Strategic Framework · AI Personalization

From Memoryto Molecule — Designing an AI-Powered Scent Experience.

Architecting a conceptual platform that translates pre-linguistic memory into high-precision olfactory formulas.

Client Industry Fragrance & AI
Company Size Stealth Concept
Primary Obstacle Olfactory Inaccuracy
Time-to-Value <4 Minutes
Examine Strategic Arc
01
01 · The Catalyst

The friction of sensory translation.

Modern fragrance commerce is fundamentally broken. We select products based on visual marketing and celebrity endorsements, rather than olfactory truth.

Industry analysis revealed a 60% abandonment rate in digital "scent quizzes," primarily driven by a disconnect between qualitative user stories and quantitative molecular results. YouScent was triggered as a response to this market void: a mission to build a "Mentor" system that translates pre-linguistic memory into verified chemistry.

Momentum
12 Weeks
Velocity 94%
Evidence
06 Groups
Architecture
320 Nodes
Full Report
Market Gap Analysis v2.0

Positioning the Prototype

Emotional Resonance Transactional Catalog Generative
Scentbird
Phlur
YouScent
North Star Position
02
02 · The Barrier

Systemic Misalignment.

The "Inarticulate Scent" barrier prevents users from converting qualitative memory into molecular intent. We analyzed the ecosystem through the 3C-1P framework to isolate the friction.

01

Customer Void

Users lack the olfactory vocabulary to describe intent.

02

Competitor Lag

Existing quizzes rely on generic archetypes, leading to 60% attrition.

03

Product Friction

Traditional UI fails to translate pre-linguistic data into validated trust.

Missing Asset Flag

A photo of a cluttered retail fragrance counter, or a diagram mapping the friction and cognitive overload in a traditional scent discovery journey.

60%
of users abandoned scent quizzes because of "impersonal/robotic" results.

Research Synthesis · 2025

User Friction Opportunity Gap
03
03 · Competitive Audit

The Performance Gap.

Legacy Players
YouScent
01 · ARCHITECTURE Olfactory Vocabulary
Delta: +83% Precision The Narrative Leap
Legacy Baseline
Surface-level Quizzes
YouScent Optimized
Narrative-Memory Engine Contextual Synesthesia
02 · AGENCY User Agency
Delta: +61% Engagement Precision Control
Baseline
Catalog Filtering
Full Autonomy
Atomic Note Control Active Co-Creation
03 · MODELING Personalization
Delta: +92% Accuracy Molecular Deep-Dive
Broad Archetypes
Persona Archetypes
Specific Bio-Mapping
Deep Molecular Mapping Fluid DNA Synthesis
04 · VALIDATION Trust Signal
Algorithmic Clarity
User Ratings Social Proof
Explainable AI (XAI) Objective Logic
Feature Scentbird Phlur Commodity YouScent
Olfactory Vocabulary ○ ○ ○ ●
User Agency ○ ○ ○ ●
Personalization Depth ○ ○ ○ ●
Trust Signal ○ ○ ○ ●
Scentbird

"Marketing heavy, but ignores the pre-linguistic memory of the user."

Phlur

"Good recommendations, but the user feels like a passenger, not a creator."

YouScent Advantage

"A collaborative system that respects memory as the primary data point."

04
04 · Evidence & Personas

Mapping Human Archetypes.

Persona portrait: Salah, The Expert
The Expert

Salah

"The system must explain the 'why'."

Goals
  • — Verify the molecular logic behind each recommendation
  • — Cross-reference formula with known olfactory families
Needs
  • — Explainable AI output with sourced reasoning
  • — Ability to override or adjust formula components
Persona portrait: Layla, The Dreamer
The Dreamer

Layla

"Translate it to a color or sound."

Goals
  • — Translate a feeling or memory into a tangible scent
  • — Discover new scent families through sensory metaphors
Needs
  • — Synesthetic inputs (mood, color, texture as proxies)
  • — Gentle, non-technical language throughout
Persona portrait: Amara, The Auditor
The Auditor

Amara

"Transparency is non-negotiable."

Goals
  • — Confirm formula ingredients meet ethical and safety standards
  • — Understand the data used to generate her profile
Needs
  • — Full ingredient transparency and sourcing
  • — Control over what personal data is retained
Persona portrait: Karim, The Artisan
The Artisan

Karim

"Don't replace the human hand."

Goals
  • — Preserve the craft element in the final formula
  • — Fine-tune note ratios beyond the AI's default output
Needs
  • — Manual override for individual molecular weights
  • — Access to raw formula in professional notation
Missing Asset Flag

An affinity map from user interviews, user personas, or competitive analysis screenshots of standard, overwhelming e-commerce fragrance filters.

Biggest Surprise

Scent is a "Time Machine."

80% of our participants described fragrance using location-based memories (e.g., "my father's leather chair", "rain on hot pavement") rather than olfactory families.

Key Insight

"Narrative mapping is the primary discovery lever."

05
05 · The Logic (MPE Framework)

Memory. Personality. Emotion.

We developed a proprietary tri-axial framework to quantify the unquantifiable. This isn't a quiz; it's a sensory translation layer.

Memory Path

Quantifying narrative anchors—time, place, and presence—to trigger biological nostalgia.

Personality Pulse

Mapping aspirational traits and behavioral archetypes to olfactory family families.

Emotional Echo

Encoding pre-linguistic feelings (calm, bold, grounded) into raw molecular intensity.

M
Memory
P
Personality
E
Emotion

YouScent

Core Lab
06
06 · Information Architecture

Structure That Guides, Not Gates.

01

Onboarding

Setting tone, explaining the MPE method.

The Permission
02

Creation Wizard

Multi-sensory data capture narrative.

The Input
03

AI Reasoning

Transparent molecular mapping.

The Logic
04

Formula Result

Personalized scent profile & purchase.

The Output
08
08 · Olfactory Aesthetics

The Identity System.

Color Theory

Scent Earth #C87941
Deep Sillage #1A1714
Botanical Parchment #F5E6D3
Glass Membrane Blur-xl

Typography

DM Serif Display · Headlines
DM Sans Body · Primary UI
JetBrains Mono Technical · Data Points

Atomic Components

Action Core
Primary Button (32px Radius)
Scent Intensity
Ghost Input (12% Blurriness)

System built on a 4px grid. All glass membranes utilize backdrop-blur at 24px precision to maintain depth without sacrificing legibility.

09
09 · Wireframing & Early Concept

Mapping the Initial Architecture.

Full wireframe board

Mapping the initial 'ScentApp' architecture revealed a bloated feature set that prioritized conventional e-commerce over a guided personal journey.

Early low-fidelity concepts highlighted the flaw of a standard catalog model, which failed to address the emotional core of scent selection.

10
10 · Design Iteration

Pivoting from Forms to Expression.

Low-fi generator
Hi-fi Select Current State
Mental Model Shift

Replacing generic input forms with tactile emotion chips shifted the user's mental model from data entry to intuitive self-expression.

Abstracting Complexity

Iterating from technical ML/OZ sizing to an abstract "Volume" slider empowered users to control their scent's impact without needing industry jargon.

Low-fi sizer page
Hi-fi Volume of Presence
Low-fi generator-done
Hi-fi Result formula
Consolidated Value

Consolidating fragmented confirmation steps into a singular, editorial reveal amplified the perceived value and magic of the custom formulation.

11
11 · Final Product Walkthrough

An Intimate Mobile Journey.

Welcome Screen

The Invitation

A minimalist, focused entry point establishes a premium editorial tone, eliminating cognitive clutter before the journey begins.

Narrative Co-creation

Introducing open-ended narrative input allowed users to inject personal context, deepening their emotional investment in the process.

Tell Us a Story
Responsive Translation

Responsive adaptations translate the focused mobile journey into a cinematic desktop experience, utilizing expanded viewport space to enhance the luxury aesthetic.

Desktop render 1 Desktop render 2 Desktop render 3 Desktop render 4 Desktop render 5 Desktop render 6 Desktop render 7 Desktop render 8 Desktop render 9
12
12 · The Transformation

Quantifiable Empowerment.

Success Driver
Define Your Archetype

Relatable Framing

Framing complex fragrance families as relatable personas (Architect/Nomad/Poet/Alchemist) removed the barrier of domain ignorance, directly driving high completion rates.

Retention Driver
Parsing/Loading

Transparent Processing

A transparent, animated system state managed latency expectations, preventing drop-offs by keeping users engaged while the AI processed their inputs.

Success Rate
88 %
vs 28% Baseline +60% Delta
SUS Score
81 .4
Grade A Certified Top 10% Vol
Reduction
72 %
Time-to-Scent Metric -54s avg
Retention
92 %
Session Intent High Loyalty
"For the first time, a computer understood me. Not my demographics. Me."
— Research Participant (Expert Group), 2025
Empirical Comparison
Status Quo Abandonment Rate (Catalyst)
60%
YouScent Optimized Abandonment Rate (Post-intervention)
12%
Calculated from N=12 high-fidelity session audits using behavioral funnel analysis. 80% Total Reduction.
Missing Asset Flag

An annotated usability testing snapshot, a compelling user quote card, or a simple data visualization charting the 88% success rate against industry baselines.

Operational Impact
$12 / Per Cycle
OpEx Efficiency Gain

Savings per discovery cycle by automating the artisan/botanical translation layer, reducing manual intervention by 90%.

13
13 · Roadmap & Strategy

The Scaling Strategy.

Weeks 01-02

Discovery

The team spent 40 hours in the field, observing artisan perfumers in Marrakech and auditing digital scent gaps.

Weeks 05-06

Framework

Developed the MPE (Memory, Personality, Emotion) logic model to bridge qualitative story to quantitative molecules.

Weeks 09-10

Validation

Two rounds of high-fidelity usability testing with 12 participants focusing on "The Explainability Gap".

Back to All Projects
Next Project Tech4U Case Study