Honor of Kings; the only search results are APK download sites and automated ranking sites, with no press and no Reddit. Which creatives ran and how much was spent isn't visible in public ad libraries, which doesn't mean nothing was spent.

Where the money comes from

Two revenue streams: ads and in-app purchases. $12.99 to $36.99 most likely corresponds to weekly, monthly, and annual subscription tiers (inference). In places like Mexico and Argentina, where users have low willingness to pay, the model is to recover part of install costs through ads first, then take subscriptions from the small share of users who will pay. This is the playbook Vietnamese and Turkish publishers use most often.

It hasn't charted on any Top Grossing list, so its paid scale is probably still small, and there's no public revenue data. The real economics also aren't in Lingtutor alone: three English-learning apps share one content and AI base, and whichever one pays back fastest gets more budget. Right now Lingtutor is the best performer. Its sibling app Wiss AI has about 57k installs, and Saya has 5,000 more (AppBrain).

Who it's competing with

AI-powered English learning is one of the most crowded categories on Google Play's Education charts. In the last 7 days alone, 27 more English apps than Lingtutor were newly added (Google Play, 11-06):

The one thing Lingtutor does differently from the others is that it builds on reading, with AI conversation as a supporting feature, closer to the Beelinguapp and LingQ approach, while most rivals are pure AI chat partners. But this difference isn't why it grew: its install numbers depend on ad budget and per-country install costs.

Few people know about Lingtutor, but "AI English app" is no secret at all.

If it were you

A one-line takeaway: Don't build yet another AI English app. What's worth learning from Lingtutor is this move: an old app that wasn't selling gets no product rebuild, just an AI feature that's easy to show in ads, a rename, and a new first line in the description, and then it's scaled up again through paid ads.

Who this suits: small teams that already have one or two apps that are doing modestly, can run app-install ads on Meta, TikTok, and Google, and know their CPI and payback by country.

Who it doesn't suit: indie developers hoping to grow through product quality, store search, or community word of mouth. In this category, product quality barely affects install volume, and without ad budget this path doesn't work.

How to do it:

• Start with your old apps: find a solid feature whose selling point you can't explain, then add an AI feature you can demo in a 15-second video (photo recognition, real-time correction, one-tap generation), put "AI" in the title, and lead the description with the new feature.
• Don't build in English learning. This "old app plus AI repackaging" approach works best in categories not yet saturated by dozens of paid-acquisition teams, or in narrow English-speaking niches: a specific exam, a specific profession (nurses, seafarers, customer service), or Latin Americans learning Japanese or Korean.
• Don't try to compete on AI conversation itself, since ChatGPT does it better for free. What keeps people around is content leveling, streaks, and review reminders, which is exactly Lingtutor's original foundation.

How to acquire users: Follow its skeleton: use video ads to demo the new feature; start with Spanish-speaking countries where installs are cheap and the education free chart is easy to enter, such as Mexico and Argentina, and aim for the top twenty of the category chart to pick up some organic store traffic; run ads plus subscriptions as a dual revenue model, with cheap countries recovering costs through ads and countries that can afford it, like the US and Spain, paying through subscriptions.

When to move: No rush. The window in English learning has already been largely filled; the "old app plus AI repackaging" play isn't time-bound, but the novelty of "AI" in ads is fading, so the earlier you apply it to a category where no one has repackaged this way, the better.

What would mean we got this wrong: If Lingtutor breaks into the top 50 of a country's education grossing chart over the next two or three months, it would suggest these small shops' repackaging is stronger on payments than it looks, and the numbers would be worth recalculating. Conversely, if its daily installs fall back to a few hundred, or tarunow moves the budget to another app, it would show that growth sustained by "adding AI and renaming" was a short burst, with payback never actually worked out.