Checkpoint 2 — Slide Deck
EXE101 · Group 01
Class SE1933-NJ
AI platform that simulates customer behaviour

MARKET
LAB

Turn scattered customer data into actionable insights — before you launch a product or campaign.

Instructor
Vương Tiểu Oanh
Mentor
Phạm Thanh Hương
Team
Group 01 · SE1933-NJ
15 min + 10 min Q&A
Presentation Flow · 15′ + 10′ Q&A

AGENDA

Market Lab · Checkpoint 2
I
Slides 3–6

Market Analysis

Alex · Vũ
  • Market Size & Growth
  • Customer Needs, Attitudes & Behaviour
  • Competitive Analysis
  • SWOT Analysis
II
Slides 7–14

Target Customers

Tùng · Bích · Vũ
  • Research Design & Survey Structure
  • Survey Analysis & In-depth Interviews
  • Key Customer Insights
  • Customer Personas & Target Priority
III
Slides 15–18

Value Proposition & USP

Quyền
  • Value Proposition
  • Unique Selling Points
  • Positioning Map
  • Summary & Next Steps
IV · Conclusion

Summary of key findings & strategic direction for Market Lab.

EXE101 · Group 01 · SE1933-NJ
1.1–1.2 · Market Analysis

MARKET SIZE & GROWTH

Owner · Alex
Market definition — intersection of 3 segments
  • AI Market Research — Research powered by AI
  • AI-Powered Customer Insights — Real-time customer behaviour analysis
  • Synthetic Persona Simulation — Simulating customer reactions with AI before market entry
AI in Market Research34.2%
CAGR / yr
$4.6B · 2025$36.8B · 2032
Generative AI43.3%
CAGR / yr
$23.1B · 2024$90.6B · 2028
Overall AI Market×4
over 5 years
~$260B · 2025>$1,200B · 2030
🇻🇳 Opportunity in Vietnam
82%
small SMEs reporting growth in 2024
44%
SMEs investing in AI 2024 — double 2023
78.8%
population using Internet, early 2025
95.4%
Internet users on ≥1 social network
$350M
Vietnam Digital Marketing Market
1.3 · Market Analysis

CUSTOMER NEEDS, ATTITUDES & BEHAVIOUR

Owner · Alex
🎯 Needs
  • Need to understand customers before deciding on product, price, campaign
  • 47% of researchers face budget constraints but must still ensure quality
  • Focus groups $4,000–5,000 → too high for SMEs & startups
  • Agencies need fast insights for pitches within a 5–7 day deadline
💡 Attitudes & Values
  • Prioritise cost saving, shorter time, simpler analysis
  • 87% of users/providers worry about AI reliability
  • Trust factors: reliability, fairness, data protection, explainability
🛒 Purchasing Behaviour
  • SME/Startup: free Google Forms, existing data, rely on experience
  • Agency: online surveys, social listening, AI tools for qualitative analysis
  • Decision: cost · speed · data quality · ease of use · AI reliability
📈 Trends

89% of researchers use AI tools regularly or experimentally · 69% have integrated synthetic data · strong rising demand to cut costs & shorten research cycles.

1.4 · Market Analysis

COMPETITIVE ANALYSIS

3 direct competitors · Alex
Selection criteria: same product segment (AI Persona / Synthetic Research) · same customer base (SMEs, agencies, research teams) · same price segment.
Synthetic Users
Yabble Virtual Audiences
Minds AI
Product
AI interviews, concept testing, problem discovery
Synthetic respondents; concept/pricing/messaging testing
Validated multi-persona panels for B2B teams
Customers
Researchers, PMs, agency owners
Research teams, agencies, large enterprises
B2B: product, marketing & research teams
Price
~$2–60 / interview
From ~$8,900 / year (high)
Quote-based, not public
Strengths
Fast, concept testing, integrates own data, 85–92% accuracy
Strong concept/pricing/messaging, backed by YouGov
Multi-persona, research-grade, 80–95% accuracy
Weaknesses
Not localized for VN, lacks VN behavioural data, hard to budget for SMEs
High price, enterprise-oriented, not localized for VN
Hidden pricing, focus on large B2B, not present in VN
Market Lab Opportunity
Localize VN + subscription + real-user validation
Cheaper, fits VN SMEs, adapts to VN behaviour
Target B2C SME VN, public pricing, VN personas
💡 Biggest gap

No AI Persona platform is yet localized for Vietnamese SMEs at an affordable price with real-user validation.

1.5 · Market Analysis

SWOT ANALYSIS

Owner ·
Strengths
  • AI automation: build personas in minutes instead of weeks
  • Data-driven personas from real behavioural, demographic & psychographic data
  • Multi-Agent Architecture using the OCEAN model — consistent, psychologically deep
  • Strong localization: VN language, culture & consumer behaviour
Weaknesses
  • Depends on the quality of input source data
  • Only fits text-based research — not a replacement for usability testing
Opportunities
  • 44% of VN SMEs invested in AI in 2024 (2× vs 2023)
  • Strong rising demand for cost-effective research solutions
  • Few direct competitors localized for SMEs in VN
  • AI in Market Research CAGR 34.2% (2025–2032)
Threats
  • International rivals could localize for VN at any time
  • SMEs may think free ChatGPT/Gemini is enough
  • Data-privacy concerns when sharing CRM/internal data with AI
  • Trust in AI-generated insights is still low
2.1 · Target Customers

RESEARCH DESIGN

Owner · Tùng
📋 Quantitative Survey
  • Goal: measure trends & pain points at scale
  • Tool: Google Forms — branching by industry
  • Questions: multiple choice, Likert 1–5, open-ended
  • 17 responses (B2B target ≥30)
🎙 In-depth Interviews
  • Goal: deeply understand each individual's context, motivation & barriers
  • Format: semi-structured, open-ended · in person + video call
  • 5 interviews (meets the minimum required)
🔗 Combining both methods
  • Cross-validate: quantitative figures + qualitative perspective
  • Analyse in parallel → derive shared insights
👥 Sampling Criteria
Industry
Marketing, startup, agency, F&B, retail, education, SMEs in VN
Experience
Joined surveys, campaigns, product dev, insight analysis
Geography & Age
Living/working in VN (HCM, HN, ĐN) · 18+
DMU Role
User · Influencer · Decision Maker · Buyer
Sampling
Purposive + convenience · network, LinkedIn, MKT communities
2.2 · Target Customers

SURVEY STRUCTURE & SAMPLE

Owner · Tùng
📐 Questionnaire structure — 10 parts (branching by industry)
1
Screening
Filter the right respondents by industry & role
6
AI Acceptance
Trust & willingness to use AI for research
2
Business Background
Industry, type, size, research experience
7
Willingness to Pay
Suitable price for an AI research tool
3
DMU Role
Role in the buying decision (user/influencer/buyer)
8
Purchase Consideration
ROI, data security, ease of use, payment, VAT
4
Research Behaviour
How they collect insights today
9
Feature Preference
Persona, concept/price/message testing, export
5
Pain Points
Budget, time, respondents, analysis
10
Final Feedback
Overall rating, likelihood to use Market Lab
📊 Sample breakdown — 17 responses
Marketing Agency64.7%
11 people · IMC / Digital / Branding · serving 5–10 clients · 10–50 employees · revenue 10–50 bn VND/yr
Retail / Fashion23.5%
4 people · 80% Marketing Exec · 60% with >20 locations · revenue 10–50 bn or >200 bn VND/yr
F&B~12%
2–3 people · from 1 location to chains >20 outlets · operating >5 years · mainly use sales data
2.3 · Target Customers

SURVEY ANALYSIS — PAIN POINTS

Owner · Bích
Agency · 11
Retail / Fashion · 4
F&B · 3
Hard to collect data
Slow data collection — 72.7%
Data scattered across sources — 100%
No dedicated research team — 100%
Hard to analyse
Hard to turn data → insight — 63.6%
Hard to predict customer trends — 80%
Online surveys are biased — 66.7%
Time pressure
Pitch deadlines only 5–7 days — 83.3%
Weekly/quarterly decisions — 40%
No systematic research process
Budget
Must use old reports (Nielsen/Kantar) — 50%
No budget to hire research agencies
Losses ~5–20M VND from wrong decisions
💬 Open-ended

Decisions needing insight most: running ads (Agency 91.7% · Retail 100%) · designing promotions 83.3% · choosing messaging 83.3% · adjusting price & market expansion 75–80%. → Insight needs recur weekly, tied directly to revenue generation.

2.4 · Target Customers

IN-DEPTH INTERVIEW — WHO & WHAT

Owner · Bích
👤 5 interviewees
DT
Dương Trần (Du Du)
Founder/CEO — Marketer Được Việc Agency
Cosmetics, F&B, beauty clinic, dental (Japanese/US/Korean brands entering VN)
MH
Vũ Minh Hoàng
CEO/Founder — CITO Agency
Performance marketing, e-commerce (mother & baby, beauty, edu)
VA
Vũ Thị Vân Anh
Strategic Planner — Ore IMC Agency
F&B, retail, beauty, lifestyle, education, service brands
ND
Hoàng Ngọc Diệp
Marketing Manager — Labooong
F&B (campaigns, product launch, competitor monitoring)
+1
5th interviewee
Conducted via personal network
❓ Key questions (semi-structured)
1
Current process
From brief → insight, how long? Which steps?
2
Tools & data sources
Which tools/methods in use? Main data sources?
3
Barriers with AI
Biggest barrier to using AI? Encountered AI errors?
4
Evaluating Market Lab
How should it be positioned? At which stage?
5
Suitable industries
Which industries best for Market Lab to add value now?
2.4 · Target Customers

INTERVIEW FINDINGS

Conclusions by interviewee · Bích

Agency IMC / Branding
Du Du · Vân Anh

  • Need: fast insights for brief/pitch; brief→strategy takes 1 week research + 1 month positioning
  • Pain: data fragmented across sources, time-consuming to consolidate
  • AI barrier: hallucination — needs validation with real users
  • ML role: AI research co-pilot, strengthen proposals before pitching

"AI helps find data & build survey frameworks — but the big idea still has to come from people"

Performance Agency
Minh Hoàng

  • Need: data to define segments, tactics & conversion as soon as a client onboards
  • Pain: juniors depend on seniors; lack structured research tools to scale
  • AI barrier: general AI tends to agree with the user, lacks critical thinking; limited grasp of VN behaviour
  • ML role: reduce research time, test hypotheses, auto-segment

"If a tool creates customer data that feels 'human' and cuts cost vs an agency → businesses will try it"

F&B Brand
Ngọc Diệp

  • Need: consolidate data from sales + reviews + social to choose test directions before launch
  • Pain: AI creates generic personas not convincing enough for F&B
  • AI barrier: low reliability without real data; worried about data security
  • ML role: help choose test directions — NOT replace real testing

"If it only creates generic personas, it's not convincing enough for an F&B brand to pay for"

🔑 Conclusion

Agency = primary target. F&B & Retail should be expanded use cases once we have case studies.

2.5 · Target Customers

FROM DATA TO ACTIONABLE INSIGHTS

Owner · Bích
1
Not short on data — short on synthesis
100% of agencies & retail have data but it's fragmented → ML must be a tool that consolidates + filters noise, not just creates personas
4
F&B is not the MVP main target
F&B depends on taste, location, in-store → real testing is faster. Focus on E-com/Retail & SaaS first
2
AI accepted as a support tool
87% worry about reliability; hallucination is the biggest barrier → need clear data sources, validation, explainable outputs
5
Industries with clear behavioural data fit better
E-com: views, add-to-cart, reviews · SaaS: trial, onboarding, churn → primary: E-com/Retail & SaaS
3
'AI Persona' isn't enough without real data
All 3 interviewees asked where the data comes from → shift messaging: 'AI turns data → insights' instead of 'AI creates personas'
6
Agency = secondary target & amplification channel
Agencies serve many clients → each agency is a multiplier for Market Lab (distribution channel)
2.6 · Target Customers

CUSTOMER PERSONAS

3 key personas ·
K
Nguyen Minh Khoi
Marketing Manager · E-commerce · 29 · HN
25–30M VND/month · E-commerce Startup — Fashion & Lifestyle
Pain points

Surveys take 3–5 days, biased results; ChatGPT inconsistent; budget 3–5M/month; CEO questions data reliability

Desired gains

Test concepts in hours; personas matching target; dashboard reports for CEO/CFO

Budget & buying

3–5M/month · self-decides if <5M · Freemium → Trial → Monthly

"I need to know if a campaign will work BEFORE running it — not after losing the whole budget"
H
Tran Quang Huy
Product Manager · SaaS/Tech · 31 · HCM
35–50M VND/month · SaaS Startup — Fintech/HRTech/EdTech
Pain points

Feedback fragmented across channels; validating features is costly; stakeholders challenge the roadmap

Desired gains

Validate ideas before dev; fast JTBD; evidence for the roadmap; reduce risk of building the wrong thing

Budget & buying

5–15M/month · shares decision with Eng Manager · Trial → Pilot → Full team

"I don't want to build a feature only to find out 3 months later that users don't use it"
T
Le Thu Trang
Strategic Planner · Agency · 28 · HCM
20–35M VND/month · Creative/Digital Marketing Agency
Pain points

Pre-pitch research is time-consuming; 5–7 day deadlines; hard to build multi-industry personas; clients doubt data credibility

Desired gains

Build personas in hours; insights for pitches; explore many segments; raise win rate

Budget & buying

2–10M/month · team lead approves small spend · Free → Trial project → Team sub

"With a 5-day deadline, I don't have time to survey 100 people for one brief"
2.6 · Target Customers

FINAL INSIGHT & TARGET PRIORITY

Owner ·
📌 Final Insight Statement
Market Lab should be positioned as an AI tool that turns scattered customer data into actionable insights — rather than just a tool that creates AI personas. In the MVP: prioritise E-commerce/Retail & SaaS/Tech (clear behavioural data); Agencies act as a secondary target / distribution channel.
🎯 Target priority in the MVP
★★★
Primary Target

E-commerce / Retail & SaaS / Tech

Rich, clear behavioural data; AI can analyse and generate value from day one.

★★
Secondary Target

Marketing Agencies

Frequent research needs, many clients → multiplier effect when using Market Lab.

Expanded Use Case

F&B & Retail/Fashion Brands

Real needs but dependent on real testing; suitable once case studies exist.

3.1 · Value Proposition & USP

VALUE PROPOSITION

Owner · Quyền
🔴 The Problem

Fragmented data

Teams need to understand customers faster before deciding. Data exists but is fragmented — time-consuming to consolidate, hard to turn into actionable insight.

🟢 The Solution

AI synthesis layer

An AI platform that synthesises data from many sources, analyses behaviour, builds personas by segment, and proposes actionable insights — not generic personas.

💰 The Benefit

Faster, lower-risk decisions

Shorten research, reduce reliance on intuition, decide based on data. Agencies save pitch time; businesses reduce launch risk.

🎯 Differentiator

Data-first, AI-second

AI is a processing layer — it doesn't 'fabricate' personas. VN localization. Structured research vs ChatGPT; pricing fits SMEs better than Yabble.

📣 Value Proposition Statement
Market Lab helps businesses and agencies turn scattered customer data into actionable insights — by using AI to analyse customer behaviour, identify segments, and test early assumptions before launching products or campaigns.
3.2 · Value Proposition & USP

UNIQUE SELLING POINTS

Owner · Quyền
A USP sits at the intersection of: what you do wellwhat customers wantwhat competitors can't do yet.
1
🤖 Simulate Behaviour Before Launch
Create AI Customers that simulate how each group reacts to product, price & message before launch — reduce uncertainty, save real-test budget.
4
📋 Structured Scenario Testing
Create personas, set up scenarios, save results, compare options, track — not just 'prompt & get answer' like ChatGPT.
2
⚡ Faster & Cheaper than Traditional Research
Focus groups $4,000–5,000 & take weeks. Market Lab: insights in hours, 3–15M VND/month — helps filter ideas.
5
🔄 Data Flywheel — smarter the more you use it
Accumulates scenarios & usage data → simulations grow more accurate. Creates natural switching cost, hard to replace.
3
🇻🇳 Localization for the VN market
Understands VN language, culture, consumer behaviour & business context — what Synthetic Users, Yabble, Minds AI can't do yet.
3.2 · Value Proposition & USP

USP STATEMENT & POSITIONING

Owner · Quyền
📣 USP Statement

Market Lab creates AI Customers localized for the Vietnamese market, helping businesses simulate how customers react to product, price & message before real rollout — faster & cheaper than traditional research, more structured than ChatGPT.

📍 Positioning: Market Lab vs. Alternatives
ChatGPT / Gemini
Synthetic Users / Yabble
Market Lab
Price
Free / low
$2–60/interview · $8,900/yr
Fits VN SMEs (subscription)
Localization VN
❌ No
❌ No
✅ Yes
Structured research
❌ Prompt-based
⚠️ Yes but complex
✅ Scenario system
Data accumulation
❌ Not stored
⚠️ Limited
✅ Data flywheel
Real-user validation
❌ No
⚠️ Limited
✅ Validation mechanism
Result explanation
⚠️ Inconsistent
⚠️ Limited
✅ Explainable + sourced
IV · Conclusion

SUMMARY & NEXT STEPS

Market Lab · Checkpoint 2
✅ Completed in Checkpoint 2
  • Market Analysis: AI in Market Research $4.6B → $36.8B · CAGR 34.2% · clear VN opportunity
  • Customer Research: 17 survey responses + 5 in-depth interviews
  • Key Insight: not short on data — short on the ability to synthesise it into actionable insights
  • Target Priority: Primary E-com/Retail & SaaS · Secondary Agencies · Expanded F&B
  • Value Proposition: AI turns scattered data into actionable insights
  • USP: AI Customers localized for VN + structured scenario testing + data flywheel
🔮 Next steps (Checkpoint 3)
  • Complete the MVP prototype: AI Persona Creation + Scenario Testing
  • Add surveys to reach a minimum of 30 B2B respondents
  • Build a real-user validation mechanism to boost credibility
  • Develop first case studies with E-commerce/Retail or SaaS
  • Set subscription pricing suitable for VN SMEs (3–10M VND/month)
Q&A

Thank you, instructors & mentor!

Market Lab · EXE101 · Group 01 · SE1933-NJ
← / → · space