Language learning appMar 3 to May 31, 2026
31M views from twelve accounts and one stubborn format
A language learning app wanted to reach Gen Z without paying influencer rates. We found one repeatable format, ran it across 12 learner personas and delivered 31M views in three months.
- Client
- Language learning app
- Accounts
- 12
- Platforms
- TikTok, Instagram
- Engagement
- 3 months, fully managed
Impact at a glance
About the client
A mobile app that teaches Spanish, French and Japanese through short daily conversations. Most of its users are 18 to 30 and found the app through App Store search.
The product team was strong. The marketing team was one person.
The challenge
Growth had flattened at the limits of App Store search, and paid installs were costing more than $4 each. Two influencer campaigns had each produced a single spike and then disappeared.
They wanted a steady source of new users from social, without adding headcount or managing creators themselves.
How we ran it
We spent the first two weeks testing six formats across four accounts, then committed all twelve accounts to the one that clearly won. The client's marketer joined a weekly review; everything else ran through our team.
- 01
One format, many languages
"Getting it wrong in public": a learner tries a phrase in an everyday situation, gets it slightly wrong and is corrected on the spot. Same structure, endless variations.
- 02
Learners, not teachers
Every persona was a learner documenting their own progress, so the mistakes felt relatable rather than embarrassing.
- 03
Comments become the next post
We answered the best comments with new videos, turning single hits into short series that kept people coming back.
The posts that carried it
Three of the 4 posts that passed 1M views.
Ordering coffee in Madrid went very wrong.
301K likes · 5.4K commentsDay 90 of Japanese: can I understand the train announcement?
176K likes · 2.9K commentsMy French roommate grades my accent, no mercy.
148K likes · 3.8K commentsResults
- 31.0M total views, Mar 3 to May 31, 2026
- 4 posts over 1M views
- 3.1x more app downloads attributed to organic social
Weekly views
4 accounts testing formats, then 12 on the winner
View data table
| Week | Views |
|---|
What we learned
- 01
A fixed format makes volume easy. Once the structure was set, every account could post daily without the quality slipping.
- 02
Mistakes beat mastery. Posts where the learner got it wrong outperformed polished mini-lessons by about 3 to 1.
- 03
Replies extend a hit. Answering top comments with new videos kept each breakout going for roughly two more weeks.