Skip to content

How did Duolingo use streak goals to improve retention?

Duolingo added a streak-goal choice immediately after users started a new streak. The team did not store or use the chosen value, but the act of choosing created commitment; Ali reported that users retained better weeks after seeing it, and a later version with no preselected option produced stronger goal selection without causing drop-off.

  • By Ali Abouelatta
  • 3 visuals
  • 1 min read
  • retention · positive friction · onboarding · habit formation
Duolingo streak-goal picker before and after a 14-day selection.
The composite shows an unselected picker with a disabled Continue button beside a selected 14-day goal with an enabled Continue button. It does not establish which state appeared first on entry.1012 × 834opens the full image

The evidence

The evidence is the before/after UI: a lightweight goal-selection screen became a commitment device even though the selected value was not operationally used.

What changed

  • Control: users saw a streak goal screen with a preselected option.
  • Experiment: users had to actively choose a goal before moving on.
  • The selected goal did not change the downstream app experience.

Why it matters

  • Positive friction can raise motivation when the user understands the goal and the effort feels meaningful.
  • The commitment moment matters even when the product does not personalize based on the selected value.

Every visual

  • Duolingo streak-goal picker with the 14-day option selected.
    The visible state has a selected 14-day goal and an enabled Continue button; the screenshot alone cannot establish whether the selection was a default or a user action.876 × 1122opens the full image
  • Duolingo unselected and selected streak-goal states.
    The composite repeats the picker state comparison; it is not a separate post-selection confirmation screen.1012 × 834opens the full image

Build from these references

Sources

  1. [1]

    Lazyweb Research: Duolingo streak goals

    The thread says the goal was not stored or referenced, users retained better for weeks, and the higher-friction follow-up did not create drop-off.

    Source

Methodology

Product-experiment write-up by Ali Abouelatta, dated on November 21, 2022. Source links are listed below.