How to use this guide.

Three things to know before you start: how the steps connect, what to prepare, and when to use the reference library.

Navigating the steps

Who does what

Every step has a recommended primary role, but collaboration is encouraged throughout. The lead designer should facilitate the full flow and own the ethics audit sign-off, but UX researchers, product managers, ethics reviewers, and behavioral scientists can all contribute at different stages.only one designer also can do all the steps individually based on project scope and budget availability

Track your progress

Each step page has a "Mark complete" checkbox. Your progress is stored locally in your browser, so returning to the guide picks up where you left off. The progress bar in the top navigation reflects your current state.

Onboarding: framework walkthrough

Before the team applies any framework tool, every participant should share a mental model of the process they are about to navigate. Step 1 is the team-level equivalent of onboarding.

Purpose

Give every team member an annotated Double Diamond with all ten steps overlaid in their correct phase, showing handoffs, platform shifts (Miro → Figma → Miro), and feedback loops. The activity is deliberately lightweight; its function is alignment, not instruction.

Miro board setup

Figjam (Figma collaborative Board) is also other option if the team prefer it.

• Board name: 00 — Framework Overview
• Central canvas: Double Diamond diagram with all ten step-cards positioned in the correct phase zone
• Four phase zones: Discover, Define, Develop, Deliver — each as a framed region
• Step-cards are clickable and link to their respective boards (01–08)
• Sidebar frame with glossary: Fogg terms (Spark / Signal / Facilitator), System 1/2 notation, six mode names

Facilitation activity

Walk the group around the diamond in sequence, spending two minutes on each step. For every step, surface two questions:
(1) what output does this step produce?
(2) what decision does that output unlock in the next step?

The goal is not to teach the framework in depth (that happens step by step later) but to make the overall arc visible so no one is later confused about why, for example, Figma is gated until Step 6.

Outputs

• Shared mental model of the ten-step flow
• Explicit team commitment to staying in Miro for Steps 1–5 (no premature Figma jumps)
• Named owner for each step

Common pitfalls

• Skipping onboarding because "everyone knows the Double Diamond." The framework-specific overlay is the value ,NOT the Double Diamond itself.
• Treating this step as a presentation rather than a dialogue. Questions here prevent confusion in Steps 2–5.

Research: behavioral journey mapping with cognitive annotations.

A standard journey map tracks touchpoints, emotions, and pain points. A behavioral journey map adds three annotation layers at every touchpoint: motivation, ability, triggr. These annotations are the raw material for Steps 3 and 4.

Purpose

Produce a journey map where each touchpoint carries seven data points: user action, emotion, motivation level (−2 to +2), ability level (−2 to +2), trigger level (−2 to +2). Without these annotations, downstream mode diagnosis becomes guesswork.

in next step we use all findings from last map in category (under fogg component) board and we can list all related cognitive biases which ended to that pain point.

Method options

Select one or more based on time available. Use the sprint-compressed one-hour behavioral brief (from Study 3 Theme 4) when discovery time is limited.

Miro board setup

• Board name: 02 — Behavioral Journey Map
• Horizontal swim-lane template, seven rows: touchpoints, actions, emotions, motivation (−2 to +2), ability (−2 to +2), trigger (−2 to +2)
• Evidence sidebar: verbatim quotes as sticky notes attached to each touchpoint
• Unvalidated flag: amber sticky on any annotation based on intuition rather than evidence

Outputs

• Fully annotated behavioral journey map
• Ranked list of "critical touchpoints", where motivation, ability, or trigger is deficient
• Evidence citations for each annotation

Common pitfalls

• Confusing journey mapping with persona writing. Personas are static archetypes; behavioral journeys are sequences of cognitive states over time.

Cognitive state diagnosis: the traffic-light card deck.

The framework's core diagnostic act. Map each critical touchpoint to one of six cognitive modes using a triage-style card deck

Purpose

Collapse the complexity of COM-B, Fogg, TTM, and dual-process theory into a single rapid classification : the "good enough fast enough" triage principle proposed by P01 during the Study 3 workshop, drawing on emergency response system design.

The six modes

Each card is A6 format, colour-coded red / amber / green using the triage metaphor. Click any card below to jump to its full reference entry.

Facilitation activity

For each critical touchpoint from Step 2, the team collectively drags the matching mode card onto the touchpoint. Disagreements are valuable ,they surface ambiguous evidence. Where the team cannot agree within two minutes, the touchpoint is flagged "needs more research" and returned to Step 2. This avoids the Study 2 pattern of forced stage assignments producing unstable downstream design.

Miro board setup

• Board name: 03 — Cognitive State Diagnosis & 3-traffic-light card deck.
• Six mode cards as draggable Miro components (template library provided)
• Journey map from Step 2 imported as background layer
• Disagreement zone: parking area for contested touchpoints

Outputs

• Every critical touchpoint annotated with mode (1–6)
• Each mode tagged with nudge entry point (Facilitator / Spark / Signal / excluded)
• Contested touchpoints flagged for return to Step 2

Common pitfalls

• Annotating System 1/System 2 from memory of Kahneman. Use the heuristic: if the user would stop and think, it is S2; if they would act without thinking, it is S1.

Bias mapping: the cognitive under-layer and problem-need linkage.

Modes diagnose the "what" of the user's state. Biases diagnose the "why." Surface the specific cognitive biases operating beneath each diagnosed mode and connect them to the unmet user need.

Purpose

The same mode can have different underlying biases in different domains. Mode 3 Motivated-Stuck in hydration is often attentional (salience deficit); in exercise it is present bias; in medication adherence it can be optimism bias ("I'll remember"). Different biases call for different mechanisms in Step 5. Without the bias layer, mechanism selection is mis-targeted.

Canonical bias-per-mode starting points

System 1 biases (fast / intuitive / automatic)

Operate under cognitive absorption, time pressure, or low deliberation. Triggered by salience, affect, or pattern-matching to prior cues.

System 2 biases (slow / deliberative / effortful)

Operate under explicit reasoning, analysis, and choice evaluation. Often produce errors of over-analysis, choice overload, or delayed consequence weighting.

Facilitation activity

For each mode-annotated touchpoint, pull 2–4 bias cards into the per-touchpoint grid.
Constraint: S1-dominant modes draw from the S1 bias set; S2-dominant modes from S2; mixed modes draw from either with justification.
For each bias, write one sentence on why it operates here (drawing on Step 2 evidence), then one sentence stating the unmet need it creates. The triple bias, why, unmet need is the decision-unlocking artefact for Step 5.

Miro board setup

• Board name: 04 — Bias Map
• Bias card library grouped into System 1 and System 2 frames (see Bias Library in sidebar reference)
• Per-touchpoint grid: three columns : Bias | Why it operates here | Unmet need it creates
• Identity-threat watch-zone: explicit callout for mindfulness or self-concept-sensitive touchpoints

Outputs

• Bias-need mapping per critical touchpoint
• Flagged high-risk biases (loss aversion in Mode 6, scarcity in any vulnerable state) carried forward as ethics-audit concerns for Step 7

Nudge selection: Spark / Signal / Facilitator and the 23 mechanisms.

Translate the diagnostic work of Steps 2–4 into a shortlist of candidate nudge mechanisms. The selection logic is deliberately simple so it can be applied under sprint pressure.

The selection decision tree

A single diagnostic question at the top routes to one of three branches:

Browse the 23 mechanisms

Filter below by trigger category. Each card shows ethical risk, primary bias target, and the modes where it fits best.

Miro board setup

• Board name: 05 — Nudge category selection
• Decision tree at top (primary barrier → trigger category)
• Three mechanism card library frames (23 cards total, downloadable as templates)
• Per-touchpoint shortlist canvas: 2–4 selected cards with written rationale linking each to mode, bias, and unmet need from Step 4
• High-risk mechanism flag zone

Outputs

• Candidate mechanism shortlist per critical touchpoint (typically 2–4 mechanisms)
• Written rationale linking each mechanism to mode, bias, and unmet need
• High-risk mechanism flags carried forward to Step 7 for ethics audit

Figma ideation: Nudge component library

The only step that leaves Miro or Figjam (collaborative boards). The switch is deliberate, this is where cognitive strategy becomes concrete visual artefact.

Purpose

Designers open the pre-built Nudge Component Library (a Figma Community file) and use it to ideate concrete interface treatments for each mechanism shortlisted in Step 5. Keeping visual tools isolated to one step prevents designers from jumping to interface craft before the behavioral and ethical work is done.

Component annotation schema

Every component in the library carries a standardised annotation block visible in the Figma sidebar. The annotation is the translation layer between cognitive strategy and interface implementation.

Example: Facilitator/ Default

we have 23 different digital nudge mechanisms which has underlayers in cognitive bias, dula system thinking and ethics

Activity

For each mechanism on the Step 5 shortlist, duplicate the component from the library into the ideation workspace, customise to the specific product context: brand typography, domain-specific copy (hydration vs mindfulness, for example), integration with surrounding interface. Produce 2–3 variations per mechanism. These will be evaluated in Step 7 (ethics) and Step 8 (stakeholder alignment) before final design synthesis.

Outputs

• 2–3 Figma design variations per shortlisted mechanism
• Annotated component instances showing which parameters were customised
• Export-ready frames for import back into Miro (Step 7 operates on these frames)

Common pitfalls

• Treating Step 6 as the whole design process. It is ideation only. Final feature design happens in Step 9 after ethics and stakeholder review.
• Ignoring the library's built-in annotations. Stripping them converts an evidence-based component into a decoration.

Ethics audit

Before any design leaves ideation, every shortlisted nudge passes through the five-question ethics audit. This step converts the abstract principles into a documentable, defensible, sprint-compressed protocol.

Run the audit

Try it below with a design you are considering. Mark each question Pass or Fail. Any single Fail halts progression, the design must be revised before proceeding to Step 8.

Intensity calibration

In parallel with the audit, place each design on the four-level intensity ladder.

Outputs

• Signed audit per design element (designer, product owner, ethics lead)
• Any design failing an audit question returned to Step 6 for redesign
• Audited designs tagged with intensity level for Step 9 synthesis

Stakeholder and business alignment.

An ethically audited, evidence-grounded nudge is useless if the designer cannot defend it to a product manager asking for something extractive. Step 8 converts behavioral rationale into stakeholder language.

The four stakeholder templates

Facilitation activity

Populate each template from the work produced in Steps 3, 4, 5, and 7. The behavioral design brief is always mandatory. Schedule a 30–45 minute stakeholder review — not as a presentation but as a shared artefact inviting stakeholder modification within ethical thresholds. Proposed modifications that would violate Step 7 thresholds are met with the dark pattern exclusion record.

Miro board setup

• Board name: 08 — stakeholder and business alignment.
• Four template frames (one per template), auto-populated from upstream boards where possible
• Business-goal vs. user-goal alignment grid: explicit trade-offs with a "no dark pattern" watermark
• Decision log: timestamped record of stakeholder modifications and ethical pushbacks

Outputs

• Signed stakeholder brief
• Business-alignment adjustments documented (within ethical thresholds)
• Any dark-pattern proposals formally excluded with evidence-based rationale

Interface design synthesis.

The synthesis point where everything produced in Steps 1–8 converges into the production-ready interface specification. Return to Figma for execution but use Miro as the brief.

The synthesis canvas

Each feature-level design decision is a node. Every node carries six required annotations drawn from earlier steps.

Outputs

• Annotated Miro synthesis canvas
• Final Figma spec with per-feature annotation (mode, mechanism, bias, intensity, dual-process)
• Design review sign-off incorporating Step 7 ethics outcome and Step 8 stakeholder brief

Evaluation: A/B testing and the Nudge Effectiveness Score.

Conventional A/B testing measures whether a change increased a metric. It does not measure whether that increase came at the cost of user autonomy. Step 10 closes the measurement blind spot. we can consider this step for our later work, as it is more relevant for post-launch evaluation rather than the design process itself.

Try the NES calculator

Enter per-condition metrics below. The formula penalises engagement achieved at the cost of intrusiveness; a nudge with high clicks but high perceived intrusiveness scores lower than a quieter nudge with modest clicks and low intrusiveness. we can not calculate the long term effecticvenss as it might have some different variable envolved

Decision rule

A/B test protocol

Metrics Collected

Feedback loop

Post-test, document lessons in the learning log and feed them back to Step 2 for the next design cycle. The framework is explicitly iterative, the NES informs the next journey map, which informs the next diagnosis, and so on.

Outputs

• NES per condition, documented
• Adoption decision with full evidence trail
• Lessons-learned log feeding the next design cycle (close the loop)

Framework action points

The Digital Nudging Framework v2.0 has some key action points. Each one operationalised in specific steps of the ten-step implementation.

Action A — Behavioral State Diagnosis

Six-mode diagnostic organised around the traffic-light card deck. Modes map onto TTM stages while incorporating dual-process annotations absent from standard TTM applications.

• Mode 1 Unaware (red, S1 dominant) — no schema
• Mode 2 Undecided (amber, S1/S2 conflict) — ambivalence
• Mode 3 Motivated-Stuck (amber, S2 frustrated) — friction-blocked
• Mode 4 Primed (green, S2 ready) — both motivation and ability present
• Mode 5 Forming (green, S2→S1) — habit consolidating
• Mode 6 Retreating (red, S2 defensive) — post-lapse reactance

Action B — Nudge Type Selection with Psychological Evidence

All 23 digital nudge mechanisms organised by Fogg trigger type. Each annotated with primary cognitive bias target, compatible modes, ethical risk rating, and health app example.

Selection logic: motivation absent → Spark; ability constrained → Facilitator; motivation and ability present but cue absent → Signal; uncertain → Facilitator first.

Action C — Intensity Calibration and Ethical Thresholds

Four-level intensity ladder (0/1/2/Dark) with cognitive load profile and ethical acceptance criteria per level. NES formula specified as the quantitative instrument. Five-question Ethics Audit converts abstract principles into pre-launch, documentable protocol. Dark nudge exclusion list: fabricated urgency, obstructed opt-out, shame-based framing, competitive social comparison, bait-and-switch defaults, negative self-image exploitation.

Action D — Dual-Process Targeting

Three cognitive mode profiles — System 1 (automatic), System 2 (deliberate), Transitional (mixed / stress-disrupted / novelty-demanding). For each: contextual description of when active in health app use, aligned nudge strategy, theoretical basis, and explicit guidance on what to avoid. Transitional profile specified with particular care — the state where reactance and self-efficacy damage risk is highest.

Action E — Health Domain Guidance

Domain-specific cognitive barrier profiles and nudge sequence recommendations for hydration, posture, mindfulness (Study 1 empirical, n=88 per domain) plus physical activity and sleep (Study 3 practitioner-validated, W notation). Each domain entry specifies characteristic cognitive barrier profile, empirically recommended nudge sequence, domain-specific design considerations, and representative participant voice.

The six modes — full reference.

The traffic-light card deck. Click any card to navigate to Step 3, where the deck is used in context. Print versions are A6 format for physical workshops.

23 nudge mechanisms.

Caraban et al.'s taxonomy of digital nudges, organised by Fogg trigger type. Filter by category, then match each mechanism to your diagnosed barrier.

Risk legend

Low risk — safe in most contexts · Medium — condition-dependent · High — conditional or categorical exclusion

Bias library.

Cognitive biases organised by the dual-process system they operate within. S1-dominant modes should draw from the S1 set; S2-dominant modes from S2. Mixed modes may draw from either with explicit justification.

System 1 biases (fast / intuitive / automatic)

Operate under cognitive absorption, time pressure, or low deliberation. Triggered by salience, affect, or pattern-matching to prior cues.

System 2 biases (slow / deliberative / effortful)

Operate under explicit reasoning, analysis, and choice evaluation. Often produce errors of over-analysis, choice overload, or delayed consequence weighting.

File architecture

Each mechanism ships with four domain variants (hydration, posture, mindfulness, general). Worked ethical-vs-dark-pattern comparisons are on a separate page for reference during Step 7 audits.

Component annotation schema

Every component carries the same annotation block in the Figma sidebar:

Distribution

Published to Figma Community under CC BY 4.0. A companion GitHub repository contains the ethics audit checklist as a downloadable PDF and the NES Evaluation Worksheet as a fillable form. Feedback is collected through an embedded Google Form linked from the Figma Community publication.

Study traceability matrix.

Each of the ten steps is connected to at least one empirical finding from the three-study research programme. This matrix is the framework's transparency guarantee.

Framework statement