Abstract
This study analyzes how the principles of persuasive psychology are applied in digital product design, distinguishing strategies that promote virtuous habits from those that foster compulsive behaviors. Using a qualitative and exploratory approach grounded in a literature review of Kahneman, Norman, Eyal, Fogg, Dutra and Warren, a comparative case study of four applications — Duolingo, Apple Health, TikTok and Temu — was conducted through the lens of the Hook Model and the Manipulation Matrix. The results show that trigger, action, variable reward and investment are ethically neutral mechanisms, whose impact on user autonomy is determined by the designer's intentional calibration. Duolingo and Apple Health emerge as Facilitators, orienting engagement toward real, lasting benefit; TikTok is classified as a Dealer for actively suppressing the user's decision-making capacity; and Temu is reclassified as a Peddler — not a Dealer — for delivering the promised product, even as it amplifies consumption through artificial urgency and scarcity. The study concludes that the designer's ethical responsibility lies in orienting these mechanisms, and proposes practical guidelines grounded in the Regret Test and the Manipulation Matrix to support more conscious design decisions in the Brazilian context.
Keywords: persuasive psychology; behavioral design; Hook Model; cognitive biases; design ethics.
Introduction
In the contemporary moment, digital devices have stopped being mere tools and become extensions of human routine, shaping behaviors and identities through repetitive practices and daily rituals. At the center of this interaction is experience design (UX/UI), which uses principles of persuasive psychology to engage the user spontaneously. However, mechanisms like cognitive biases and variable rewards are frequently calibrated in opposite directions: while some products aim to facilitate virtuous habits, others are designed to foster compulsive behavior and dependency.
The central problem of this research lies in understanding how these same psychological triggers are adjusted in product design to generate such distinct outcomes. On one hand, engagement engineering can serve as a "painkiller" for real pain points, promoting health and education; on the other, it can act as a "vitamin" that appeals to fleeting emotional needs, leading to online gambling addiction or excessive consumption. Although consolidated habit-formation models exist, such as Nir Eyal's Hook Model and Fogg's Behavior Model (2019), the literature shows a gap regarding a deep comparative analysis of the ethical calibration of these mechanisms in today's Brazilian digital context.
The justification for this study rests on designers' ethical responsibility: as "facilitators," they must decide not only whether they can "hook" users, but whether they should do so in service of human well-being. The social relevance of the research is heightened by the rise of disorders such as ADHD and digital addictions that demand new professional postures, such as the concept of HX Design, which prioritizes the human being beyond the screen (DUTRA, 2020).
The general objective of this paper is to conduct a comparative analysis of how the principles of persuasive psychology are applied in product design, distinguishing strategies that promote autonomy from those that foster manipulation. To this end, the methodology adopted is a qualitative and exploratory study grounded in a literature and documentary review of authors such as Kahneman, Norman, Eyal and Dutra.
Literature review
Understanding contemporary digital design requires a deep analysis of how technology shapes human behavior through deliberately engineered psychological mechanisms. This section explores the theoretical foundations underlying the duality between healthy habits and compulsive behaviors in the digital environment.
The duality of thought and cognitive biases
Information processing by the user does not happen in a purely rational way. According to Norman (2004), the emotional design perspective holds that the human mind operates on three levels: visceral, behavioral and reflective. While the visceral level responds immediately to aesthetics and sensory stimuli, the reflective level is responsible for interpreting long-term values and meaning. Interfaces oriented toward digital addiction frequently overload the visceral level to bypass the reflective level's critical analysis, facilitating automatic decisions.
In this scenario, cognitive biases act as mental shortcuts that influence choice architecture. As Dutra (2020) observes, small adjustments in how information is presented can lead the user toward behaviors anticipated by the designer, exploiting automatic brain tendencies that precede conscious reflection.
Engagement engineering and the Hook Model
The transformation of a product into a routine habit is explained by Eyal (2014) through the Hook Model, made up of four stages: trigger, action, variable reward and investment. The ultimate goal is spontaneous engagement, getting the user to return to the product repeatedly without the need for aggressive advertising.
A critical element in this cycle is variable reward. Research indicates that dopamine levels rise significantly when the brain anticipates an uncertain reward, suppressing the areas associated with judgment and reason while activating the areas linked to wanting and desire. Eyal distinguishes products that act as "painkillers," solving real pain points and needs, from products that act as "vitamins," appealing to fleeting and potentially addictive emotional needs.
The anthropology of habit and identity
Beyond the technical mechanics, habits carry a central philosophical and anthropological dimension. According to Warren (2016), most of human life is driven by daily rituals and repetitions that — whether the individual is aware of it or not — shape their identity and desires. What Warren calls the "liturgies" of the ordinary suggests that repetitive practices in digital interfaces are not neutral; they instill in the user a particular vision of the "good life," defining what they love and worship. The ethical calibration of design thus determines whether these "digital liturgies" lead to autonomy or dependency.
Neurodesign and cognitive duality (Systems 1 and 2)
Neurodesign is defined as the application of neuroscience and psychology to create more effective designs, seeking to understand how the brain responds to visual stimuli and decision-making (BRIDGER, 2015). Central to this understanding is dual-process theory: System 1 operates unconsciously, fast and intuitively, relying on patterns and rules of thumb to generate impulsive reactions. System 2, by contrast, is slow, logical and requires deliberate mental effort, and is frequently avoided by the brain to conserve energy.
In digital design, the effectiveness of engagement lies in the ability to appeal to System 1. Interfaces oriented toward addiction frequently exploit processing fluency — the brain's bias toward simple, easily decoded images — to bypass System 2's critical analysis. Through a visual "salience map," the designer can direct the user's attention to elements that activate the reward system before any conscious reflection even occurs (BRIDGER, 2015).
Habit engineering and the dopamine system
Habit formation occurs when the product becomes the "first thing that comes to mind" solution for an internal discomfort, such as boredom or loneliness. The Hook Model uses variable reward to exploit the brain's dopaminergic system. Neurobiology studies indicate the nucleus accumbens is more active in anticipation of reward than upon receiving it. Reward variability creates a focused state of desire that suppresses areas associated with judgment and reason, strengthening the user's cycle of spontaneous return.
Ethics in design: the Manipulation Matrix
Given the power of these psychological mechanisms, Eyal (2014) proposes the Manipulation Matrix as an ethical decision-support tool for designers and entrepreneurs. The matrix is organized around two fundamental questions: "Would I use the product myself?" and "Does the product materially improve the user's life?"
Based on these answers, creators are classified into four quadrants:
Facilitator. The designer uses the product and believes it improves the user's life. This is considered the ethical ideal, since the creator has "skin in the game" and understands the consequences of use.
Peddler. The creator believes the product helps the user but would not use it personally. There is a risk of disconnection and lack of empathy, often resulting in ineffective or merely promotional solutions.
Entertainer. The product is used by the creator, but it does not durably improve the user's life. This is ephemeral entertainment, such as casual games, that depends on constant novelty to avoid being discarded.
Dealer. The designer creates a product they would not use and that does not benefit the user. This quadrant represents pure exploitation and manipulation, aimed solely at profit through the creation of harmful addictions.
Complementing the matrix, the "Regret Test" serves as an operational metric: if the user, fully aware of the designer's intentions, would still take the action without regret, the strategy is classified as persuasion; otherwise, it constitutes coercion.
Method
This research is characterized as a qualitative and exploratory study, grounded in a literature and documentary approach to understand the calibration of psychological triggers in digital design. The method is structured in three main stages: literature review, comparative interface analysis, and ethical discussion.
Literature and documentary research
The first stage consisted of a literature review, surveying and critically analyzing foundational works on persuasive psychology, cognitive biases and behavioral design. Reference authors consulted included Kahneman, Eyal, Norman and Dutra, aiming to map engagement mechanisms and choice architecture in digital products. In addition, documentary research was conducted across technology outlets and UX/UI guidelines to contextualize current practices in the Brazilian market.
Comparative case study
The empirical phase of this research adopts a comparative case study strategy, which allows for a dense, holistic and in-depth analysis of the internal workings of the selected interfaces. The object of analysis is divided into two distinct groups, selected for their relevance in the current digital landscape and for the clarity with which they apply psychological triggers:
Group A (Virtuous Habits). Made up of Duolingo and Apple Health. Duolingo is characterized as a "Facilitator" that uses daily notifications and gamification mechanics (such as the "streak" system) to turn language learning into a positive daily routine. Apple Health uses reward anticipation through physical activity goals (such as step counts) to motivate healthy, consistent behaviors.
Group B (Vicious Habits). Made up of TikTok and Temu. TikTok is analyzed for its ability to generate compulsive engagement through an infinite feed of short videos and algorithmically personalized variable rewards, creating a retention cycle based on the "hunt reward." Temu represents high-retention digital retail, where design is focused on the constant search for deals and instant gratification, exploiting scarcity bias and the pleasure of "discovering" material goods.
Analysis procedures
Data was processed using trigger-mapping technique, identifying how elements of scarcity, social proof and variable reward are implemented within each group. The qualitative analysis was guided by pattern-matching technique, seeking to identify whether the design patterns observed in the interfaces matched the theoretical predictions of the Hook Model and the levels of emotional design. Finally, the results were checked against the Manipulation Matrix in order to categorize the ethical impact of each strategy on user autonomy.
Results
The comparative mapping of psychological triggers across the four interfaces analyzed revealed patterns consistent with the Hook Model's predictions, while also revealing fundamental ethical differences in how these mechanisms are calibrated. Results are presented by group and, at the end, synthesized in a comparative table.
Group A — Virtuous habits
Duolingo and Apple Health share an engagement architecture based on the Hook Model, but with an ethical calibration that preserves and strengthens user autonomy. In both cases, external and internal triggers are oriented toward solving real pain points — language learning and physical health, respectively —, framing them as "painkiller" products in Eyal's (2014) taxonomy. The investment the user accumulates (acquired skills, health history) represents genuine value that belongs to the user, not the platform.
Duolingo. Duolingo applies all four stages of the Hook Model with a clear orientation toward user benefit. The external trigger, embodied in the daily notification with the Duo mascot, activates the internal trigger of loss aversion around the streak — a mechanism Kahneman (2011) describes as one of the most powerful cognitive biases. Action is made easy through five-minute gamified micro-lessons, aligned with Fogg's (2019) principle that reducing friction is decisive for habit formation. Variable reward, delivered through weekly social-ranking leagues, taps into what Eyal (2014) calls the Reward of the Tribe, while accumulated investment in XP, trophies and streaks builds a positive identity around continuous learning, as noted by Warren (2016). Table 1 details this mapping.
Table 1 — Trigger mapping — Duolingo
| Hook stage | Mechanism | Concrete example | Theoretical reference | Class. |
|---|---|---|---|---|
| External trigger | Personalized daily push notification with emotional appeal | "You lost your streak! Don't leave Duo waiting." — sad owl icon at 8pm | Eyal (2014): the external trigger is the cycle's entry point; notifications reactivate a dormant habit | Facilitator |
| Internal trigger | Streak creates fear of losing continuity (loss aversion) | Consecutive-day counter on display; losing the streak hurts more than never starting one | Kahneman (2011): the pain of losing is ~2x greater than the pleasure of an equivalent gain | Facilitator |
| Action | Five-minute micro-lessons with gamified UX; minimal access barriers | Lesson starts with one tap; visible progress bar; immediate audio feedback | Fogg (2019): Behavior = Motivation × Ability × Trigger — reducing friction raises the action rate | Facilitator |
| Variable reward | Weekly leagues with social ranking; outcome uncertain until week's end | Gold League: the user only learns whether they moved up or down on Sunday night | Eyal (2014): Reward of the Tribe — the need for belonging and social recognition | Facilitator |
| Investment | Accumulated XP, trophies, streak and profile — a sunk identity cost | 300-day streak represents a built identity; quitting means losing who I am | Warren (2016): repetitive practices instill identity — what we do shapes what we love | Facilitator |
Source: prepared by the author (2025).
Apple Health. Apple Health structures its engagement mechanisms around the visualization of the three activity rings — Move, Exercise and Stand — creating a visual feedback system that operates directly on Norman's (2004) reflective level: an incomplete ring generates an identity-level tension that motivates action. The external trigger of near-goal proximity notifications exploits the near-miss effect, while haptic feedback upon closing each ring provides immediate sensory reinforcement. Surprise achievements constitute Eyal's (2014) Reward of the Hunt, rewarding the user's actual performance in a variable way. Investment in longitudinal health data represents the group's highest exit cost, since the accumulated history is genuinely valuable and non-transferable, as shown in Table 2.
Table 2 — Trigger mapping — Apple Health/Fitness
| Hook stage | Mechanism | Concrete example | Theoretical reference | Class. |
|---|---|---|---|---|
| External trigger | Daily goal alerts with a sense of near-completion | "You're 847 steps from closing your rings today!" — at 9pm with a nearly full bar | Eyal (2014): an external trigger with a sense of goal proximity increases the likelihood of action | Facilitator |
| Internal trigger | Incomplete rings create discomfort with one's self-image | Opening the app and seeing the Exercise ring in yellow creates an identity tension that motivates a walk | Norman (2004): reflective level — the user interprets the incomplete ring as inconsistent with their values | Facilitator |
| Action | Haptic feedback and animation on closing each ring; multimodal sensory reinforcement | Vibration and visual burst on completing the Move ring — immediate kinesthetic pleasure | Bridger (2015): multimodal feedback (visual + tactile) reinforces the reward circuit by activating System 1 sensory responses | Facilitator |
| Variable reward | Surprise achievements and badges for unexpected personal records | A "Best week ever" badge appears without warning — the user never knows when the next one is coming | Eyal (2014): Reward of the Hunt — reward variability keeps the dopaminergic system engaged | Facilitator |
| Investment | Longitudinal health history — personal data that is irreplaceable over years | Years of heart-rate, sleep and activity graphs create a high emotional switching cost | Eyal (2014): data investment creates an exit cost that retains the user around genuine value | Facilitator |
Source: prepared by the author (2025).
Group B — Vicious habits
TikTok and Temu both use the Hook Model for maximum retention, but diverge in their ethical classification on the Manipulation Matrix. While TikTok falls into the Dealer quadrant, Temu shows characteristics that position it more precisely as a Peddler — a distinction explored in the sections below.
TikTok. TikTok represents the most extreme case of compulsive engagement among the platforms analyzed. Its design architecture is built to systematically suppress Kahneman's (2011) System 2: the absence of a home screen, continuous autoplay and the lack of any time limit create an environment that demands active conscious effort to stop — not to continue. The internal trigger exploits real pain points such as boredom and loneliness, but does not resolve them; instead, the platform amplifies these states by creating a dependency on constant stimulation. Variable reward is operated by machine-learning algorithms that precisely calibrate the frequency of viral content injection, replicating the slot-machine mechanism described by Eyal (2014). Applying Eyal's (2014) Regret Test, the intentional design of autonomy suppression classifies TikTok in the Dealer quadrant, as shown in Table 3.
Table 3 — Trigger mapping — TikTok
| Hook stage | Mechanism | Concrete example | Theoretical reference | Class. |
|---|---|---|---|---|
| External trigger | Personalized push featuring a high-engagement-probability video selected by the algorithm | "@user posted something you'll love" — content chosen before any conscious decision | Bridger (2015): System 1 is activated before System 2 — the external trigger bypasses critical analysis | Dealer |
| Internal trigger | Product positioned as a reflex solution for any emotional discomfort (boredom, loneliness) | No home screen — the feed starts automatically; opening the app is a conditioned response to boredom | Eyal (2014): habit forms when the product becomes the "first thing that comes to mind" solution for internal pain | Dealer |
| Action | Infinite autoplay feed; intuitive vertical scroll; no pause button or time limit | No choice to make — the next video plays automatically; stopping requires active conscious effort | Norman (2004): visceral level dominates — the interface overloads sensory stimuli to suppress the reflective level | Dealer |
| Variable reward | The algorithm unexpectedly injects viral content every few videos — calculated variability | The user never knows when the "perfect" video is coming — continuous anticipation suppresses rational judgment | Kahneman (2011): the nucleus accumbens is more active in anticipation — variability creates focused, compulsive desire | Dealer |
| Investment | Content created, followers, preference history — an ecosystem that traps through switching cost | Switching platforms means losing followers, creations and the algorithm that "knows me" deeply | Eyal (2014): investment of reputation and content creates an exit cost that serves the platform, not the user | Dealer |
Source: prepared by the author (2025).
Temu. Temu's ethical classification on the Manipulation Matrix demands a more nuanced analysis than TikTok's. At first glance, mechanisms such as urgency timers, strikethrough pricing and artificial scarcity suggest manipulative design; however, the platform does effectively deliver the promised product to the user — there is a concrete material benefit in the form of low-cost products that arrive at their destination.
The crucial distinction lies in the matrix's two questions: "Does the product materially improve the user's life?" — a partially positive answer, given that the product is delivered — and "Would the designer use the product?" — an equally positive answer, since Temu's executives and designers are plausible consumers of the service themselves. This profile corresponds to the Peddler quadrant: the creator believes the product benefits the user, but the design reveals a disconnect between the stated intent and the real effects of compulsive engagement, which can lead to unplanned impulse spending.
Unlike the Dealer — who would not use the product and does not believe in its benefit — the Peddler operates in a zone of ethical ambiguity: the benefit exists, but the engagement mechanisms are calibrated to maximize consumption beyond the user's real need. Table 4 details this mapping.
Table 4 — Trigger mapping — Temu
| Hook stage | Mechanism | Concrete example | Theoretical reference | Class. |
|---|---|---|---|---|
| External trigger | Push notification with an expiring offer timer and artificially scarce stock | "Only 2 left! Offer expires in 03:42:11" — personalized countdown push notification | Kahneman (2011): scarcity bias + loss aversion — "not losing" motivates more than "gaining" | Peddler |
| Internal trigger | Purchase gamification positioned as a solution for the emotional need for discovery and bargain-hunting | A "wheel of fortune" during onboarding promises a prize — creates an expectation of winning before the first purchase | Dutra (2020): novelty bias and sunk cost explain why the user keeps investing time and money beyond real need | Peddler |
| Action | One-tap checkout with strikethrough pricing and an immediate discount in visual focus | Neon orange button; original price struck through in gray next to the discounted price — a decision made in seconds | Norman (2004): visceral level — warm color + strikethrough price activate urgency before rational cost analysis | Peddler |
| Variable reward | Product scroll as a deal hunt; coins and coupons with deliberately obscure value | "You won 47 Temu coins!" — the real-currency equivalence is opaque; accumulation hooks the search for more | Eyal (2014): Reward of the Hunt for material resources — uncertainty about real value keeps the loop active | Peddler |
| Investment | Gamified affiliate program: inviting friends generates credits and progressive rewards | "Invite 3 friends and get $50" — the user invests time, relationships and personal credibility in the platform | Eyal (2014): social investment is the most powerful; the user recruits new consumers into the system | Peddler |
Source: prepared by the author (2025).
Comparative synthesis
Table 5 synthesizes the main dimensions of analysis across the four applications, showing how the same Hook Model mechanisms produce radically different impacts on user autonomy depending on the ethical calibration adopted by the designer.
Table 5 — Comparative synthesis — ethical calibration of engagement mechanisms
| Dimension | Duolingo | Apple Health | TikTok | Temu |
|---|---|---|---|---|
| Type of pain addressed | Real learning (painkiller) | Physical health (painkiller) | Boredom/loneliness (ephemeral vitamin) | Discovery/bargain (artificially amplified) |
| User autonomy | Preserved and strengthened | Preserved and strengthened | Actively suppressed by design | Partially compromised by urgency design |
| Reward variability | Functional — reinforces a virtuous habit | Functional — rewards real performance | Exploitative — a digital slot machine | Exploitative — obscures real monetary value |
| Investment cost | Skill acquired by the user | The user's own health-data archive | Irrecoverable lifetime | User's financial resources and social network |
| Regret test | Passed — an informed user has no regrets | Passed — an informed user has no regrets | Failed — design hides its real intentions | Partial — delivers the product, but hides impulse manipulation |
| Matrix classification | Facilitator | Facilitator | Dealer | Peddler |
Source: prepared by the author (2025).
The results show that the distinction between virtuous and vicious habits does not lie in the mechanisms themselves — all four applications use triggers, variable rewards and investment — but in the ethical orientation of those mechanisms. The Facilitator aligns engagement with the user's real benefit; the Peddler delivers value but amplifies consumption beyond need; the Dealer actively suppresses autonomy to maximize retention.
Final considerations
The comparative analysis conducted in this study showed that the Hook Model's mechanisms — trigger, action, variable reward and investment — are, in essence, ethically neutral. What determines whether a product forms virtuous or vicious habits is not the existence of these mechanisms, but the intentional calibration the designer applies at each stage of the cycle. Duolingo and Apple Health use the same four stages as TikTok and Temu; what distinguishes them is the orientation of these triggers: in the former, engagement serves a real, lasting benefit for the user; in the latter, it serves the maximization of screen time and consumption, often at the expense of the user's own autonomy.
This finding confirms the study's central hypothesis and highlights a critical gap in the training of digital designers: engagement technique is widely taught and documented, but the ethical question that should precede any design decision — who does this mechanism actually work for? — rarely receives the same curricular weight. The concept of HX Design, which positions the human being as the ultimate recipient of any digital experience, offers a promising path to close this gap, by proposing that user well-being be treated as a product metric, not a secondary constraint.
The reclassification of Temu from the Dealer quadrant to the Peddler quadrant, proposed in this study through a rigorous application of Eyal's (2014) two Manipulation Matrix questions, represents a relevant analytical contribution. The distinction is not merely taxonomic: it implies that not all manipulative design is equivalent in its ethical severity. Temu delivers the promised product — there is a concrete material benefit —, but it amplifies consumption beyond real need through artificial urgency and scarcity. Recognizing this gradation is essential for criticism of persuasive design to be precise, and therefore more effective.
Based on these results, the following ethical guidelines are proposed for product design and UX professionals:
- Apply the Regret Test as a mandatory step in the design process: before implementing any engagement mechanism, the designer must be able to answer whether the user, fully informed of the product's intentions, would take the action without regret. A negative answer should be treated as a project blocker, not an acceptable risk.
- Map the pain the product claims to address, and check whether the engagement mechanism resolves it or merely exploits it. Painkiller products, which address real needs, sustain engagement over the long term; products that exploit pain without resolving it create dependency without genuine loyalty.
- Evaluate investment cost from the user's point of view: what the user deposits into the product — time, data, money, social relationships — should generate proportional, returnable value for the user, not merely a switching cost that traps them in the platform.
- Adopt the Manipulation Matrix as a review tool in product rituals such as sprint reviews and retrospectives, integrating the ethical perspective into the continuous development flow, not just the concept phase.
However, it would be naive to place the entire ethical responsibility for what gets designed on the designer alone. Most design professionals operate within corporate structures where strategic guidelines — including decisions about which engagement mechanisms to implement and at what intensity — are set by senior management, retention targets and business models established long before any brief reaches the product team. In this scenario, the designer is often an executor of decisions they did not make, which does not exempt them from responsibility, but repositions where its primary locus lies.
This structural tension is documented empirically by Marques and Falcão (2026), in a study published in UERJ's Arcos Design journal, which conducted semi-structured interviews with designers and developers working on digital betting platforms in Brazil. The results reveal that professionals recognize design's active role in inducing compulsive behavior, but report that internal pushback is systematically limited by hierarchies, performance targets and the absence of formal spaces for ethical reflection within organizations. The study concludes that the debate on design ethics needs to shift its focus from the end user to the conditions of production of interfaces — that is, to the organizational structures that shape what a designer can or cannot question.
This scenario is particularly critical in Brazil's current context, marked by the explosion of the iGaming market following the regulation of sports betting under Law No. 14,790, in December 2023. Central Bank data show that, in August 2024, Brazilians wagered approximately R$20 billion in online bets in a single month — of which R$3 billion came from beneficiaries of the Bolsa Família welfare program (BRAZILIAN CENTRAL BANK, 2024). The explosion of UX and product design jobs at iGaming companies sparked a heated debate within the Brazilian design community about the ethical limits of selling one's creative labor to companies whose business model structurally depends on exploiting cognitive biases. Platforms such as "Jogo do Tigrinho" ("The Little Tiger Game") have become symbols of this impasse: their interfaces exploit intermittent reinforcement, artificial urgency and language that equates betting with entertainment — the deliberate use of the verb "play" instead of "bet" — configuring what UX literature calls deceptive patterns (MAIS QUE VISUAL, 2024).
Given this scenario, it is proposed that the designer's ethical agency operate on two complementary levels. The first is the choice of employer: in deciding which company to sell their labor to, the designer exercises a form of power that precedes any product decision. Working for a company that falls into the Dealer quadrant — one that would not use its own product and does not believe in its benefit to the user — means putting technical and creative knowledge in service of exploitation. This choice has concrete ethical implications beyond the individual, since every qualified professional who joins a predatory ecosystem increases its capacity to cause harm.
The second level is strategic action within organizations. Even operating under the constraints of the capitalist employment system, designers can and should seek an active role in product decisions, acting as a user advocate in roadmap meetings, proposing well-being metrics alongside retention metrics, and documenting the ethical implications of the decisions presented to them. This is not naivety about corporate power dynamics, but recognition that omission is also a choice, and that internal criticism — even when it does not prevail immediately — creates records, precedents and, eventually, more responsible organizational cultures. The strategic designer, then, is one who balances business objectives with a commitment to user well-being, without abandoning either.
Finally, it is acknowledged that this study has limitations inherent to its qualitative approach and its scope of four applications which, while representative, do not exhaust the diversity of the digital ecosystem. Future research could broaden the scope to include online betting platforms and adult content networks — edge cases of the Dealer quadrant —, as well as conduct longitudinal studies with real users to measure the impact of the identified mechanisms on objective well-being indicators. Building an ethical interface-audit protocol, based on the Manipulation Matrix and the Hook Model, stands out as the most immediate and applicable next step of this research for the Brazilian professional context.
References
BRAZILIAN CENTRAL BANK. Note on sports betting (bets). Brasília: BCB, 2024.
BRIDGER, Darren. Neurodesign: como o neurodesign pode aumentar o engajamento, a influência e o poder de persuasão da comunicação visual. São Paulo: Editora Cultrix, 2015.
DUTRA, Rian. Enviesados: como os vieses cognitivos influenciam o design e o comportamento do usuário. São Paulo: Editora Alta Books, 2020.
EYAL, Nir. Hooked: how to build habit-forming products. New York: Portfolio/Penguin, 2014.
FOGG, B. J. Tiny Habits: the small changes that change everything. Rio de Janeiro: Editora Sextante, 2019.
KAHNEMAN, Daniel. Thinking, fast and slow. New York: Farrar, Straus and Giroux, 2011.
MAIS QUE VISUAL. Jogo do Tigrinho: um problema também de UX. Substack, nov. 2024.
MARQUES, Daniel; FALCÃO, Thiago. Design persuasivo e mediação tecnológica em plataformas digitais de apostas: práticas e dilemas éticos de designers e desenvolvedores. Arcos Design, Rio de Janeiro, v. 19, n. 1, 2026.
NORMAN, Donald A. Emotional design: why we love (or hate) everyday things. New York: Basic Books, 2004.
WARREN, Tish Harrison. Liturgy of the ordinary: sacred practices in everyday life. Downers Grove: InterVarsity Press, 2016.
Capstone thesis — Postgraduate Specialization in Emotional Design, Centro Universitário Belas Artes de São Paulo.