Habit design (Hooked)
Variable Reward — Self
Unpredictable mastery rewards
What it is
Rewards of the Self are sought for the intrinsic feeling of mastery, competence, and completion. Game progress, streaks, skill acquisition, leveling up — at varying intervals.
Why it works
Rewards of the Self are pursued for their own sake — the intrinsic satisfaction of competence, mastery and completion, with no other person required. Self-Determination Theory, developed by Deci and Ryan, identifies competence as one of three basic psychological needs alongside autonomy and relatedness, and shows that intrinsic motivation is far more durable than motivation driven by external rewards. Their research also carries a warning designers routinely ignore: attaching extrinsic rewards to an activity someone already enjoys can reduce their intrinsic motivation for it, an effect known as overjustification. The variability matters here as much as elsewhere. Predictable progress becomes mechanical, while unexpected evidence of one''s own improvement — a skill revealed, a comparison to a past self — lands as genuine reward.
How to apply it in design
Surface unexpected progress moments — surprise badges, occasional skill reveals, glimpses of mastery the user didn't know they had. Reward identity ("you're becoming X"), not action ("you did Y").
Where you'll see it
- Progress the user did not know was being tracked, revealed at the right moment
- Comparison against the user's own past performance rather than other people
- Difficulty that adapts so the task stays just beyond current ability
- Completion states that feel finished rather than merely stopped
- Feedback that names the capability gained, not the action performed
Example
Duolingo's variable XP and league system rewards persistence with unpredictable status moments — your league rank is the variable.
Failure mode
Linear, predictable progress bars without surprise stop being rewarding the moment the goal becomes mechanical. 80% loading is not a reward.
When not to use it
Avoid mastery mechanics where the activity is genuinely a chore. Nobody wants to level up at expense reports, and gamifying obligatory admin usually reads as condescending. Be careful too with intrinsically motivated behaviour: adding points and badges to something the user already enjoys can crowd out their own reason for doing it, which is the overjustification effect in practice. And progress systems are counterproductive in tools used sporadically, where returning after a gap to a broken streak or a decayed score punishes the user for having a life.
The ethical line
Progress must reflect something real. Vanity metrics that inflate to make users feel accomplished — arbitrary point totals, meaningless levels, completion bars for actions of no consequence — are flattery, and users work this out faster than product teams expect. The more damaging pattern is the manufactured loss: streaks engineered so that breaking them feels like failure, which converts an intrinsic reward into anxiety and turns a tool for building capability into an obligation. If the user would feel guilty rather than disappointed on missing a day, the mechanic has been tuned past the point of serving them.
What to measure
Track measured capability rather than activity — accuracy, speed or quality on comparable tasks over time — and check that the product''s own progress display correlates with it. If users level up without getting better, the system is rewarding attendance.
Frequently asked questions
What is a variable reward of the self?
It is Nir Eyal's category for intrinsic rewards sought for the feeling of mastery, competence or completion, with no social component. Progress, skill acquisition and levelling up all belong here, and the reward is strongest when the evidence of improvement arrives unexpectedly rather than on a predictable schedule.
Why do progress bars stop being motivating?
Because a predictable increment carries no new information. Once the user can forecast exactly what the next action will add, completing it becomes mechanical rather than rewarding. Unexpected reveals — a skill they did not know they had built, a comparison to their own earlier work — restore the variability the mechanism depends on.
What is the overjustification effect?
It is the finding from Deci and Ryan's research that adding extrinsic rewards to an activity someone already enjoys can reduce their intrinsic motivation for it. In products it means that gamifying a genuinely enjoyable task can make it feel like work, and removing the points later leaves less motivation than existed before.
Are streaks a good design pattern?
They work when they support a habit the user already wants and forgive normal life, and they fail when breaking one feels like failure. A streak that produces guilt has converted an intrinsic reward into an obligation, which is the opposite of what mastery rewards are for.
Related skills
Source: Eyal — Hooked (Ch. 4)
Practise Variable Reward — Self on a real brief
Quest gives you a client brief that targets this skill and 60 minutes to ship an iteration.