5-Department Personas
12 min read
Five departments use the same data for different work goals and performance measures (KPIs). Selecting a department in the persona switcher changes the sidebar order, card emphasis, and chatbot tone to match that department's tasks.
At a Glance
| Code | Department | Mission | Core KPIs | Primary Scenarios |
|---|---|---|---|---|
| P1 | Brand Marketer | Own-brand campaign ROAS · New product launch | ROAS / CTR / New product adoption rate | S1 · S2 · S5 |
| P2 | Consumer & Trend Insights | Persona · Trend · Competitor analysis | LTV / Persona share / SOV | S2 · S4 · S6 |
| P3 | D&A · MarTech | Data · model quality · MarTech operations | Model accuracy / Drift / Pipeline SLA | S2 · S3 · S6 |
| P4 | CRM · LG Members | Member tier · Journey · Loyalty | Tier conversion / Active members / NPS | S2 · S7 · S8 |
| P5 | MD · Channel Sales | SKU · Channel sell-through · Inventory | Channel GMV / Inventory turn / New product adoption | S1 · S2 · S5 |
P1. Brand Marketer
Profile
- Position: Brand Marketing (by BU — Beauty: The History of Whoo · Su:m37 · OHUI / HDB: Elastine · Perioe / Refreshment: Coca-Cola · Fanta)
- Language tone: Campaign / ROAS / Conversion / Response rate / Retargeting / New product launch
- Decision unit: Weekly campaign operations, monthly budget reallocation, quarterly new product launches
- Data perspective: Channel × Campaign × Segment × Brand
KPIs
| Metric | Definition |
|---|---|
| ROAS | Campaign revenue ÷ Campaign cost |
| CTR / Response rate | Actions vs. impressions/sends |
| New product adoption rate | 30-day post-launch sales target achievement |
| Revisit rate | 30/60/90 days |
Chatbot Tone Examples
- "Break down this month's The History of Whoo new product launch campaign response rate by channel"
- "Predict SNS reaction for next week's Coca-Cola Zero campaign"
Primary Scenarios
- S1 Natural language search — "Women in their 30s, unpurchased for 90 days + organic preference" → instant targeting
- S2 Chatbot — invokes
campaign_simulator,attribution_calc - S5 Campaign ROAS — Bayesian channel mix
P2. Consumer & Trend Insights
Profile
- Position: Consumer & Market Insights / Trend Lab
- Language tone: Cohort / LTV / Persona / SOV / Trend / Churn
- Decision unit: Quarterly segment definitions, monthly trend reports
- Data perspective: Cohort × Behavior × External signals (social · trends · competitors)
KPIs
| Metric | Definition |
|---|---|
| LTV | Cumulative purchase amount over 12/24 months |
| Churn rate | Ratio of members with no purchase in 90/180 days |
| Persona share | Distribution across lifestyle personas |
| SOV | Share of voice in SNS · search for a category/brand |
Chatbot Tone Examples
- "Common patterns among members entering top 10% LTV in the last 3 months"
- "Newly emerging keywords in Olive Young reviews + connected first-party SKUs"
Primary Scenarios
- S4 Persona matching + clustering — Kids mom · Gold miss · Single household · Senior · Trendsetter
- S6 External signal fusion — All 4 types combined
P3. D&A · MarTech
Profile
- Position: AI/ML / Data Platform / MarTech operations
- Language tone: Accuracy / Drift / Pipeline / RMSE / AUC
- Decision unit: Model deployment cycle, data quality SLA
- Data perspective: Model performance × Data quality × Pipeline health
KPIs
| Metric | Definition |
|---|---|
| Model accuracy | Per recommendation/prediction/classification model |
| Data quality score | Missing · duplicate · outliers |
| Pipeline SLA | Daily/hourly processing success rate |
| Model drift | Input distribution change |
Chatbot Tone Examples
- "Cohorts where recommendation-model CTR dropped in the last 30 days + drift"
- "Statistical detection of anomalous transaction signatures"
Primary Scenarios
- S3 Insight cards — Code Interpreter + matplotlib
- S6 External signal fusion — correlation · lag analysis
P4. CRM · LG Members
Profile
- Position: Membership operations / Loyalty / CS
- Language tone: Tier / Points / Benefits / Retention / Consent / NPS
- Decision unit: Monthly tier-conversion campaigns, quarterly benefit revamps
- Data perspective: Tier × Payment × Channel × Time
KPIs
| Metric | Definition |
|---|---|
| Tier conversion rate | Regular → Silver → Gold → VIP |
| Active members | Active members over 30/90 days |
| Point usage rate | Used vs. accrued |
| NPS | Membership satisfaction |
Chatbot Tone Examples
- "Common patterns of members whose activity dropped 6 months after entering VIP"
- "Absolute send-block guard for non-consenting members"
Primary Scenarios
- S7 Omnichannel journey — Owned mall → SNS → Olive Young → Mart → Repurchase
- S8 Guardrails — Bedrock Guardrails + consent + minor cosmetics
P5. MD · Channel Sales
Profile
- Position: Merchandising (MD) / Channel sales (owned mall · marts · H&B · convenience stores · QSR)
- Language tone: GMV / Inventory / Shelf / Order / Turn / Sell-through / PB / New product
- Decision unit: Weekly orders, monthly category reviews, quarterly channel evaluations
- Data perspective: Brand × Category × Channel × Time
KPIs
| Metric | Definition |
|---|---|
| Channel GMV | Per-channel revenue |
| Inventory turn | Revenue ÷ average inventory |
| New product adoption rate | 30-day post-launch revenue |
| Channel share | Owned mall vs. Mart vs. H&B vs. Convenience store |
Chatbot Tone Examples
- "Top 5 channels/stores where Perioe sales dropped last week and the causes"
- "Order guide for hot packs, hand creams, and cocoa when -5°C is forecast"
Primary Scenarios
- S5 Campaign ROAS — channel · store unit
- S1 Natural language search — SKU turn · new product shelves
How the Persona Switcher Affects the UI
| UI Element | P1 | P2 | P3 | P4 | P5 |
|---|---|---|---|---|---|
| Sidebar emphasis | S1·S2·S5 | S4·S6 | S3·S6 | S7·S8 | S1·S5 |
| Insight card order | Campaign/ROAS | LTV/SOV | Drift/Quality | Tier/NPS | GMV/Inventory |
| Chatbot system prompt | Marketer | Analyst | Data scientist | Operations manager | MD/store manager |
| Recommended queries | "This week's campaign ROI?" | "High-value cohort?" | "Anomalous transactions?" | "VIP churn risk?" | "Channels where Perioe sales dropped?" |
| Guardrail strength | Marketing consent required | PII masking | Answer confidence | Consent · tier | Protect store sales info |
Demo Usage Tips
- During S2 chatbot stage, demonstrate the persona switcher — issue the same query "What caused Perioe's sales drop this month?" in P1·P3·P5 sequentially to emphasize differences in answer tone, KPIs, and tool calls
- S8 Guardrails stage should be demoed in the P4 persona — consent · minor cosmetics guards activate automatically
- S7 Omnichannel journey should be demoed in the P4 persona — a 90-day timeline of a single member, emphasizing member behavior crossing BUs