Designers say guilt, streaks, and hooks keep users coming back better than delight. Others say some friction should stay when it helps users make better choices

Designers argue two ways about retention: engineered guilt and streaks work better than delight, or friction should stay where it builds judgment.
One essay says Duolingo's guilt and loss aversion beat delight, Coinbase and Duolingo job ads staff streaks, and Eyal's habit loop is said to fit AI assistants.
Against that, essays call obstructive cancellation flows a deliberate business decision and argue for keeping friction where it helps people choose.
About nine sources sit on the hooks side and thirteen on the friction side, so the argument is live rather than settled.

sources

+18

+14

last 30d

+11

First seen 5 June 2026

Sources saying it

20mentions, since 29 Jun

3 jobs

3 articles

6 social

Decide which side you are on before the retention review. Streaks that work and streaks users resent are the same feature.
Your retention measure should include how many people turned the reminders off.
Keep one piece of friction on purpose where the user is about to make a costly choice, and write down why.

Draw the streak reset moment. Guilt is doing the work there, and it is your call how hard it pushes.
Design one confirmation step that slows a risky action, and defend it in review with the outcome it protects.
A hook that works is still a hook. Show the team what the interface looks like from the tired user's side.