Ethical boundaries in habit forming design skill

Distinguishing ethical engagement from dark patterns, protecting vulnerable users, and building products that genuinely improve lives.

by wondelai·MIT license·★ 2,235 Stars on the repo·GitHub ↗

Use now

Files of Ethical boundaries in habit forming design

wondelai/main1 file
ethical-boundaries.md
Show the full text333 lines

Ethical Boundaries in Habit-Forming Design

Distinguishing ethical engagement from dark patterns, protecting vulnerable users, and building products that genuinely improve lives.

Table of Contents

  1. The Ethics Question
  2. The Manipulation Matrix Revisited
  3. Dark Patterns vs. Ethical Engagement
  4. Protecting Vulnerable Users
  5. Questions to Ask Your Team
  6. Ethical Design Practices
  7. Regulatory Landscape
  8. Case Studies: Getting It Right
  9. The Long-Term View
  10. Ethical Design Checklist

The Ethics Question

Every habit-forming product must answer: "Does this make users' lives better?"

The Hook Model is a tool. Like any tool, it can build or destroy. The responsibility lies with the creator.


The Manipulation Matrix Revisited

Nir Eyal's framework for evaluating your product:

Maker Uses It Maker Doesn't Use
Materially Improves Life Facilitator ✓ Peddler ⚠️
Doesn't Improve Life Entertainer Dealer ✗
Facilitator (Ideal)

You use the product AND it makes lives better.

Examples: Duolingo (language learning), fitness apps, productivity tools

Characteristics:

  • Founder would genuinely recommend to family
  • Users report their lives are better
  • Engagement correlates with positive outcomes
Peddler (Caution)

Product improves lives, but maker doesn't use it.

Examples: Baby products (if you don't have kids), industry-specific tools

Risk: You may not deeply understand user needs.

Mitigation: Extensive user research, advisory users, humility about assumptions.

Entertainer (Acceptable)

Product doesn't necessarily improve life, but maker uses it.

Examples: Games, entertainment apps

Key question: Is it honest about what it is?

Acceptable when:

  • Users know they're being entertained
  • Time spent is a conscious choice
  • No deception about value
Dealer (Avoid)

Product doesn't improve life AND maker doesn't use it.

Examples: Exploitative mobile games, clickbait apps

Red flags:

  • You wouldn't want your kids using it
  • Business model depends on user regret
  • Success metrics conflict with user wellbeing

Dark Patterns vs. Ethical Engagement

What Separates Them
Dark Pattern Ethical Engagement
Tricks users into actions Motivates users toward their goals
Hides true costs/consequences Transparent about tradeoffs
Exploits cognitive weaknesses Works with human psychology
Benefits only the company Creates mutual value
Users feel regret after Users feel satisfied after
Specific Comparisons

Notifications:

  • Dark: "You're missing out!" (FOMO manipulation)
  • Ethical: "Your friend replied to your comment" (genuine value)

Variable rewards:

  • Dark: Slot machine mechanics in non-gambling apps
  • Ethical: Personalized content that matches interests

Urgency:

  • Dark: Fake countdown timers that reset
  • Ethical: Genuine deadlines with clear reasons

Investment:

  • Dark: Making it hard to leave/delete data
  • Ethical: Building value users want to keep

Social proof:

  • Dark: Fake reviews, inflated numbers
  • Ethical: Genuine testimonials, real metrics

Protecting Vulnerable Users

Who Is Vulnerable?
Group Vulnerability Special Considerations
Children Developing brains, less impulse control Parental controls, time limits, age-appropriate content
Teens Identity formation, peer pressure Mental health considerations, privacy protection
Those with addiction history Susceptible to compulsive patterns Avoid gambling-like mechanics
People in crisis Emotional vulnerability Don't exploit negative emotional states
Elderly May not understand manipulation Clear, honest interfaces
Design Safeguards

For children:

  • No manipulation through social pressure
  • No in-app purchases without parental consent
  • Time limit recommendations
  • Age-appropriate rewards (not gambling mechanics)

For addiction-prone mechanics:

  • Usage dashboards showing time spent
  • Built-in break reminders
  • Easy-to-find "take a break" features
  • No punishment for not engaging

For everyone:

  • Clear paths to disengage
  • Honest representation of value
  • No exploitation of negative emotions
  • Respect for attention and time

Questions to Ask Your Team

Before Building
  1. "Would we want our children/parents using this product this way?"
  2. "Does our business model require users to develop unhealthy habits?"
  3. "What happens if users engage exactly as we hope they will?"
  4. "Are we proud of what we're building?"
During Development
  1. "Are we testing for addictive potential?"
  2. "Do our success metrics align with user wellbeing?"
  3. "Are we designing ways for users to moderate their own use?"
  4. "Would users thank us if they knew all our engagement tactics?"
After Launch
  1. "Do users report their lives are better?"
  2. "Are we seeing concerning usage patterns?"
  3. "Would we be comfortable with media coverage of our tactics?"
  4. "Are we solving problems or creating them?"

Ethical Design Practices

Transparency

Tell users what you're doing:

  • "We use notifications to help you stay on track with your goals"
  • "Our feed shows content we think you'll enjoy based on your activity"
  • "We track [X] to improve your experience"

Provide control:

  • Notification preferences with meaningful options
  • Usage dashboards
  • Easy opt-outs
  • Data portability
Alignment of Incentives

Good alignment:

  • Subscription model where engagement = value delivered
  • Success-based pricing
  • User outcomes as company OKRs

Poor alignment:

  • Ad model where more time = more money (regardless of value)
  • Engagement metrics divorced from user benefit
  • Growth at any cost
Building in Guardrails

Time-based:

  • "You've been scrolling for 30 minutes. Take a break?"
  • Daily usage limits (opt-in or default)
  • Bedtime reminders

Behavior-based:

  • Slow down compulsive actions (confirmation before purchase)
  • Friction for potentially regrettable actions
  • Cool-off periods for significant decisions

Self-selected:

  • Let users set their own limits
  • Provide tools for self-monitoring
  • Support users who want to reduce usage

Regulatory Landscape

Current Regulations

Children (COPPA, GDPR-K):

  • Parental consent for under-13 data collection
  • Restrictions on marketing to children
  • Special protections for children's data

Dark patterns (FTC, EU):

  • False urgency is deceptive advertising
  • Hidden fees violate consumer protection
  • Difficult cancellation processes under scrutiny

Addiction (emerging):

  • Some jurisdictions regulating "addictive" app features
  • Loot box regulations in games
  • Proposals for "attention protection" laws
Where It's Heading

Expect increased regulation of:

  • Notification practices
  • Variable reward mechanics
  • Design for children/teens
  • Transparency requirements
  • Right to disengage

Proactive approach: Build ethical practices now before they're mandated.


Case Studies: Getting It Right

Apple Screen Time

Apple built tools that potentially reduce engagement with their ecosystem:

  • Daily/weekly usage reports
  • App limits
  • Downtime scheduling
  • Content restrictions

Why: Long-term trust > short-term engagement

Instagram (Partial)

Implemented some ethical features:

  • Time spent dashboard
  • Notification controls
  • Like count hiding (option)

Mixed: Still heavily optimized for engagement, but added user controls.

Duolingo

Balances habit formation with learning:

  • Streak mechanics motivate consistent practice
  • Lessons designed for genuine skill building
  • Notifications timed for optimal learning

Works because: User goals (learn language) align with engagement (daily practice).


The Long-Term View

Business Case for Ethics
Short-term Sacrifice Long-term Gain
Lower initial engagement Higher lifetime value
Slower growth Better retention
More user controls Stronger brand trust
Transparent practices Lower regulatory risk
The Reputation Test

Ask: "If a journalist wrote about our engagement tactics tomorrow, would we be proud?"

If the answer is no, reconsider the tactic.

Building for Humans

The goal of the Hook Model isn't addiction—it's helping people form habits that serve them.

Ethical products:

  • Help users achieve their goals
  • Respect time and attention
  • Build genuine value
  • Enable healthy use patterns
  • Can be proud of what they've built

The ultimate measure: Does the world have more happy, fulfilled people because your product exists?


Ethical Design Checklist

Before shipping any engagement feature:

  • Does this serve the user's stated goals?
  • Would we be comfortable explaining this tactic publicly?
  • Have we built in user controls?
  • Have we considered vulnerable populations?
  • Does user success align with our success?
  • Can users easily disengage if they choose?
  • Would we want this feature in a product we use?
  • Are we optimizing for value or just time spent?
1# Ethical Boundaries in Habit-Forming Design
2 
3Distinguishing ethical engagement from dark patterns, protecting vulnerable users, and building products that genuinely improve lives.
4 
5 
6## Table of Contents
71. [The Ethics Question](#the-ethics-question)
82. [The Manipulation Matrix Revisited](#the-manipulation-matrix-revisited)
93. [Dark Patterns vs. Ethical Engagement](#dark-patterns-vs-ethical-engagement)
104. [Protecting Vulnerable Users](#protecting-vulnerable-users)
115. [Questions to Ask Your Team](#questions-to-ask-your-team)
126. [Ethical Design Practices](#ethical-design-practices)
137. [Regulatory Landscape](#regulatory-landscape)
148. [Case Studies: Getting It Right](#case-studies-getting-it-right)
159. [The Long-Term View](#the-long-term-view)
1610. [Ethical Design Checklist](#ethical-design-checklist)
17 
18---
19 
20## The Ethics Question
21 
22Every habit-forming product must answer: **"Does this make users' lives better?"**
23 
24The Hook Model is a tool. Like any tool, it can build or destroy. The responsibility lies with the creator.
25 
26---
27 
28## The Manipulation Matrix Revisited
29 
30Nir Eyal's framework for evaluating your product:
31 
32| | **Maker Uses It** | **Maker Doesn't Use** |
33|--|-------------------|----------------------|
34| **Materially Improves Life** | **Facilitator** ✓ | **Peddler** ⚠️ |
35| **Doesn't Improve Life** | **Entertainer** | **Dealer** ✗ |
36 
37### Facilitator (Ideal)
38 
39You use the product AND it makes lives better.
40 
41**Examples:** Duolingo (language learning), fitness apps, productivity tools
42 
43**Characteristics:**
44- Founder would genuinely recommend to family
45- Users report their lives are better
46- Engagement correlates with positive outcomes
47 
48### Peddler (Caution)
49 
50Product improves lives, but maker doesn't use it.
51 
52**Examples:** Baby products (if you don't have kids), industry-specific tools
53 
54**Risk:** You may not deeply understand user needs.
55 
56**Mitigation:** Extensive user research, advisory users, humility about assumptions.
57 
58### Entertainer (Acceptable)
59 
60Product doesn't necessarily improve life, but maker uses it.
61 
62**Examples:** Games, entertainment apps
63 
64**Key question:** Is it honest about what it is?
65 
66**Acceptable when:**
67- Users know they're being entertained
68- Time spent is a conscious choice
69- No deception about value
70 
71### Dealer (Avoid)
72 
73Product doesn't improve life AND maker doesn't use it.
74 
75**Examples:** Exploitative mobile games, clickbait apps
76 
77**Red flags:**
78- You wouldn't want your kids using it
79- Business model depends on user regret
80- Success metrics conflict with user wellbeing
81 
82---
83 
84## Dark Patterns vs. Ethical Engagement
85 
86### What Separates Them
87 
88| Dark Pattern | Ethical Engagement |
89|--------------|-------------------|
90| Tricks users into actions | Motivates users toward their goals |
91| Hides true costs/consequences | Transparent about tradeoffs |
92| Exploits cognitive weaknesses | Works with human psychology |
93| Benefits only the company | Creates mutual value |
94| Users feel regret after | Users feel satisfied after |
95 
96### Specific Comparisons
97 
98**Notifications:**
99- Dark: "You're missing out!" (FOMO manipulation)
100- Ethical: "Your friend replied to your comment" (genuine value)
101 
102**Variable rewards:**
103- Dark: Slot machine mechanics in non-gambling apps
104- Ethical: Personalized content that matches interests
105 
106**Urgency:**
107- Dark: Fake countdown timers that reset
108- Ethical: Genuine deadlines with clear reasons
109 
110**Investment:**
111- Dark: Making it hard to leave/delete data
112- Ethical: Building value users want to keep
113 
114**Social proof:**
115- Dark: Fake reviews, inflated numbers
116- Ethical: Genuine testimonials, real metrics
117 
118---
119 
120## Protecting Vulnerable Users
121 
122### Who Is Vulnerable?
123 
124| Group | Vulnerability | Special Considerations |
125|-------|---------------|----------------------|
126| Children | Developing brains, less impulse control | Parental controls, time limits, age-appropriate content |
127| Teens | Identity formation, peer pressure | Mental health considerations, privacy protection |
128| Those with addiction history | Susceptible to compulsive patterns | Avoid gambling-like mechanics |
129| People in crisis | Emotional vulnerability | Don't exploit negative emotional states |
130| Elderly | May not understand manipulation | Clear, honest interfaces |
131 
132### Design Safeguards
133 
134**For children:**
135- No manipulation through social pressure
136- No in-app purchases without parental consent
137- Time limit recommendations
138- Age-appropriate rewards (not gambling mechanics)
139 
140**For addiction-prone mechanics:**
141- Usage dashboards showing time spent
142- Built-in break reminders
143- Easy-to-find "take a break" features
144- No punishment for not engaging
145 
146**For everyone:**
147- Clear paths to disengage
148- Honest representation of value
149- No exploitation of negative emotions
150- Respect for attention and time
151 
152---
153 
154## Questions to Ask Your Team
155 
156### Before Building
157 
1581. "Would we want our children/parents using this product this way?"
1592. "Does our business model require users to develop unhealthy habits?"
1603. "What happens if users engage exactly as we hope they will?"
1614. "Are we proud of what we're building?"
162 
163### During Development
164 
1651. "Are we testing for addictive potential?"
1662. "Do our success metrics align with user wellbeing?"
1673. "Are we designing ways for users to moderate their own use?"
1684. "Would users thank us if they knew all our engagement tactics?"
169 
170### After Launch
171 
1721. "Do users report their lives are better?"
1732. "Are we seeing concerning usage patterns?"
1743. "Would we be comfortable with media coverage of our tactics?"
1754. "Are we solving problems or creating them?"
176 
177---
178 
179## Ethical Design Practices
180 
181### Transparency
182 
183**Tell users what you're doing:**
184- "We use notifications to help you stay on track with your goals"
185- "Our feed shows content we think you'll enjoy based on your activity"
186- "We track [X] to improve your experience"
187 
188**Provide control:**
189- Notification preferences with meaningful options
190- Usage dashboards
191- Easy opt-outs
192- Data portability
193 
194### Alignment of Incentives
195 
196**Good alignment:**
197- Subscription model where engagement = value delivered
198- Success-based pricing
199- User outcomes as company OKRs
200 
201**Poor alignment:**
202- Ad model where more time = more money (regardless of value)
203- Engagement metrics divorced from user benefit
204- Growth at any cost
205 
206### Building in Guardrails
207 
208**Time-based:**
209- "You've been scrolling for 30 minutes. Take a break?"
210- Daily usage limits (opt-in or default)
211- Bedtime reminders
212 
213**Behavior-based:**
214- Slow down compulsive actions (confirmation before purchase)
215- Friction for potentially regrettable actions
216- Cool-off periods for significant decisions
217 
218**Self-selected:**
219- Let users set their own limits
220- Provide tools for self-monitoring
221- Support users who want to reduce usage
222 
223---
224 
225## Regulatory Landscape
226 
227### Current Regulations
228 
229**Children (COPPA, GDPR-K):**
230- Parental consent for under-13 data collection
231- Restrictions on marketing to children
232- Special protections for children's data
233 
234**Dark patterns (FTC, EU):**
235- False urgency is deceptive advertising
236- Hidden fees violate consumer protection
237- Difficult cancellation processes under scrutiny
238 
239**Addiction (emerging):**
240- Some jurisdictions regulating "addictive" app features
241- Loot box regulations in games
242- Proposals for "attention protection" laws
243 
244### Where It's Heading
245 
246Expect increased regulation of:
247- Notification practices
248- Variable reward mechanics
249- Design for children/teens
250- Transparency requirements
251- Right to disengage
252 
253**Proactive approach:** Build ethical practices now before they're mandated.
254 
255---
256 
257## Case Studies: Getting It Right
258 
259### Apple Screen Time
260 
261Apple built tools that potentially reduce engagement with their ecosystem:
262- Daily/weekly usage reports
263- App limits
264- Downtime scheduling
265- Content restrictions
266 
267**Why:** Long-term trust > short-term engagement
268 
269### Instagram (Partial)
270 
271Implemented some ethical features:
272- Time spent dashboard
273- Notification controls
274- Like count hiding (option)
275 
276**Mixed:** Still heavily optimized for engagement, but added user controls.
277 
278### Duolingo
279 
280Balances habit formation with learning:
281- Streak mechanics motivate consistent practice
282- Lessons designed for genuine skill building
283- Notifications timed for optimal learning
284 
285**Works because:** User goals (learn language) align with engagement (daily practice).
286 
287---
288 
289## The Long-Term View
290 
291### Business Case for Ethics
292 
293| Short-term Sacrifice | Long-term Gain |
294|---------------------|----------------|
295| Lower initial engagement | Higher lifetime value |
296| Slower growth | Better retention |
297| More user controls | Stronger brand trust |
298| Transparent practices | Lower regulatory risk |
299 
300### The Reputation Test
301 
302Ask: "If a journalist wrote about our engagement tactics tomorrow, would we be proud?"
303 
304If the answer is no, reconsider the tactic.
305 
306### Building for Humans
307 
308The goal of the Hook Model isn't addiction—it's helping people form habits that serve them.
309 
310**Ethical products:**
311- Help users achieve their goals
312- Respect time and attention
313- Build genuine value
314- Enable healthy use patterns
315- Can be proud of what they've built
316 
317**The ultimate measure:** Does the world have more happy, fulfilled people because your product exists?
318 
319---
320 
321## Ethical Design Checklist
322 
323Before shipping any engagement feature:
324 
325- [ ] Does this serve the user's stated goals?
326- [ ] Would we be comfortable explaining this tactic publicly?
327- [ ] Have we built in user controls?
328- [ ] Have we considered vulnerable populations?
329- [ ] Does user success align with our success?
330- [ ] Can users easily disengage if they choose?
331- [ ] Would we want this feature in a product we use?
332- [ ] Are we optimizing for value or just time spent?
333 

Discussion