# "Your K-pop Crush Teaches You Korean": a Shanghai team puts an AI lover into education, with one week of ads

> Source: https://stringers.io/en/apps/play_com.lovingo.ai.html (Stringers / 什么值得做, generated 2026-11-20)
> Stringers tracks which small products suddenly took off each day, investigates why one by one, and distils playbooks from them. Numbers come from 15 databases we collect daily, each with source and date.
> For agents: these investigations and the underlying databases (verified revenue, App Store / Google Play charts, ad libraries, GitHub, Product Hunt…) are queryable live over MCP — tools feed (daily breakouts) / plays (playbooks) / hot / products / data. Setup: https://github.com/komako-workshop/stringers (the user signs in with Google at stringers.io/en/start.html to get a personal URL). Site index: https://stringers.io/llms.txt

> It hasn't really taken off yet. From Nov 2 to 11 it ran a Google ad in 9 European countries with the line "Learn Korean from your K-pop crush." Installs went from 574 to 2,762. The Japan Education chart has a low bar, and about two thousand installs was enough to get it into the top 120.

[Mobile App](https://stringers.io/en/apps/index.md) · Android · written 11-13

Lovingo: AI Language Partners · Dayfold HK Limited

# "Your K-pop Crush Teaches You Korean": a Shanghai team puts an AI lover into education, with one week of ads

**What it is** An Android app where you chat in a foreign language with an AI romance character. You can upload photos to create a character, and chats include translation, shadowing with pronunciation correction, a vocabulary notebook, and affection levels with memories that unlock. It supports 6 languages.

**Why it's getting attention** It hasn't really taken off yet. From Nov 2 to 11 it ran a Google ad in 9 European countries with the pitch "Let your K-pop crush teach you Korean." Installs went from 574 to 2,762. The Japan education chart has a low bar, and roughly two thousand installs was enough to rank in the top 120.

**C** **How you could follow it** Don't spread across 6 languages. Pick one language pair (Western K-pop fans learning Korean, or Japan and Korea learning from each other) and go deep. Make short videos with an original idol-style character, copy its ad structure, and test on a small budget. Judge success by day-7 retention and first-week payments.

C grade: Lovingo is still testing itself, with no ratings and no reviews. "AI lover + language learning" has been around for three years without anyone scaling it. The only paid signal comes from Dokitalk, which uses real human teachers. What's worth learning is this team's approach of switching categories to test, and its ad hook. Following it now means you'd be validating its demand for it. Few people know about it yet

Est. monthly revenue (Google Play) **under $1K** We estimated this from its rank on country top-grossing charts and real install counts (it isn't on the grossing chart, so confidence is lower), calibrated against real revenue; typical error is 2–3x. Read it as a tier, not an exact number.

Data (updated daily) · first launch 2026-09-18 · 11-20 installs 4,234

Android · Total installs worldwide

Data 2026-11-07 → 2026-11-20 (14 days, 11 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.

Lovingo is an Android app that lets you chat in a foreign language with an AI romance partner, listed in Google Play's education category. It recently reached #112 on the Japan education free chart and looks like it's quietly gaining traction. In fact it has been live for 7 weeks with only 574 installs, and this week's growth matches a European ad that only ran for 9 days. The team behind it also has an AI character chat app with 280k installs. This time they're taking the same underlying capability and trying it in a different category. Below: what it is, where this small volume came from, whether anyone has made "AI lover + language learning" work, and whether you should follow it.

### A 25-second ad

Starting Nov 2026 2, a 25-second vertical video began running on Google ad placements in Belgium, Germany, Denmark, Spain, France, the UK, Italy, the Netherlands, and Sweden (Google Ads Transparency Center, advertiser DayFold HK Limited).

The first shot shows someone holding up a phone to film an Asian male dancer on stage, in a black tank top with a silver necklace. A female English voiceover reads the caption: "What if your K-pop crush could teach you Korean?"

In the next shot, the photo is uploaded to the app's "Create Your Character" page, which says "Upload a photo and an audio clip, and let AI bring your character to life." A caption pops up: "Wait... I can really talk to him?" The man in the photo comes alive, holds a microphone, and says in Korean: "안녕, 자기야! (Hi, babe!) Let's learn a cute Korean phrase today. Repeat after me: 보고 싶어 (I miss you)." The camera cuts to a blonde woman sitting on a bed, reading it aloud to her phone. The man replies: "Almost! Try again slowly. You're doing great!" The final screen shows three preset characters: café friend Alex, travel blogger Mina, and art student Chloe, with "GET IT ON Google Play" below.

This ad last ran on Nov 11. Another video from the same advertiser stopped on Nov 10. From Nov 12 onward, this advertiser has no ads running.

### What it is

Pick an AI character, or create your own as in the ad, and chat in one of 6 languages: English, Japanese, Korean, French, Spanish, or Chinese. The chat interface offers translation, playback of model pronunciation, shadowing, and replay of your own recordings. New words are saved to a vocabulary notebook along with the conversation where they appeared. The other half is the romance game: affection levels, collecting stamps, and unlocking "memory" images and videos as the relationship progresses.

The official site's headline is a single line: "Learn a language. Through romance." The homepage features only two Japanese characters, Aoi Tanaka from Tokyo and Ren Sato from Kyoto. The user terms describe it as "AI language learning designed around conversations and relationships with fictional characters," for users 18 and older only, and prohibit sexual and explicit content.

The money side is still rough. Google Play lists in-app purchases at $1.99–14.99 each, and the terms say subscriptions, usage quotas, and paid features are "if offered" only. There is no iOS version. The privacy policy names Google Gemini for chat, ElevenLabs and MiniMax for voice, ThinkingData for behavior analytics, and Adjust for ad attribution. Using Adjust suggests paid acquisition was planned from day one.

### Who made it: a Shanghai team running a portfolio

The developer is DayFold HK Limited, registered in Hong Kong in May 2025, with an address in Central Plaza, Wan Chai. Another app from the same company, Plotta: Chat in Comics, is also an AI character chat product. Its feature is that what you say turns into comic panels that move the story along. The advertiser that runs Plotta's TikTok ads is "Shanghai Bai್ri Shanian Technology Co., Ltd.", so the team behind it is a Shanghai going-global group.

Plotta is the app that actually got traction for this company. It hit Google Play in Jul 2026 13, and by Nov 14 it had 286k installs. Since late October it has held #30–120 on the Entertainment free chart in the US, UK, Canada, and Australia (Google Play, 11-14). Paid acquisition drove it: Google Ads started in Jul 20 and covered 31 countries; TikTok Ads started in October, and in early Nov 7–10 dozens of creatives went live in a single burst. Third-party install estimates show a stepped curve. On Oct 22 alone it added about 30k installs (AppstoreSpy estimate), which looks like concentrated spend.

Lovingo takes Plotta's character chat, voice, and generation capabilities, reframes them, and moves them from entertainment to education for a second try.

### Where this week's volume came from

| Date | Total installs | Which charts it appeared on that day |
|---|---|---|
| 11-07 | 574 | South Korea education free #185 |
| 11-08 | 1,491 | Japan education free #143 |
| 11-10 | 1,888 | Japan education free #194 |
| 11-14 | 2,762 | Japan education free #112 |

Source: Google Play install counts and chart rankings.

Released on Sep 18, but only 7 installs by Nov 574. For 7 weeks there was almost no movement. Paid ads started on Nov 2, and installs rose several-fold within a week. The two events line up exactly in time.

"Ranked #112 on Japan education free" sounds good, but the bar is low. On the same chart in Nov 14, a #110 flashlight app gained only 7 installs over 3,587 days, and several apps ranked #105–108 gained only 7–450 installs over 670 days. Lovingo only needed 7 installs worldwide in 2,188 days to make that list, so the ranking says little about how many Japanese users it has.

There's another mismatch: the ads ran in Europe, but the chart placements were in Japan and Korea. It doesn't appear on the education charts of the seven countries where it advertised (Germany, France, Italy, Spain, the Netherlands, Sweden, and the UK). Whether the Japan and Korea numbers come from organic search driven by the Japanese store page, or from ads that don't show up in the transparency center, can't be determined from the outside (unverified).

As of Nov 14, it has no ratings, no reviews, and no appearances on any top-grossing charts. Nothing about it shows up on X or TikTok, and there is no official account. This is a small-budget test: look at one week of install cost and payments, and stop if the numbers are bad. It has not proven itself yet.

### "AI lover + language learning": three years in, no clear leader

Lovingo did not come up with this combination on its own:

| Product | Where | Scale | How it differs from Lovingo |
|---|---|---|---|
| LoveLingo: My Ai Girlfriend | iOS, listed 2023-08 | US: 63 ratings | Closest concept: the "girlfriend" is a native speaker. Solo developer, girlfriend only, rebranded in late October as an "AI language partner" |
| Turn English conversation into AI girlfriend, AI boyfriend, AI lover | Android, Japan | Very small, never charted | Teaches English only, with a "friend → lover → partner" progression and outfit changes |
| AI英会話 AI LOVE ENGLISH | iOS, Japan, 2025-03 | JP store: 23 ratings | Pitch is learning English with the feeling of fandom |
| 도키톡 Dokitalk | Android / iOS, Korea, 2026-10 | 408 installs on Play, yet ranked #23 on the Korea education top grossing chart | Koreans learn Japanese from a "Japanese almost-lover"; the other side is a real Japanese teacher, not AI |
| メローズ AI彼女・AI彼氏 | iOS, Japan, 2026-02 | 656 ratings in the Japan store, #22 on lifestyle top grossing | Pure AI lover, no language learning |
| Praktika / Speak / Talkpal | Both platforms | Play 6.4M–21.3M | AI tutor, a "teacher" rather than a "lover", heavy paid acquisition |
| Character.AI / Talkie | Both platforms | Play 60M+ | General AI companionship; many people already use it to practice languages, but there is no vocabulary notebook or shadowing |

Source: Google Play and App Store listings and charts, mid-November.

Three things stand out. First, combining AI lovers with language learning has been tried for 2023 years, and in three years no one has scaled it. The gap may be because no one has done it well, or because the demand is small. Second, the only clearly paying product is Dokitalk: with a little over four hundred installs it reached the top 30 of Korea's education top grossing chart, and what it sells is real human teachers. People will pay to learn a language within a "romantic" setup, but there is no evidence yet that the AI version can collect that money. Third, Japanese users pay for AI lovers (メローズ) and also for AI English conversation (Speakbuddy ranks #8 on Japan's education top grossing chart). Whether combining the two sells better hasn't been proven by anyone.

Compared with these, Lovingo has the most complete product: 6 languages, custom character creation, and learning tools and romance hooks all in one chat interface. But it has no reviews and no ratings, so there's no sign anyone is staying for these features. Its volume this week was bought with ads and has little to do with how good the product is.

### The most useful move in the ad

The 25-second video ties "learn a language through romance" to a specific person: a European woman who loves K-pop and wants to learn Korean. That is much easier to grasp than the website's line "Learn a language. Through romance." Fans are already some of the most motivated learners of Korean and Japanese, and the hook "he teaches you himself" can be explained in ten seconds and turned straight into a short video.

Be careful, though, about the method it demonstrates: upload one photo and one audio clip, and AI turns that person into a character that can speak. The ad shows an idol on stage. Building characters from real idols' faces and voices is a gray area for portrait rights, voice rights, and app store review. If someone uses the feature to recreate a popular idol and it spreads, the app itself will be the first to get in trouble.

### If it were you

**Verdict (C)** : You can learn something from it, but following it isn't worth much. Lovingo is still testing itself. Similar products have existed for three years without scaling, and the only visible paid signal comes from the human-teacher version. Copying it now means re-validating the demand on its behalf. What's worth taking is the team's approach and the hook in that ad.

**Who it suits** : Small teams that already have AI character chat (conversation, voice, and avatar generation all integrated), have run paid acquisition before, and want to add a new category cheaply. Or people who are native Japanese or Korean speakers, already active in fandom communities, and understand the fan audience.

**Who it doesn't suit** : Anyone building AI character chat from scratch (Character.AI and Talkie have a high bar to match, and Praktika and Speak spend heavily on paid acquisition). Anyone relying on organic store traffic (the bar for ranking on education charts is low, but the volume it brings is also small). Anyone planning to get traffic by recreating real idols.

**How to do it** : Don't copy its spread across 6 languages. Pick one pair and go deep first: Western K-pop fans learning Korean, or Koreans learning Japanese and Japanese learning Korean (Dokitalk is already collecting money in Korea). Use original characters and build an "idol feel," and don't touch real people's faces or voices. Launch on Google Play in the education category first. Set up attribution and payment tracking on day one, and use one metric as the success line: on day 7, how many people are still chatting, and whether anyone is paying to unlock the relationship or story.

**How to get users** : Make short videos using its ad structure: a line like "What if he could teach you X?" → the character calls you "babe" in the target language → you repeat after him, get corrected, and get praised. Run Google App campaigns and TikTok ads with one country, one language pair, and a small weekly budget, then check install cost and first-week payments. At the same time, post the same kind of content on TikTok under tags like learn Korean and learn Japanese, which Lovingo has not done at all.

**When to act** : There is no hard deadline, since it hasn't proven itself yet. One signal to watch: Lovingo released a new version on Nov 13. If it restarts ads, swaps creatives, or adds countries before December, that suggests the first round of tests satisfied them, and the category will heat up. Whoever moves first will gain the advantage then.

**What would mean I got it wrong** : ① If Lovingo passes 50k installs and hundreds of reviews by 2027 January, or cracks the top 50 of any Japan or Korea education chart, people are paying for AI romance-based language learning. My call was too conservative. ② If it stops buying ads and installs flatline at a few thousand by year-end, the team with the data has walked away from it, and this opportunity should be dropped. ③ If Character.AI, Talkie or Duolingo adds romance characters plus vocab lists and shadowing, there's no room left for a small team.

A final note for builders: relaunching a capability that already works in a new category, with a new pitch, is the cheapest way to test. AI character chat that's been run in entertainment can move to education, where chart thresholds are lower, ads are easier to get approved, and users have a legitimate reason to pay. Whether it sells after the switch still depends on the data from the first week of ads.

## Ads it's running

**Google** has been seen 1 times. It is not running ads now. The earliest was 2026 Nov 2.

[Google · video · 9 days](https://adstransparency.google.com/advertiser/AR00423674033846878209/creative/CR12202946611373932545?region=anywhere)

These are the longest-running ads in this set. Long runs usually mean the creative is working. Data comes from the official Meta, Google and TikTok ad libraries and updates daily; open the official library to view the original creative.

[Android store page ↗](https://play.google.com/store/apps/details?id=com.lovingo.ai)

**Research notes: competitor table, acquisition evidence, full research report**

Competitors

| Name | Where | Launched | Scale | How it differs from this one |
|---|---|---|---|---|
| LoveLingo: My Ai Girlfriend | iOS | 2023-08 (2026-10-28 sent 2.0, rebranded as a language partner) | US: 63 ratings | Closest concept, but only "girlfriend", iOS only, indie dev. Lovingo lets you build your own characters, teaches 6 languages, and has shadowing and a vocab notebook |
| Turn English conversation into AI girlfriend, AI boyfriend, AI lover (com.efriends.chat) | Android | Not verified | Has never charted on any of our tracked lists. Very small | Japan-only, teaches English only, has relationship progression and outfit changes. Lovingo is multilingual and uses Gemini, ElevenLabs, and MiniMax for voice and dialogue |
| AI英会話 AI LOVE ENGLISH (HANAUTA) | iOS | 2025-03 | JP store: 23 ratings | The hook is learning English through oshi-katsu (fan-devotion) culture, English only. Lovingo's core is the romantic relationship |
| 도키톡 Dokitalk | Android | 2026-10-05 | 408 installs on Play; Korean education top-grossing #23; iOS Korea education top-grossing #37 | The other side is a real Japanese teacher, in a flirtatious relationship with the user, not AI. Already showing payment signals |
| AI英会話スピークバディ | Android | 2017 | Play 799K; Japan education top-grossing #8 | AI character conversation, but course-based, no romance |
| Praktika / Speak / Talkpal | Android | 2019–2023 | Play 6.4M–21.3M | AI tutor, not a lover, heavy paid acquisition |
| Character.AI / Talkie | Android | 2023 | Play 70.9M / 60.9M | General-purpose AI companionship. Users spontaneously use it to practice languages, with no learning tools |
| Duolingo Video Call(Lily) | Big-company features | 2024 | Duolingo Play 996M | Fixed-character video calls with a complete learning system, no romance |
| Plotta: Chat in Comics (sister product) | Android | 2026-07-13 | Play 286K; #30–120 on the Entertainment top free charts in the US, UK, Canada, and Australia | The same AI character chat engine, placed in entertainment, making comic storylines. Lovingo is the education-category version of it |

Where users come from

- **Confirmed** **Google App Campaigns ads** : advertiser DayFold HK Limited (AR00423674033846878209) ran 2 video creatives, one landing on com.lovingo.ai. 11-02 first shown, 11-11 last shown, targeting 9 European countries. Installs rose from 574 to over 1,888 in the same window (googleads library).

- **Confirmed** **Attribution and analytics SDKs (set up for paid acquisition)** : the privacy policy says Adjust measures acquisition and conversion, and ThinkingData tracks retention and paid conversion.

- Inferred **Store search / low-competition charts** : the Japan store page has a Japanese description; it ranks #112 on Japan's education free chart, and nearby apps gained only a few hundred to 3,000+ installs over 7 days.

- Inferred **TikTok / creators / X / media** : TikHub finds no related videos, the @lovingo.ai account doesn't exist, and x-radar shows 0 mentions.

- Inferred **Plotta cross-promotion** : no evidence found.

## Lovingo: AI Language Partners (com.lovingo.ai) research report

Research date: data card refreshed to 2026-11-14. Sources: our googleplay / appstore / googleads / tiktokads / metaads libraries (agent-im server), iTunes lookup, Jina web search and page reads, TikHub TikTok search, x-radar.

### Bottom line

Lovingo hasn't broken out yet. It's the second product from a Shanghai going-global team (Hong Kong shell company DayFold HK Limited). It reuses the character-chat tech from their first AI comic character chat app **Plotta** (286K installs on Play), repackaged as "learn a language through romance" and listed in Play's education category. In its first 7 weeks it got only 574 installs. The "7 days, 4.8x growth" we saw when it was recalled coincides exactly with a set of Google App Campaigns video ads it ran from 11-02 to 11-11 across 9 European countries, so that surge came from paid testing, not organic traction. Ranking #112 on Japan's education free chart is a low bar; a couple hundred global installs a day is enough. By 11-14 it had 2,762 cumulative installs, no ratings, no reviews, and no discussion on X or TikTok.

### 1. What it is

- **Product** : pick or create an AI character and chat with them in the target language. Chats include translation, native-speaker pronunciation samples, shadowing, and playback of your own recordings. New words are saved to a vocab notebook along with the conversation they came from. Also includes relationship / affection meters, collecting stamps, and unlocking "memory" images and videos, a dating-sim-style progression. Supports 6 languages: English, Japanese, Korean, French, Spanish, Chinese [Play store page hl=en/ja; lovingo.ai/terms section 2].

- **Positioning** : the website title is "Learn a language. Through romance." The homepage features only two Japanese characters: Aoi Tanaka, "curious creator" from Tokyo, and Ren Sato, "quiet romantic" from Kyoto [lovingo.ai homepage HTML]. The terms describe it as "AI language learning built around conversations and relationships with fictional characters," limit it to users 18 and older, and explicitly ban pornographic and explicit content. The Play age rating is 16+ (strong language) [lovingo.ai/terms sections 4, 11.2; Play Japan page].

- **Pricing** : Play in-app purchases of $1.99–14.99 each. The terms mention subscriptions, usage quotas, and paid features only as "where offered" [googleplay library app_snap; terms section 14].

- **Versions** : 2026-09-18 on Play, 11-01 update 1.1.3,11-13 update 1.1.4, with release notes all reading "Optimize user experience" [googleplay library]. **No iOS version** : searching "lovingo" in the US and Japan iTunes stores returns nothing [iTunes search 2026-11-14].

- **Store page localization** : the Japan store page description is in Japanese, while the title is still in English [Play hl=ja&gl=jp].

- **Tech stack** (disclosed in privacy policy): language model is Google Gemini, voice uses ElevenLabs and MiniMax, behavior analytics uses ThinkingData (Shuji Technology), attribution uses Adjust [lovingo.ai privacy policy sections 5, 7]. ThinkingData plus MiniMax is a typical pairing for Chinese going-global teams, and integrating Adjust shows paid acquisition was planned from day one. The Play data safety page says "no data collected," which contradicts the privacy policy.

### 2. Who built it

- **DayFold HK Limited** : registered under Hong Kong business registration number 78129797,2025-05-12, address on the 65th floor of Central Plaza in Wan Chai [totalcredit.hk / ltddir.com / Play developer info]. It also operates dayfold.ai, tagline "Be the comic artist of your life," with only policy links on the page.

- **A Shanghai company behind it** : the advertiser running TikTok ads for sister product Plotta is **Shanghai Bairi Shanniantech Co., Ltd.** , registered in China, biz_id 7688229736434171922 [tiktokads library adv_map / advertiser]. This company also made an AI community app, "Haohao" [haohao.today; App Store]. It has a similar name and focus to Shanghai Baiburmeng Technology ("White Day Dream Online", CEO Meng Shiyang, AI infinite-scenario interactive fiction, raised tens of millions in funding), but **whether the two are related has not been verified** .

- **Sister product Plotta: Chat in Comics** (com.plotta.app): AI comic character chat, where what users say generates comic panels that drive the story. Listed on Play as 2026-07-13, with 11-14 installs 286,008 and 1,665 ratings. From 10-29 to 11-14 it consistently charted on the Entertainment free charts in the US, UK, Canada, and Australia (#30–120) and the Entertainment top-grossing chart in the Philippines [googleplay library app_day / chart_rank]. The iOS version launched 2026-08-25 and has 14 ratings in the US store [iTunes]. Plotta's playbook is paid acquisition: Google Ads started 2026-07-20 and covered 31 countries [googleads library watch]; TikTok ads started 10-13 (UK), and from 11-07 to 11-10 it ran dozens of creatives [tiktokads library ad]. AppstoreSpy's estimated install curve is step-shaped, e.g. 10-08 at +14,365 in a single day and 10-22 at +29,902, which looks like the shape of concentrated ad spend [googleplay library app_day_ext, AppstoreSpy estimates].

**Verdict** : this is a going-global team running a matrix of apps. Plotta validated the "AI character + romance / story" capability, and Lovingo takes it into education and tries again with a "learn a language" pitch. The recall rationale holds up.

### 3. What the data actually shows

| Date | Total installs | increase | Charted the same day |
|---|---|---|---|
| 11-07 | 574 | — | South Korea education free #185 |
| 11-08 | 1,491 | +917 | Japan education free #143, Korea #185 |
| 11-09 | 1,491 | 0 | Japan #169 |
| 11-10 | 1,888 | +397 | Japan #194 |
| 11-14 | 2,762 | +874 (4 days total) | Japan #112 |

Source: googleplay library app_snap / chart_rank. No chart appearances from 11-11 to 11-13, so there are no observations.

- **Nearly silent for the first 7 weeks** : launched 9-18, and by 11-07 had only 574 installs. AppstoreSpy has no backfilled history for it either (0 rows in app_day_ext), which suggests it never reached a thousand or two installs before then.

- **The timing of the surge matches the ads** : in Google's Ads Transparency Center, DayFold HK Limited has an advertiser account AR00423674033846878209 focused on Lovingo. It has 2 video creatives, one of which is an app install ad landing on com.lovingo.ai. Both are **first shown 11-02, last shown 11-10 / 11-11** , targeting Belgium, Germany, Denmark, Spain, France, the UK, Italy, the Netherlands, and Sweden. Ad spend dropped to zero starting 11-12 [googleads library creative / advertiser_snap]. Installs rose from 574 to over 1,888 over that same period.

- **Ranking #112 on Japan is a low bar** : on the Japan education free chart, the #11-14 app, Smart Flashlight, gained only 110–7 spots in 3,587,105 days, and several apps ranked #108–7 gained 450–670 over the same period [googleplay library v_chart]. Lovingo gained 7 globally over 2,188 days, enough to reach #112, which says little about how many Japanese users it has. The Korean chart at #185 is the same situation.

- **Ads run in Europe, but charts in Japan and Korea** : our charts cover 7 European storefronts (Germany, France, Italy, Spain, Netherlands, Sweden, UK), and Lovingo charted in none of them. That suggests the European education charts are harder to crack than its install volume would imply. How the Japan and Korea installs were sourced can't be separated: it could be organic search from the Japanese store page, or ad spend not visible in the transparency center (unverified).

- **No payment signals** : no top-grossing chart appearances, 0 ratings, 0 reviews.

### 4. Competitors: who else does "AI characters + emotion + learning a language"?

| Product | Where | Launched | Scale | How it differs from Lovingo |
|---|---|---|---|---|
| LoveLingo: My Ai Girlfriend(Mehmet Ali Demir) | iOS | 2023-08; 2026-10-28 sent 2.0, rebranded as "AI language partner" | US: 63 ratings | Closest concept: "girlfriend" is a native speaker with her own hometown and daily life. Indie dev, iOS only, just one girlfriend |
| Turn English conversation into AI girlfriend, AI boyfriend, AI lover (com.efriends.chat) | Android (Japan) | Not verified | Has never charted on any of our tracked lists. Very small | Teaches English only, native Japanese UI, relationship progression from "friend → lover → partner", 70 scenarios, outfit changes |
| AI英会話 AI LOVE ENGLISH (HANAUTA) | iOS (Japan) | 2025-03 | JP store: 23 ratings | Hook is learning English with an "oshi-katsu" feel. Japanese local team, English only |
| 도키톡 Dokitalk(from the unknown Corp.) | Android / iOS (Korea) | 2026-10-05 | 408 installs on Play, yet ranked #23 on Korea's education top-grossing chart; iOS Korea education top-grossing #37 | Koreans learning Japanese from a "Japanese flirting partner" who is a **real Japanese teacher** , not AI. Willingness to pay is already visible |
| Linqo(Lionster Sdn. Bhd.) | iOS / TikTok | 2026-05 | TikTok 616 followers | English speaking practice partner. TikTok creatives push "a little romance", not a dating storyline |
| AI英会話スピークバディ | Android / iOS | 2017 | Play 799K; Japan education top-grossing #8 | Japan's most mature AI character English conversation app. Course-based, no romance |
| Langaku(Mantra) | Android / iOS | 2025-01 | Play 323K; Japan education top-grossing #9 | Learn English through comics. Built on content IP, no chatting |
| Praktika / Speak / Talkpal / Loora | Android / iOS | 2019–2024 | Play 6.4M–21.3M | AI tutors and virtual teachers, positioned as "teacher" not "lover", with heavy paid acquisition |
| Character.AI / Talkie(MiniMax) | Android / iOS | 2023 | Play 70.9M / 60.9M | General AI companion. Many users already practice languages with it, but it has no vocab notebook or shadowing tools |
| Duolingo Video Call(Lily) | Big-company features | 2024 (Max tier) | Duolingo Play 996M | Video chats with a fixed character, Lily. Full learning system, no romance |
| メローズ AI彼女・AI彼氏 | iOS (Japan) | 2026-02 | JP store 656 ratings, Lifestyle Top Grossing 22 | Pure AI lover, no language learning, but shows Japanese users will pay for AI lovers |

Data sources: Google Play / App Store libraries, iTunes lookup, Play / Instagram pages.

**Competitive takeaways** :

- The combo "AI lover + language learning" has existed for 2023 years, and never got big. No leader. Lovingo isn't the first.

- Small teams have popped up in both Japan and Korea. Dokitalk uses real human tutors and hit the Korea Education Top Grossing top 30 with just 400-plus installs. That suggests paying demand for "flirty relationship to learn a language" in Korea, but the scale is still tiny.

### 5. Growth tactics: the evidence found

1. **Google App Campaigns ads (confirmed)** : see section 3. 11-02 to 11-11 ran ads in 9 European countries, with 2 video creatives. The ad timing overlaps with the install spike.

2. **Adjust attribution (confirmed, prepping for paid acquisition)** : the privacy policy states it uses Adjust to "measure acquisition and conversion".

3. **Store search / Japanese store page (inferred)** : the JP listing was localized into Japanese, and low entry barriers got it charted in the JP and KR education charts. Charting itself brings some organic traffic.

4. **TikTok / creators / social content (none found)** : TikHub searches for "lovingo" and "lovingo ai" return nothing relevant, and the @lovingo.ai and @lovingoai accounts don't exist. Plotta does have @plotta.ai (80 videos, 798 followers).

5. **X / media / Reddit (none found)** : over 90 days on x-radar, "lovingo", "plotta" and "Dayfold" each returned 0 posts. Web search only surfaced store pages, APK mirrors and the official site.

6. **Meta and TikTok EU ads (not found)** : neither the metaads nor the tiktokads library lists Lovingo as an advertiser. Those libraries only track products on our list, so not finding it doesn't mean it didn't run ads.

7. **Plotta cross-promotion (unverified)** : no evidence found.

### 6. What's the differentiation, and is it why it broke out?

- **What's different** : it puts the retention hooks of a companion app (affection meter, memory unlocks, collection) and the tools of a learning app (translation, shadowing scores, vocab saved with its source conversation) into one chat interface. It's listed under **Education** , not Entertainment or Social, supports 6 languages at once, and lets users build their own characters. LoveLingo only has a "girlfriend" and is iOS only; the Japanese apps only teach English. Lovingo looks broader and more polished by comparison.

- **Why Education is useful** : Education charts have lower entry barriers; the "language learning" framing makes a companion product easier to get through review and run ads for, and gives users a legitimate reason to pay.

- **This isn't why it took off, because it hasn't taken off yet** . The current volume is fully explained by a week of Google paid acquisition tests plus the low-barrier Japan chart. No reviews or ratings, so there's no evidence users are staying for the differentiation.

### 7. What others can learn

- **Reuse the capability, switch category** : the same character chat, voice and generation stack was first validated in Entertainment with Plotta (286k installs), then packaged as "Education + romance" for the Education category. Education has less competition and lower chart barriers, and it's easier to justify payment.

- **Test cadence** : launch on Play first, skip iOS for now, wire up Adjust and ThinkingData. Then 6 weeks after launch, run two video creatives in Europe on Google App Campaigns for a week or more, check conversion, and stop if it doesn't work. This is the standard "small budget to test CPI and payment" approach.

- **The Korea signal** : Dokitalk teaches Japanese through real flirty relationships and hit the top 30 in Education Top Grossing with just 400-plus installs. If you want to build here, consider starting with the Japan-Korea mutual learning niche (Koreans learning Japanese, Japanese learning Korean) instead of generic "6 languages".

### 8. Still unclear

- Whether the Japan and Korea installs are organic or paid: the Ad Transparency Center only shows European ads.

- After Google ads 11-11, did they stop, change accounts, or swap creatives; and will 11-14's new version 1.1.4 keep scaling?

- How much it earns: it hasn't charted on any top grossing list, and in-app purchase volume isn't verified.

- Whether Shanghai Baiteng Shanian Technology and Shanghai Baimeng Technology (Meng Shiyang) are related.

- Whether it's running Meta or TikTok ads outside Europe (our ad libraries don't cover that).

This article was written by AI from public data and the data we collect daily. Numbers are sourced, and inferences are labeled.
