Duolingo Is Built for the User Who Came for the Activity
Duolingo was built for users who came for the activity of language learning. In India, more new users are coming for what the language unlocks.
I haven’t seen Duolingo’s internal numbers. What follows is observation, framed by years of building learning products for users in India and globally. The pattern might be wrong. The frame still seems worth laying out.
Duolingo has invested seriously in India already. The Duolingo English Test as an outcome-anchored product. Hindi-medium courses. Regional pricing on Duolingo Super. The frame I’m laying out sits at a layer those investments don’t directly address: the engagement system in the first week of the core app, where users decide whether the daily ritual is worth coming back to.
What I keep coming back to: the design assumes a user who came for the activity. Many Indian users came for the outcome.
The Motivation Is Outside the App
Most users in India don’t open Duolingo because they want to learn a language for fun. They open it because they need to clear an interview, or get a better job, or feel less self-conscious in English-speaking professional settings. The motivation is external. The destination is outside the app.
Duolingo’s design works in the opposite direction. It rewards the activity, not the outcome. The streak is the streak. The league is the league. The XP is the XP. None of these reward systems make a claim about whether the user is getting closer to the actual thing they came for.
The same gap shows up at the lesson level. A new user needs to feel today’s lesson is moving them closer to their goal. Vocabulary lessons on food, family, or daily routines, even well-built ones, don’t tell the user they are closer to job-ready. The skill-build is happening; it just isn’t legible to the goal.
For users who fell in love with the activity (and there are many), none of this is a problem, because the activity is the destination. For users who came for an outcome, neither the rewards nor the lessons speak the language of the goal. They are earning XP and practicing how to talk about food and family. What they wanted was a sense of being closer to an interview.
The System Is Heavy Before the Habit Forms
Streaks, leagues, XP, quests, hearts. For a user who’s been on the app for months, these aren’t five systems but a single integrated reward layer that texturizes daily practice. For a user opening the app on day three, they’re five systems, each with its own logic, each with its own consequences for missing a day.
The cost of complexity is highest before a habit has formed. The user who already shows up daily can absorb new mechanics as variations on the routine. The user who’s still deciding whether to come back tomorrow has to evaluate each system on its own.
Multiple loops competing for attention is fine when attention is given. It is expensive when attention is being earned.
In India specifically, the cost of missing a day is higher. Routines are less stable. Time fragments more. Missing a day or two is normal; the streak system treats it as failure. For a user not yet hooked, the “you broke your streak” message can land as evidence that they’re not a Duolingo person.
Before the Habit Forms, the Job Is Different
None of this is to say Duolingo’s system is wrong. It is one of the most carefully built engagement systems in consumer software. It works because it solved habit formation. The system rewards consistency, builds a daily ritual, and turns the ritual into identity (I’m a streak person now).
But the system rewards consistency in users who are already showing up. Before the habit forms, a different version of the product probably has to do a different job: convince the user that the time spent today connects to the goal they had when they downloaded. Once the habit forms, the engagement loops can take over. The question is whether that earlier, simpler, more goal-anchored version is worth running for the first week or two.
Probably yes. Especially in a market where most new users came for what’s outside the app.