Persona For User Story
| Date | Version | Details |
|---|---|---|
| 2025/11/27 | v1.0 | Initial documentation |
| 2025/11/28 | v1.1 | Add photos & Appendix |
Key characteristics describing photo search and organization behaviors (Ascending scale from 1 to 5).
| Behavioral Variable | Scale Description (1 ↔ 5) |
|---|---|
| Organization Habit | Spontaneous (1) ↔ Systematic (5) |
| Tagging Willingness | Avoids Manual Tagging (1) ↔ Active Tagging (5) |
| Trust in AI Automation | Prefers Manual Control (1) ↔ Trusts AI Automation (5) |
| Search Style | Scrolling/Browsing (1) ↔ Text/Natural Language Query (5) |
| Context Sensitivity | Entity-focused (Face/Place) (1) ↔ Activity/Context/Story (5) |
| Sharing Tendency | Personal Storage (1) ↔ Active Sharing/Collaboration (5) |
| Privacy Sensitivity | Low (1) ↔ High (5) |
| Time Availability for Organization | Little to None (1) ↔ Regular Investment (5) |
| Device Diversity | Single Device (1) ↔ Multi-Device/Cloud (5) |
Mapping of behavioral characteristics for the Primary Persona (Haneul) and Secondary Persona (Seo-woo).
- Haneul (Primary): Based on interviewees including a fashion-loving international student and a painter.
- Seo-woo (Secondary): Based on interviewees including a golf caddy and a clothing store owner (solopreneur).
| Behavioral Variable | Haneul (Primary) | Seo-woo (Secondary) |
|---|---|---|
| Organization Habit | 4 | 2 |
| Tagging Willingness | 4 | 2 |
| Trust in AI Automation | 3 | 4 |
| Search Style | 5 | 4 |
| Context Sensitivity | 5 | 4 |
| Sharing Tendency | 3 | 5 |
| Privacy Sensitivity | 4 | 3 |
| Time Availability | 3 | 1 |
| Device Diversity | 4 | 5 |
Four key behavioral patterns derived from user interviews and persona analysis.
| Behavioral Pattern | Details |
|---|---|
| 1. Context-First Search | • Searches using keywords related to activities or situations (e.g., "Escape room", "Club performance", "Screen golf") rather than just objects. |
| 2. Partial Tagging → Generalization | • Expects the AI to automatically group the remaining related photos after manually tagging just a few representative images. |
| 3. Minimizing Organization Effort | • Prefers to achieve bulk organization effects through a single, short labeling action, rather than spending time sorting individual items. |
| 4. Sharing vs. Archiving | • Haneul (Primary): Focuses on personal archiving and retrospection. • Seo-woo (Secondary): Prioritizes immediate sharing and repurposing content (for social media/work). |
Classification of the two personas (Haneul, Seo-woo) across four behavioral axes.
| Haneul | Seo-woo |
|---|---|
 |
 |
| Behavioral Axis | Haneul (Primary) | Seo-woo (Secondary) |
|---|---|---|
| Control ↔ Automation |
Semi-Automated / Minimal Effort Needs the remaining photos to be organized automatically after manually tagging a small batch (20-30 photos) right after an event. |
Full Automation / Speed Has almost no time for tagging. Expects accurate results within 5 seconds; the process must be faster than manually selecting photos. |
| Planner ↔ Spontaneous |
Planner / Routine Organizes photos immediately after assignments or trips; maintains a periodic organization routine (e.g., once a month). |
Spontaneous / On-Demand Strong tendency to procrastinate. Finds and utilizes photos immediately when needed for specific purposes; often requires batch processing of multiple images instantly. |
| Private ↔ Social |
Private / Archiving Focuses on personal albums for self-reflection and portfolios rather than sharing with others. |
Social / Sharing Frequently connects to social activities, such as sharing on SNS/Messengers or uploading to brand pages. |
| Episodic ↔ Atomic |
Episodic (Story-based) Searches for "chunks" of photos related to specific events or situations (e.g., "Preparing for the exhibition installation last spring"). |
Atomic (Individual Cut) Searches for specific individual shots matching precise criteria (e.g., "Shot wearing [Specific Product] where the face is not visible"). |
"If I tag just 20-30 photos, I want the rest of the related photos to be classified automatically."
- Demographics: Male, 24, Design Undergraduate Student.
- Location: Seoul, South Korea.
- Device Environment: Galaxy (Smartphone), iPad, Laptop.
- Photo Habits: Captures 1,500–2,000 photos/month.
- Main Subjects: School projects, classes, hobbies (e.g., Room Escape games).
- Character Type: Systematic Recorder who enjoys retrospection. Willing to perform minimal tagging but dislikes endless manual classification.
- Certainty in Memory Retrieval: Find photos quickly and accurately based on the context of the day, activity, people, or emotions.
- Low Effort, High Impact: Achieve mass automatic classification and clustering through a small amount of manual tagging.
- Privacy & Control: Keep sensitive photos private and have selective control over what is included in model retraining.
- Continuous Learning: Experience the system getting smarter and more personalized the more it is used.
- Scroll Fatigue: Exhausted by endless scrolling and organizing duplicate photos.
- Context Gap: Existing search (Location/Date/Face) fails to capture activities or contexts, forcing him to rely on vague memories and manual scrolling.
- Asset Retrieval: Difficulty finding past materials for class assignments, portfolios, or social media repurposing (e.g., struggling to find specific experience evidence when writing a resume).
- Sample Tagging: Spends just 2–3 minutes tagging "Key Sample Photos" immediately after an assignment, exhibition, or trip (e.g., "Escape Game", "25 F/W", "Project Series Name").
- Monthly Retrospective: Reviews auto-recommended albums and highlights once a month, making minor corrections.
- Natural Language Queries: Prefers long-tail queries like "Photos of night work during exhibition installation prep last spring."
- Accuracy: Over 95% of similar photos are correctly grouped after manually tagging only 20–30 images.
- Relevance: Accurate results appear within the Top-20 for context-based queries (e.g., "Rehearsal/Escape Game").
- Safety: Clear, intuitive, and safe settings to exclude sensitive content from AI learning.
Haneul Kim is a design major living in Seoul who records his daily life extensively, taking 1,500 to 2,000 photos every month. He works across multiple devices, moving between his smartphone, iPad, and laptop. He is accustomed to documenting various events and experiences, ranging from school assignments and exhibition preparations to his hobby, "Room Escape" games.
However, as his photo library grows, organizing and searching has become burdensome and exhausting. He frequently expresses the desire to tag just a few photos according to his standards and have the system automatically organize the rest. He wants to view the photos he needs at a glance through comfortable, natural search methods. He often experiences the frustration of scrolling endlessly through his gallery based on faint memory cues because simple date, location, or face searches are insufficient.
This inefficiency is particularly painful when looking for materials for assignments, portfolios, or Instagram content. When writing his resume or portfolio, he often wastes time asking, "When did I take that reference?" or "When did I have that specific experience?", searching one by one.
Haneul prioritizes high efficiency relative to effort. He believes it is sufficient to tag a few key photos after an event and lightly review highlights or albums once a month. He specifically wants to search using the context stored in his memory, such as "Photos of night work during exhibition installation prep last spring." Furthermore, he wants strict boundaries for his data; he wants to ensure personal photos of friends are not automatically used for training or sharing, preferring clear and intuitive settings to include data in model learning only when he explicitly chooses to do so.
Ultimately, Haneul’s goal is to enjoy organization and retrospection with minimal tagging and to find desired materials quickly via context-based search. He requires an app that serves as a "Personalized Memory Archive" featuring auto-classification, contextual search, and robust privacy controls.
"I want to pick the photos I need within 3 minutes, with auto-captions included."
- Demographics: Female, 28, Marketer.
- Location: Busan, South Korea.
- Device Environment: iPhone, MacBook (Apple Ecosystem).
- Platforms: Instagram, Naver Smart Store.
- Photo Habits: Captures ~1,000 photos/month.
- Main Subjects: Team events, marketing campaigns, personal travel.
- Character Type: Pragmatic Utility-Seeker. Has no time to organize; focuses on immediate retrieval, sharing, and repurposing.
- Immediacy: Extract a rough set of photos within 3–5 minutes (e.g., right after a shoot or just before a team meeting).
- Activity-Centric Search: Easily find specific visual contexts, such as "Holding a specific product," "Smiling shots," "Group photos," or "Shots without faces."
- Content Repurposing: Directly share to SNS or Messengers using auto-recommended tags.
- Collaboration-Friendly: Easily create shared albums with teammates and manage access permissions simply.
- Low Maintenance: Wants personalization accuracy to improve steadily without investing separate time into organizing.
- Zero Organization Time: Since she invests no time in manual sorting, the quality of AI defaults is critical.
- Privacy vs. Speed: Sensitive to copyright and privacy issues but has a strong need for rapid sharing.
- One-Tap Preference: Prefers a flow where she can simply tap a recommended tag or photo set and use it immediately.
- Speed: The system provides a ready-to-share result (Selected photos + Draft Caption) within 3 minutes of a query or upload.
- Shareability: Shared album permissions, link generation, or export to other apps must be completed within 2–3 taps.
- Relevance: High click-through and reuse rates for recommended tags.
Seo-woo Park is a 28-year-old marketer working in Busan. She relies heavily on her iPhone and MacBook for work, frequently handling promotional content for Instagram and Naver Store campaigns. She takes about 1,000 photos a month, mostly consisting of team events, marketing campaign shoots, and occasional personal travel.
Seo-woo has absolutely no spare time to organize photos meticulously. Instead, having "ready-to-use" photos at the moment she needs them is paramount. She requires a system that can extract a usable set of images within 3 to 5 minutes—whether it's right before a team meeting or immediately after a shoot. Unlike users who organize for memory retention, Seo-woo prioritizes utility, immediate use, and rapid sharing.
When searching, she prefers activity or visual context over simple dates or locations. She needs to find specific cuts like "Photos holding product X," "Smiling shots," "Group photos," or "Aesthetic shots without faces." It would be ideal if the system not only selected these photos but also generated auto-tags and captions so she could upload them to social media instantly.
Given her role, collaboration is frequent. She needs to create shared albums with teammates, manage permissions, and transfer necessary assets quickly. Therefore, the app must be collaboration-friendly—allowing sharing with tags and generating links within 2–3 touches.
While she needs speed, Seo-woo is also sensitive to privacy and copyright issues. She wants to feel safe while sharing content rapidly. Since she refuses to spend time organizing, she relies heavily on the reliable quality of AI defaults to handle personalization automatically.
Ultimately, Seo-woo's goal is to retrieve photos immediately after shooting without any manual organization, complete with captions, and share them effortlessly. Her core values are Immediacy, Collaboration Convenience, and Reliable AI Default Quality.
Description: This section summarizes the raw data collected from in-depth interviews with potential users. The insights derived from their photo management habits, storage volumes, and pain points served as the foundation for defining our Primary Persona (Haneul Kim) and Secondary Persona (Seo-woo Park).
Target Persona: Primary (Haneul Kim)
Characteristics: Users possessing tens of thousands of photos who struggle to retrieve specific past moments, inspirations, or evidentiary documents due to the sheer volume.
| ID | User Profile | Volume (Est.) | Key Behaviors | Pain Points & Needs | Connection |
|---|---|---|---|---|---|
| User A |
Art Student (Senior Year) |
~55,000 | • Participates in exhibitions 4+ times/year. • Routinely captures objects for artistic "Inspiration." |
• [Context Gap] "My gallery is a mix of artwork and random inspiration shots. It takes forever to find 'that specific texture' I saw months ago when I need it for a project." |
Context Search (Mood/Texture) |
| User B |
Global Marketer (Instagram Heavy User) |
~33,000 (Was 70k) |
• Mixed library: Selfies, Travel, Work. • Gave up migrating all photos when switching phones due to disorganization. |
• [Organization Fatigue] "I tried to organize them, but the volume was overwhelming. I ended up deleting half because I couldn't sort personal from work photos." |
Auto-Classification (Work/Life Balance) |
| User C |
Travel Enthusiast (Visited 20+ Countries) |
~66,000 | • Takes massive amounts of photos during trips. • Time-Lag Uploading: Posts to social media weeks/months after the trip. |
• [Retrieval for Retro] "I don't post immediately. When I want to upload a 'throwback' photo, scrolling back to find the best cut from a trip 3 months ago is exhausting." |
Retrospective View (Story-based) |
| User D |
Job Seeker (Recent Grad) |
~30,000 (Screenshots > 10000) |
• Captures job postings, resume evidence, and certificates. • Keeps info-heavy screenshots. |
• [Asset Retrieval] "I need to find specific certificates or proof of experience for my resume, but they are buried under thousands of random screenshots." |
Asset Search (Text/Document) |
Target Persona: Secondary (Seo-woo Park)
Characteristics: Users who view photos as temporary tools for work or sharing. They prioritize speed, curation, and immediate transmission over long-term storage.
| ID | User Profile | Volume (Est.) | Key Behaviors | Pain Points & Needs | Connection |
|---|---|---|---|---|---|
| User E | Golf Caddy |
Low (High Turnover : Frequent Add/Delete) |
• Takes photos of clients and sends them immediately. • Deletes photos often due to low storage capacity. |
• [Speed is Key] "I have zero time to edit. I need to select the best shot and send it to the client before they leave (within 5 hours). Speed is the only thing that matters." |
Immediate Sharing (Quick Selection) |
| User F |
Online Store Owner (Apparel) |
~15,000 (Frequent Deletion) |
• Takes photos of products, selects the best cuts, uploads, then deletes. • Struggles with phone storage limits. |
• [Curation Fatigue] "Picking the 'Best Cut' out of 50 similar photos (different angles/colors) is the hardest part. I need help selecting, not just storing." |
AI Curation (Best Cut) |
| User G | Interior Business Manager |
Low (Local Device) |
• Handles multiple construction sites. • Offloads site photos to Cloud/Drive to share with the team. |
• [Collaboration] "I don't want construction debris photos in my personal gallery. I move them to the team Drive immediately for sharing." |
Cloud/Team Sync (Separation) |
| User H | Grad Student | ~5,000 | • Active on Blog/Instagram. • Photo habits changed drastically after starting grad school. |
• [Lifecycle Change] Demonstrates how photo accumulation patterns change based on life stages (e.g., student vs. worker). | Adaptive System |
Based on the interviews above, the following core requirements were identified:
- The "Time-Lag" Problem (User C): Search features must support "vague memories" (context, mood) for users who organize or upload photos long after the event occurred.
- The "Semi-Auto" Sweet Spot (User A & B): Users are overwhelmed by fully manual organization but find fully automatic AI inaccurate. They need a "Few-Shot" system where tagging a small sample (20-30 photos) automatically organizes the rest.
- The "Evidence" Need (User D): Search capabilities must extend to screenshots and documents, treating photos as "Data Assets" for practical utility (e.g., Job hunting proofs, receipts).
- The "Gamified Reminiscence" (User B & C): Users abandon "Bulk Organization" due to fatigue. However, they enjoy "Rediscovering" old photos.