Stringers
Mobile App · Android · written 11-12
Dialect: Learn Your Language · Dialect studio

"The whole family talks around the table, and you can only smile along": two second-generation French migrants built an app teaching Kabyle and Wolof, and used a dozen-plus TikTok accounts to break into France's education Top Grossing chart

What it is A subscription app that teaches "heritage languages": for descendants of North African, West African and overseas-territory migrants raised in France, 5 minutes of card-based lessons a day in 22 languages the big platforms don't teach, with recordings by native speakers as the core.

Why it took off It targets the emotional pain point of second and third-generation migrants who "understand but can't speak," and Duolingo doesn't teach these languages. Since September it has run one TikTok account per language, posting daily engagement prompts like "Do you really speak it or are you faking it?" that get people arguing in the comments over who's more authentic, then used small Meta ad spend to boost the hits, pushing it as high as #8 on France's iOS free education chart.

AHow you could copy it Don't copy it in France; go find a different audience: Turkish and Kurdish communities in Germany, Moroccan communities in the Netherlands, South Asian communities in the UK, Tagalog and Cantonese families in the US. Pick languages Duolingo doesn't teach, record them fully with native speakers, add speaking practice, and copy the acquisition playbook: one short-video account per language + an engagement-prompt template + scaling up the hits with paid ads, mainly annual subscriptions.

Grade A: real demand, paid conversion already proven by France's grossing chart, and the playbook is available to any follower, while almost nobody outside the niche is paying attention. The bet is whether the "who's more authentic" comment-driven distribution still works with other migrant groups, and whether thin content causes annual renewal problems. France already has copycats, and Dialect is expanding into English and German, so the window isn't long. Circulating in small circles

Est. monthly revenue (Google Play) under $1KWe 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 2026-10-09 · 11-20 installs 3,390
Android · Total installs worldwide
1k2k3k11-0611-0911-1411-1711-203.4k
Data 2026-11-06 → 2026-11-20 (15 days, 13 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.

Dialect teaches one kind of language: the kind parents and grandparents speak at home that schools and Duolingo don't teach, such as Kabyle, Moroccan Darija, Wolof, and Guadeloupean Creole. It launched on iPhone in February 2026 and through November stayed between 25 and 47 on France's App Store education grossing chart; it only reached Google Play in October. Its volume came from one TikTok account per language and batches of "Do you really speak it or are you faking it?"-style engagement prompts. Below: what it sells, where users come from, who is already copying it, and whether you can run the same play again in another country with another group.

A five-star review: wanted to speak Kabyle with my mother-in-law

In Aug 2026 27, someone on the French App Store left Dialect a five-star review titled "J'en rêvais !" (I've been dreaming of this!):

I've always dreamed of learning Kabyle so I could talk with my mother-in-law. Thank you for making this dream come true.

Scroll through the reviews and the same sentiment keeps appearing: "Thank you, I can finally learn my father's language, we've waited a long time for an app like this" (10-27); "I looked for years for an app to learn Moroccan Darija, and after a few weeks I can hold a conversation" (05-19); "Thanks to it I can talk with my grandmother" (04-18).

The first paragraph of the store page says it all: "Do you know how to tell your grandmother 'I love you' in her language? If you grew up far from home, you know this feeling: the whole family talks around the table, and you smile along... but you can't really reply."

This is the everyday reality for millions of second and third-generation families of North African, West African and overseas-territory origin in France: understanding a bit, unable to say it. Dialect sells to that gap.

What it is: 22 "home languages" Duolingo doesn't teach

It was built by French company Dialect Studio, founded by Ahcène Bensalama and Nabil Meghni. Ahcène told the Tunisian outlet Kapitalis (2026-10-06): "We built this for people who grew up far from Tunisia and want to speak the Tunisian of home."

The product itself isn't new: 5 minutes a day of flipping cards, listening to pronunciation, taking quizzes, spaced repetition, plus a mascot-led level path and a lock-screen word widget, which is the Duolingo playbook. What's different is the languages it teaches:

  • Berber varieties: Kabyle, Chaouia, Rifian, Tachelhit
  • Three Darija varieties: Algerian, Moroccan, Tunisian
  • Wolof, Fula, Bambara, Lingala, Malagasy
  • Four Creoles: Guadeloupean, Martinican, Réunionese, Haitian
  • Kurdish, Turkish, Tamil, Vietnamese, plus Breton and Ancient Egyptian

The store page states that 10 of these have "every word recorded by people who grew up speaking the language"; the rest are mostly synthetic voices. The interface is available in 7 languages: French, English, Spanish, German, Italian, Dutch, Chinese.

Scale: the real action is on iPhone in France

Looking only at Google Play, it's small: listed 10-09, 11-13 total 2,473 installs, up 7 in 945 days, and it only reached #169 on the France Education free chart.

The volume is on iPhone:

MetricMetric
France App Store rating264 reviews, 4.69 stars (11-13); 10-28 of them are 220 reviews, adding two or three a day
France education free chart10-28 to 11-13 between #8–39, with a spike to #8 at 11-06
France education grossing chartStayed between #25–47 over the same period
Other storesBelgium education free chart #20–54; occasional education chart appearances in Canada, Switzerland, the Netherlands, Mali, Senegal and Niger
Self-reported"Over 50k people are learning their home languages with Dialect" (App Store subtitle)

The grossing chart tells the most. France is a sizable store, and a two-person team's minority-language app staying at #25–47 on the education grossing chart for weeks means people are actually buying subscriptions. How much per month can't be estimated from the outside.

On 11-06 the iOS free chart jumped to #8, Play followed two days later, and daily installs went from 0 to three or four hundred. The same week it ran a new batch of ads on Facebook and Instagram, and the Fula TikTok account posted a video that got 250k views (inferred: the two pushed each other).

Why this one: a gap the big platforms left, plus one line that hits home

First, no big players. Duolingo, Babbel and Busuu teach Haitian Creole, Turkish and Vietnamese, but not Kabyle, Wolof, Lingala, Kurdish, Tamil or Breton. One user wrote in a review: "Thank you for thinking of the language that bird forgot." (09-17)

Second, it packaged a pile of small dialects into one category. Single-dialect apps have existed for a while, like Morolingo (launched in 2024, teaching only Moroccan Darija and Amazigh), whose France rating count is 15 times Dialect's. But those apps are one dialect each and scattered across the store. Dialect's pitch is "learn the language of your home," one app for all of them, so one brand can go after a dozen-plus communities at once. Families with a Kabyle parent and an Arab parent can also find a fit: one user wrote that they spend "15 minutes a day on Tachelhit and 15 minutes on Darija" (11-09).

Third, people pay not to learn but for identity. The words that keep coming up in reviews are "roots," "origins," "grandmother," "finally." Users want to talk with family, so some will pay a higher price for it.

Where users come from: one TikTok account per language

The main account @dialect.app posted its first video in February 16, which got 165k views. From 2 to July it posted only ten or twenty videos a month, carried by a few hits: in April 15 "Dialect est enfin là !" (Dialect is finally here!) got 700k views, and in August 22 one Algerian Darija vocabulary video got 650k.

Starting in September the playbook changed. The main account posts over a hundred videos a month, and from late September each language got its own account named "X.dialect": @kabyle.dialect, @wolof.dialect, @tunisien.dialect, @chleuh.dialect, @lingala.dialect... We could find at least 13 of them, adding up to roughly 1,700 videos, with each sub-account posting two or three a day.

The content is almost all the same interactive-quiz template. Take @peul.dialect (Fula) and its November video:

Partie 4 ! Tu parles peul, ou tu fais semblant ? 🎯 Encore dix mots, trois niveaux : touriste, cousin, grand-mère. Écris ton score en commentaire 👇 Part 4! Do you really speak Fula, or are you faking it? Ten more words, three levels: tourist, cousin, grandmother. Post your score in the comments 👇

Same template, different language, different account: "Kabyles, which words do you still use?" "Kabyle or Chaoui, which one did your family pass down?" In North and West African diaspora circles, "which place's way of saying it is the real one" is a topic that starts fights by default, and once the comments heat up, the video gets pushed to more people from the same community.

The strategy depends on occasionally hitting a viral video. @peul.dialect opened its account in Sep 24, and its 81 videos drew 850k views in total. Most got only one to a few thousand views. In Nov 8, "Pour apprendre à parler le PEUL avec DIALECT !" (Learn to speak FULA with DIALECT!) hit 255k views and 856 comments, and the account reached 18k followers. The "do you really speak it or fake it" series that followed pulled in a few thousand to 38k views per video over the next few days. From Nov 9 a creator account also appeared, tagged "Paid partnership with Dialect," posting "learn Kabyle from zero in 30 days."

Paid ads came after. In Meta's ad library, its page began running ads on Jun 29, mostly boosting viral TikTok and Instagram videos as-is. From Oct 3 the ads were split by language, all with one line: "Kabyle / Wolof / Lingala... on Dialect: new words every day, native speakers, 5 minutes." In Nov 6 a batch of Lingala and Moroccan Darija ads went up. Only a dozen or so ran at once, so the scale is small. User comments match: "I stumbled on Dialect on Instagram" (08-04), "Downloaded it after seeing one Instagram video" (10-04).

Its website mydialect.io also has over three hundred pages: a landing page per language, plus guides like "best app for learning Kabyle" and "how to say hello in every language," built in 7 language versions. It already ranks near the top in search, but the traffic it brings is probably still small.

Where the money comes from: €44.99 a year, and the free version gives one set of cards a day

The French App Store price is €14.99 a month or €44.99 a year. The free version gives one set of cards and one quiz per day, and the lesson path only includes the first lesson (greetings).

The free tier is thin, so the bad reviews are mostly about price: "Without paying (40 euros) you basically can't learn anything... in three days I learned at most 10 words" (09-29); "Too expensive. I bought a year, and finished all the lessons in one evening... the so-called review is pointless" (10-26).

Another set of bad reviews is about content: Tunisian "sounds like Dutch" (07-13), Lingala "sounds like a robot" (08-29), Malagasy "the ad says native speakers, but it's obviously AI, or someone who doesn't speak Malagasy at all reading it" (10-26).

In other words, it built distribution first and is filling in content while it sells. For an annual product this is a renewal risk: someone who finishes everything in one evening probably won't pay again next year.

Who it's competing with

ProductScaleHow Dialect differs
Morolingo (2024)French iOS 4,058 ratings; Play 100k+ installsTeaches only Moroccan Darija and Amazigh; mainly free
Kalam: Learn Arabic (US, 2025)US iOS 6,814 ratings; Play 250k installs; running Meta adsTeaches regional Arabic dialects, pitches AI conversation, targets English speakers; doesn't touch Berber or African languages
Ling, uTalkPlay 2M and 1.14M installsGeneral language apps with dozens to hundreds of languages, aimed at travelers; they don't speak to "the language at home"
Lmed Taqbaylit (2025)Play 10k+ installsTeaches only Kabyle / Amazigh, free, driven by a Kabyle creator
Derja / Dziria / MghribiyaDerja: French iOS 49 ratings; Dziria: 12 ratingsCopycats, one dialect per app: 10-20's Dziria listing nearly copies Dialect's pitch: "Understand what your family says but can't answer back? … 5 minutes a day … spaced repetition"
Diaspeak (Marseille)French iOS 4 ratingsCopycat, "Arabic, Wolof, Creole and 12 other languages, to talk with your family"

(Ratings and install counts are as of early November)

Some people in France are already copying its copy word for word. But the copycats haven't gained traction yet, and Dialect itself has only reached France and Belgium.

How well known is it?

Not much. Media coverage is limited to one short article from Tunisia's Kapitalis; nobody on X has mentioned it in nearly 90 days; revenue has never been disclosed. Its reputation lives entirely in the TikTok and Instagram circles of French North and West African diaspora. Developers outside that circle have mostly not noticed it, but people inside the circle have already started copying it.

If it were you

Verdict: worth following, but don't copy it in France. Find a different audience. What Dialect proves is a playbook: find a group of diaspora children who "understand but can't speak," pick languages the big platforms don't teach, run one short-video account per language, use "how does your hometown say it" to get the comments fighting by themselves, and then boost the videos that take off with small ad spends. None of these pieces is out of reach for a follower.

Who it suits: one or two people who come from a particular diaspora community, can find native-speaker recordings, and can make short videos. Your biggest advantage is knowing the community's in-jokes and which regional wording is most likely to start a fight.

Who it doesn't suit: anyone who wants to target North and West African communities in France. Dialect already has that market locked down, and people copying its copy already exist. It also doesn't suit anyone who only wants AI-synthesized pronunciation: Dialect's most common complaint is "this isn't how we say it back home," and in this business the native speaker's voice is the product.

How to do it:

  • Change the audience. Turkish and Kurdish communities in Germany, Moroccan communities in the Netherlands and Belgium, Urdu, Bengali and Punjabi communities in the UK, Tagalog, Cantonese and Haitian Creole families in the US. First check whether Duolingo teaches the language. If it does, the room is much smaller. Dialect has opened an English account, @dialect.us, and a German interface, but neither has taken off yet.
  • Do the parts Dialect does badly. Offer fewer languages, but record every one with native speakers; add sentence and conversation practice (the most common request in the good reviews is "I want exercises where I can actually speak"); make the content deep enough that someone paying for a year doesn't finish it in one evening.
  • On price, learn from it: make the free tier a bit more limited, and make the annual plan the main product.

How to get users: one TikTok / Instagram account per language, two or three posts a day, with fixed templates: "Can you really speak it or are you faking, ten words in three levels," "which one did your family pass down," "can your kids read these 7 words." Most videos only get a few thousand views, so you need volume to wait for the one that hits. When a video does take off, boost it with small Meta spends split by language. On the website, build a "best app for learning X" and "how to say hello in X" page for each language, and slowly collect search traffic.

When to move: ideally before the first half of 2027. Dialect is already expanding into English and German, and the French copycats are building one dialect at a time. Whoever establishes themselves first in a given community makes it harder for everyone after.

Signs you've misread it:

  • If your language accounts have been running for a month with over a hundred videos and the comments still aren't fighting, that means this community lacks the "whose version is correct" drive that French North African circles have, and this distribution playbook won't work there.
  • Dialect's ranking in French education Top Grossing will likely fall in early next year, when the first batch of annual subscribers come up for renewal, which would show that renewals can't carry it. This is a business that only really works in its first year.
  • Conversely, if @dialect.us or the German version gains traction within six months, that shows this path can work, but the audience you'd pick will already be partly taken by it.

Android store page ↗

Research notes: competitor table, acquisition evidence, full research report
Competitors
NameWhereLaunchedScaleHow it differs from this one
Morolingo: Darija & AmazighiOS2024-01iOS FR 4,058 ratings; Play 100k+ installs, 4,620 reviewsOnly Moroccan Darija and Amazigh, mainly free; single-country, no "heritage language" framing, and Dialect covers a much broader set of communities
Kalam: Learn Arabic(Node Zero)iOS2025-06(Play 2025-10)iOS US 6,814, FR 490 ratings; Play 250k installs; running Meta adsTeaches regional Arabic dialects, pitches AI conversation, targets English speakers; doesn't cover Berber languages or African languages
LingAndroid2018Play 2.09M installs; iOS US 17k ratings60 general language-learning app, aimed at travelers, no North African dialect or heritage-language positioning
uTalk(EuroTalk)Android2016Play 1.14M installs150 vocabulary book, with Riffian, Wolof, Lingala, but its content is old, not organized around family scenarios, and has no community
Lmed Taqbaylit / Lmed TamaziɣtAndroid2025Play 10k+ installs; driven by Kabyle creator melissa_tamazight (238k followers)Kabyle / Amazigh only, free, leans toward a cultural movement
Duolingo / Babbel / BusuuBig-company features—Top playersCovers Haitian Creole, Turkish, Vietnamese; no Kabyle, Wolof, Lingala, Kurdish, Tamil or Breton. This gap is exactly where Dialect survives
Dziria / Derja / Mghribiya(Ismael Dubuc)iOS2026-08 / 2026-10Derja: French iOS 49 ratings; Dziria: 12 ratings; TikTok @derja has 216 3,943 followersCopycats: one app per dialect, copy lifted word for word, founder on camera for TikTok
DiaspeakiOS2026-10 (date as shown on the store listing)French iOS 4 ratingsCopycat: same positioning, "Arabic, Wolof, Creole" and 17 other languages, hasn't gained traction yet
Where users come from
  • Verified TikTok owned account network: at least 13 accounts: @dialect.app with 8,474 followers and 458 videos; @peul.dialect with 17.9k followers, 81 videos totaling 850k views; plus @tunisien, @chleuh, @kabyle, @chaoui, @wolof, @lingala.dialect, @dialect.fr, @dialect.us and others. Hits: 04-15"Dialect est enfin là" 703k, 08-22 647k, 11-08 peul 255k. The content is template-driven: "how does your hometown say it," "how many of these 8 questions do you get right" (TikHub)
  • Verified Meta (FB / IG) paid ads: Meta ad library, Page "DIalect.app": 2026-06-29 started boosting TikTok hits; 08-24 swapped in a new batch of creatives; 10-03 started language-split ads "Le kabyle / wolof / lingala... sur Dialect", plus lock-screen widget creatives; 11-06 ran a new batch. Comments also include "Les pubs nous promettent…"
  • Verified TikTok paid creator: @eldrine.dialect is tagged "Collaboration commerciale avec Dialect", running a "Learn Kabyle from zero in 30 days" series, from 2026-11-09
  • Verified Organic Instagram content: @dialect.app posts a weekly word-of-the-day and meme videos; App Store reviews read "came across it on Instagram" (08-04) and "downloaded it after seeing an Instagram video" (10-04)
  • Verified Programmatic SEO and web app: the mydialect.io sitemap lists 325 URLs: one landing page per language, plus best-app-to-learn-X and how-to-learn-X guides, across 7 language versions, with 11-09 recently updated. Not in Tranco's top 1M, so the traffic it brings is likely still small.
  • Verified media / PR: Kapitalis ran a 2026-10-06-line brief, Zoom Algérie mentioned it in passing 07-29; the press page actively pitches, listing a coverage calendar for Yennayer / Newroz and more.
  • Likely coordinated review solicitation: App Store FR shows two batches of all-5-star reviews on the same day, same version: 2026-04-14 (v1.2.0,7 reviews) and 07-07 (v1.3.0,8 reviews). Probably an update notice sent to a community.

Why Dialect: Learn Your Language (com.mydialect.app) is working

Research date 2026-11-07 (data pulled through 2026-11-13). Every number cites its source: appstore / googleplay refers to our own databases, TikTok data comes from TikHub, and Meta ads come from the facebook.com/ads/library pages.

Bottom line

Dialect is an app built by two second-generation French immigrants to teach "the language your parents spoke": Kabyle, Maghrebi Darija from three countries, Riffian, Wolof, Fula, Creole and 19–22 other minority languages Duolingo doesn't teach. It runs on 5 minutes of flashcards a day, plus recordings by native speakers. The main battleground is the French iOS store, not Play: 2026-02-24 on iOS, with Play following 2026-10-09 later. It works because three things stack:

  1. Sharp positioning: it targets the pain of second- and third-generation immigrants from France, North Africa, West Africa and the overseas departments who "understand but can't speak", with copy built around that emotional hook.
  2. A TikTok matrix with one account per language, churning out engagement-driven short videos: at least 13 accounts and roughly 1,700 videos, with a top single video at 700k views.
  3. Boosting those creatives with small Meta spends starting at 2026-06; from October, the campaigns were split by language.

The differentiator is "one app covering every heritage language, with each one labeled as native-recorded or not." That's not hard to copy, and by October copycats were already lifting the copy word for word.

1. What it is

  • The developer is Dialect Studio (a French SAS), founded by Ahcène Bensalama (born 1995, Sciences Po) and Nabil Meghni. Sources: mydialect.io/en/press and societe.com. Ahcène's personal developer account also holds two more apps, Sève (budgeting) and Saison (seasonal produce) (App Store developer page). NFX Signal lists an investor named XAnge with the same name; whether it's the same person is unverified.
  • The product: 5 minutes a day of audio flashcards, quizzes, unit-based learning paths (mascot TIPO), spaced repetition, a lock-screen word widget, and a web app. The interface is available in 7 languages (French, English, Spanish, German, Italian, Dutch, Chinese).
  • Languages: Kabyle, Chaoui, Riffian, Tashelhit; three Darija varieties (Algeria, Morocco, Tunisia); Wolof, Fula, Bambara, Lingala, Malagasy; four Creoles (Guadeloupe, Martinique, Réunion, Haiti); Kurdish, Turkish, Tamil, Vietnamese; Breton and Egyptian hieroglyphs. Native-speaker recordings cover 10–12 of these (Play / App Store descriptions and press page).
  • Pricing: the free tier gives 1 flashcard sets and 1 quizzes a day, plus the first lesson of each path. Premium is a monthly or yearly subscription; the data card lists in-app purchases at $9.99–49.99, and one reviewer mentions "40€" (App Store FR review, 2026-09-29).

2. Data: real traction is in the French iOS store

MetricValueSource
iOS release2026-02-24; the first TikTok "now on iOS" post went up on 02-26; the first wave of reviews clustered on 04-14 (v1.2.0)iTunes lookup, appstore.review
iOS ratingsFR 264 (4.69 stars); BE 23, CA 18, US 15, CH 13iTunes lookup 2026-11-13
iOS rating growthFR went from 220 (10-28) to 264 (11-13), about 2.75 reviews/day; on 7, +20appstore.app_day
iOS category chartsFR Education Top Free ranked #8–39 (11-06 climbed to #8); FR Education Top Grossing #25–47; BE Education Top Free #20–54 and Top Grossing #25–91; it has also briefly charted in Education in CA, CH, LU, NL, FI, Mali, Niger and Senegalappstore.chart_rank #10-28 to #11-13
Play installs1,528 (#11-06) → 1,932 (#11-09) → 2,267 (#11-10) → 2,473 (#11-13); on 7, +945; only on the French Education Top Free chart, at #169googleplay database, data card
Self-reported"50,000+ people learn their family's language"App Store subtitle

Judgment: In France, where chart entry is hard, the education top-grossing chart held at #25–47 for a long stretch, which means iOS subscription revenue is real. Play has only been up for a month and is a fraction of iOS volume. 11-06 The iOS free chart climbed to #8, and the jump in 11-09 and 11-10 Play daily installs from 0 to 300–400 is the same push (a viral TikTok from a language account plus 11-06 newly launched Meta ads, see below).

3. Competitors: who else teaches minority languages and dialects

ProductPlatformLaunchedScaleHow Dialect differs
Morolingo: Darija & Amazigh(AREDJAL Consulting)iOS + Android2024-01iOS FR 4,058 ratings (15× Dialect's); Play 100k+ installs, 4,620 reviewsTeaches only Moroccan Darija and Amazigh, mostly free, with an SEO blog; single country, doesn't use the "heritage language" framing
Kalam: Learn Arabic (Node Zero Inc, USA)iOS + AndroidiOS 2025-06,Play 2025-10iOS US 6,814, FR 490 ratings; Play 250k installs; running Meta adsTeaches regional Arabic dialects, leads with AI conversation, mostly English-speaking users, doesn't touch Berber languages or African languages
Ling(Simya Labs)iOS + Android2018Play 2.09M installs; iOS US 17k ratingsA multi-language general-learning app for 60, aimed at travelers. It has no North African dialect or heritage-language storytelling.
uTalk(EuroTalk)iOS + Android2016 (established)Play 1.14M installsA 150-language vocabulary book, with Lingala, Wolof and Riffian, but the content is outdated and not organized by household scenarios.
Lmed Taqbaylit / Lmed TamaziɣtAndroid2025Play 10k+ installs; driven by the Kabyle TikTok creator melissa_tamazight (238k followers), who has recently been promoting the Numidia app's waitlistKabyle and Amazigh only, free, with a strong community and Berber cultural movement vibe
GoDarija, Darija Lingo, DzLingo and othersiOS / web2025–2026Each has only single-digit to a few dozen ratingsSingle-dialect utilities
Duolingo / Babbel / BusuuBig-company features——Has Haitian Creole, Turkish and Vietnamese; no Kabyle, Wolof, Lingala, Kurdish, Tamil or Breton (checked by the Dialect press page against the 2026-09-30 course catalog)
Copycats: Derja / Dziria / Mghribiya (Ismael Dubuc)iOS + webDerja 2026-08-05,Dziria 2026-10-20Derja has 49 iOS FR ratings; Dziria has 12; the derja216 TikTok account has 3,943 followers, and its founder-on-camera video reached 55k viewsOne app per dialect; Dziria's description nearly copies Dialect's copy (see section 6)
Copycats: Diaspeak (Reda Sahraoui, Marseille)iOSThe store lists 2026-10-20 (its app id is smaller than Dialect's, possibly a repurposed older app)French iOS 4 ratings"Arabic, Wolof, Creole and 12 other languages, to talk with family and find your roots", positioning nearly identical to Dialect's

Verdict: single-dialect apps have existed for a while, and Morolingo has 15× as many ratings as Dialect. Dialect isn't first, but among the products we checked, it is the one that first made "heritage languages" into a category, put North African, West African, Caribbean and Asian immigrant languages into one app, and scaled it. There's no big-tech competitor: Duolingo not teaching these languages is the precondition for its survival.

4. Growth tactics: where users come from

4.1 TikTok account matrix (main engine, confirmed)

TikHub found at least 13 accounts, all with bios reading "Apprends le X avec Dialect", all linking to iOS or the website:

AccountFollowersLikesVideosNotes
@dialect.app (main account)8,47485k4582026-02-16 first post, 165k views
@peul.dialect17,87080k81From 2026-09-24; 81 videos total 850k views, top single video 255k (11-08)
@tunisien.dialect3,33025k130"Do you really speak Tunisian or are you faking it?" series, 150k per video (11-04)
@chleuh.dialect3,80720k94
@kabyle.dialect3,46416k89
@chaoui.dialect2,19612k106
@wolof.dialect2,00014k120
@dialect.fr1,87819k247
@lingala.dialect8164,48692
@dialect.us3075,950258English, aimed at the US
@dialect.darija / @dialect.kabyleAround 40—Each 44Short stories on history and culture
@eldrine.dialect7967,2074Paid creators: "Collaboration commerciale avec Dialect", "Learn Kabyle from zero in 30 days", from 2026-11-09

Dialect's monthly rhythm on the main account @dialect.app (based on the 450 posts pulled from TikHub):

  • 2–July: about 10–30 posts a month, carried by a few hits: 02-16 launch at 165k, 04-15 "Dialect est enfin là!" at 703k, 08-22 Algerian Darija vocabulary at 647k.
  • From September they switched to industrial volume, with more than 100 posts a month.
  • Starting late September they also spun up language sub-accounts, each posting 2–3 times a day.

Nearly all the content is templated interactive quizzes: "Kabyles, which words do you still use? Others, comment how your hometown says it", "8 questions, how many can he get right?", "Kabyle or Chaouia, which one did your family pass down?". The format is built on the regional identity and friendly rivalry inside North and West African diaspora communities, and the comment section is the distribution engine.

The comments confirm this channel: "I came across Dialect on Instagram" (2026-08-04), "Saw an Instagram video and downloaded it" (2026-10-04). Source: appstore.review FR.

4.2 Meta (Facebook / Instagram) paid ads (confirmed)

Searching "sur Dialect" (FR) in the Meta ad library returns about 170 results. Dialect's own page ("DIalect.app", id 61591940249501) ran in three phases:

  • 2026-06-29 to 08-24: boosting TikTok / IG hits as-is ("Dialect.app - 1ere application qui te permettra d'apprendre la langue de tes parents", the same creative in 6 versions).
  • 2026-08-24 to 10-04: new creatives, such as "Dialect est enfin là!" and "C'est pour ça que le Darija Marocain est sur Dialect!".
  • From 2026-10-03: ads split by language, with a shared template: "Le kabyle / wolof / lingala / créole guadeloupéen / darija algérien / rifain sur Dialect : de nouveaux mots chaque jour… prononciation de locuteurs natifs. 5 minutes…", plus a lock-screen widget creative showing "one word a day" (10-12 to 10-19). 11-06 then ran another batch (Lingala, Moroccan Darija), right when the iOS FR Education free chart hit #8.
  • The ad spend is modest: roughly a dozen running at any one time. Our metaads and tiktokads libraries don't track this product, so spend is unknown. A user comment says "Les pubs nous promettent des mots prononcés par des natifs" (2026-10-26), which also points to ads.

4.3 SEO / web version (confirmed; impact not verified)

The mydialect.io sitemap has 325 URLs: a landing page per language (apprendre-le-kabyle, learn-wolof…), plus guides like "best app to learn X", "how to learn X", and "hello in 15 languages", in 7 language versions (French, English, Spanish, German, Italian, Dutch, Chinese). The press and about pages were both updated 2026-11-09, which suggests programmatic SEO pages generated in bulk recently. On Jina, searches for "best app to learn kabyle" and "learn Moroccan darija app" already return its own guide pages near the top. mydialect.io is not in the top 1M of Tranco (traffic library), so SEO traffic is likely still small.

4.4 Media and PR (very little)

  • Only a Kapitalis 2026-10-06 article (Tunisian outlet, short piece with a founder quote) and a mention in a how-to piece by Zoom Algérie 2026-07-29.
  • The press page lists a "coverage calendar" (Yennayer, International Mother Language Day, Newroz, World Creole Day), which shows they're actively pitching.
  • x-radar searches for "mydialect" and "Dialect app kabyle" return 90 mentions in the last 0 days. On Reddit, only its own guide pages show up as indexed.

4.5 Incentivized reviews (suspected, not confirmed)

App Store FR reviews show two clusters of "all 5 stars, same day, same version": 2026-04-14 (v1.2.0,7, 2026-07-07 reviews) and 1.3.0,8 (v04-14, 700 reviews). 08-02 was the day before the video "Dialect est enfin là!" (09-21K views) went out, which suggests an update notice went to the community asking people to rate. Some reviews also name and thank "Nabil et Ahcène" (10-22), so the founders are clearly close to early users.

5. Why it took off

  1. Real pain, strong emotion: "The whole family sits around the table talking, and you can only smile along because you can't answer." Reviews keep repeating "looked for this for years", "I can finally talk to my grandmother", "reconnected with my roots" (appstore.review FR 2026-04 to 11). Willingness to pay comes from identity, not just learning.
  2. A concentrated audience the big platforms overlook: French people of North African, West African, and overseas-territory descent make up a large group, and Duolingo doesn't teach these languages. Competitors run one app per dialect, so search is fragmented. Dialect captures the demand with "one app for all" (for example, families where one parent is Kabyle and the other Arab, and some reviewers study Hilal Arabic and Darija at the same time).
  3. TikTok account matrix plus interactive quiz templates turns regional rivalry into distribution: one account per community, each with its own comment section. Over a month after the sub-accounts launched in 2026-09, @peul.dialect alone reached 850k views.
  4. Small Meta ad buys amplify the organic creatives that work; from October they targeted by language, in step with the sub-accounts.
  5. Confident subscription pricing: the free tier is small (1 card sets a day). Bad reviews complain about the price and too little free content, but top-chart rankings show conversion holds up.

6. Differentiation, and whether that's why it broke out

  • Differentiation: (1) one app covering all "heritage languages", positioned as a category rather than one app per dialect; (2) each language states openly whether the voice is a real person or synthetic, with native speakers recording the main languages; (3) every word is a family-setting word (greetings, the dinner table, relatives), with the slogan "speak like your mom".
  • Is this the reason? Positioning is a precondition for scale (it lets one brand run TikTok accounts for a dozen communities at once), but the real growth driver is the distribution playbook. The product itself is Duolingo-style cards with spaced repetition, with no technical moat, and the content quality gets criticized: the Tunisian, Malagasy, and Lingala voices sound like AI, and courses are shallow enough to finish in one evening (FR bad reviews 07-13, 08-06, 08-29, 10-26).
  • Already being copied: Ismael Dubuc's Dziria bio (2026-10-20) reads "Tu comprends quand ta famille parle, mais tu ne sais pas répondre ? … 5 MINUTES PAR JOUR … répétition espacée", almost the same copy as Dialect. He also runs one app per dialect (Derja, Dziria, Mghribiya) and posts TikToks with the founder on camera. Diaspeak (2026-10-20) uses "Arabic, Wolof, Creole, to talk with your family" as its subtitle outright.

7. What to borrow

  • The "heritage language" audience transfers: Turkish and Kurdish Germans, Moroccan Dutch and Belgians, Urdu, Bengali, and Punjabi British, and Tagalog, Vietnamese, Haitian Creole, and Cantonese second-generation Americans. Dialect has already opened @dialect.us and a German SEO page, but hasn't taken off yet.
  • One TikTok account per language, each using templates like "How does your hometown say it?" or "Do you really speak it, or are you faking?" to pull comments. Copying the same template across each account keeps costs low.
  • Meta first boosts organic hits, then splits spend by language, with creatives built on "X is now on Dialect" plus a lock-screen widget.
  • Programmatic SEO: for each language, a set of "best app to learn X" and "how to say hello in X" pages, in several interface-language versions.
  • Transparency as a selling point: each language states where its recordings come from, turning a content weakness into trust.

8. Still unclear

  • Real iOS downloads and revenue. The FR Education top-grossing chart ranks #25–47, which confirms subscription revenue is real, but monthly revenue can't be estimated from that. The "50k learners" figure is self-reported and includes web users and free users.
  • Meta and TikTok spend. Our ad libraries don't track it. The TikTok ad library (Europe) shows no ads from Dialect itself, only creator partnerships (@eldrine.dialect).
  • Whether the Meta ads for "Les Langues de Chez Nous / AKILANG" (2026-04) are connected to Dialect is unverified.
  • Whether Ahcène Bensalama is the same-named investor from XAnge (if so, that would add VC network resources and growth know-how) is unverified.
  • Whether the two clusters of 04-14-star reviews, 07-07 and 5, are natural update reminders or organized review solicitation can't be told apart.
  • Retention: bad reviews say "finished the whole course in one night" and "review is repetitive", and annual-subscription renewal rates could be a hidden risk.

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