Commitment Devices

Canonical
WendelKahneman

Confidence

86%

Cognitive Load

Medium

Evidence

production validated

Impact

feature

Ethical Guardrail

Agents must offer commitment devices only when the user explicitly expresses a goal. Must be voluntary and transparent. Never shame users for breaking commitments.

Design Intent

Humans are great at making plans and terrible at sticking to them. A Commitment Device locks in future behavior by raising the cost of deviation. They work by leveraging present bias and loss aversion to protect future selves.

Psychology Principle

Humans are great at making plans and terrible at sticking to them when future temptations arise.

Description

Mechanisms that lock in future behavior by raising the cost of deviation, bridging the gap between today's motivation and tomorrow's action.

When to use

Any long-term goal feature where follow-through is historically low -- fitness goals, savings, learning streaks, project deadlines.

Example

Strava Monthly Challenge + Stickk-style bet: User commits to Run 100 km this month + optional $25 on the line with public leaderboard visibility.

Autonomy Compatibility

SuggestConfirm

Behavioral Objective

Users honor their pre-committed goal because the cost of deviation is now higher.

  • Increased follow-through on long-term habits and goals
  • Stronger sense of personal accountability
  • Reduced decision fatigue on repeated choices

Target Actor

role

Everyday user

environment

Future temptations, busy schedules

emotional baseline

Good intentions but frequent present-bias failure

ai familiarity

medium

risk tolerance

medium

Execution Model

1

commitment_creation

User voluntarily locks in a specific future behavior.

No commitment is made despite stated goal.

2

cost_of_deviation

Make breaking the commitment meaningfully costly.

User easily abandons without feeling any real cost.

3

visibility_and_support

Keep the commitment visible and supported throughout the journey.

Commitment is forgotten until too late.

Failure Modes

Commitment feels too rigid or punitive

Make every device optional and easily adjustable

micro

User sets unrealistic commitments and fails

Agent gently suggests realistic scope before locking in

micro

Social accountability backfires into shame

Keep public commitments opt-in and positive-only

feature

Monetary devices create stress

Cap amounts and require explicit user consent

feature

Overuse leads to commitment fatigue

Limit active commitments per user

micro

Agent Decision Protocol

Triggers

  • User expresses a clear long-term goal
  • Historical follow-through on similar goals is low
  • User asks for accountability or 'help me stick to this'

Escalation Strategy

L1: Diagnose whether a commitment device is appropriate for this goal

L2: Nudge -- offer voluntary commitment with clear opt-out

L3: Restructure -- adjust commitment scope or switch device type

L4: Constrain -- limit active commitments to prevent fatigue

L5: Yield -- flag for human behavioral designer review

Example

User wants to run 3x per week -> agent offers 'Commit publicly to your friends or put $20 on the line -- I'll track it for you.'

Behavioral KPIs

Primary

  • % of commitments successfully completed
  • Follow-through rate increase vs. non-committed users
  • Retention of habit after commitment ends

Risk

  • Commitment abandonment or reset rate
  • User complaints about pressure or shame

Trust

  • User-reported sense of empowerment from commitments
  • Autonomy Dial usage when agent offers commitment options

Behavioral Signals

commitment_avoidance

goal_stated=true AND commitment_created=false

commitment_offered=true AND commitment_declined_count > 2

commitment_breaking

commitment_active=true AND deviation_count > 3 in 7 days

commitment_reset_count > 2 in 30 days

commitment_fatigue

active_commitments > 3 AND completion_rate_declining=true

user_disables_commitment_reminders=true

Decay Monitoring

Revalidate when

  • User base changes in risk tolerance or financial situation
  • New platform features enable stronger commitment mechanics
  • Cultural attitudes toward self-binding shift

Decay signals

  • Rising commitment reset or abandonment rates
  • Drop in user-initiated commitment creation
  • Feedback that commitments feel burdensome

Pattern Relationships

Related Patterns

Canonical Implementation

Strava Monthly Challenge + Stickk-style Bet: User commits to Run 100 km this month + optional $25 on the line with public leaderboard visibility

Telemetry Hooks

commitment_createdcommitment_trackedcommitment_completedcommitment_broken

Tags

accountabilityhabit-formationagent-ready