EXERCISE 4 OF 6

Deep Research Challenge

Exercise 4: Deep Research Challenge

Mr. DIY Thailand - Generative AI Workshop

Objective: Use Deep Research tools for systematic competitive intelligence gathering

Tools Required:

  • ChatGPT Plus with Deep Research feature (Option A) OR
  • Claude account - free or Pro (Option B)
  • Internet connection for AI tool access

Context & Scenario

About Mr. DIY Thailand

Mr. DIY is Thailand’s leading home improvement and lifestyle retailer with 1000+ stores nationwide. We serve budget-conscious homeowners and DIY enthusiasts with our “Always Low Prices” positioning. Our product range includes 15,000+ SKUs across hardware, household items, electrical supplies, and more.

Key Facts:

  • 1000+ stores across 77 provinces
  • THB 16.2 billion revenue (2024)
  • 15,000+ product lines
  • 98.5 million customer transactions annually
  • Competitors: HomePro, Global House, Thai Watsadu

Your Scenario

Your Task:

You are a Market Intelligence Analyst supporting Q1 2026 expansion planning. Analyze HomePro and Thai Watsadu to identify 3 underserved product categories where Mr. DIY can capture market share with 20-30% lower pricing.

Context:

Your Role: Market Intelligence Analyst Business Objective: Support Q1 2026 expansion planning with competitive intelligence

Research Focus:

  • Competitors to analyze: HomePro (โฮมโปร) and Thai Watsadu (ไทยวัสดุ)
  • Geographic scope: Thailand market specifically
  • Time frame: Research recent data (within 6 months preferred)

What You Need to Deliver:

  1. 3 specific product categories with supporting data (pricing, demand, competitor gaps)
  2. Competitor pricing analysis for each category
  3. Mr. DIY pricing recommendation (20-30% lower positioning)
  4. Estimated market size for each category in Thailand

Business Context:

  • Mr. DIY’s “Always Low Prices” positioning is our competitive advantage
  • We need to identify gaps where competitors are overpricing
  • Focus on categories with high volume potential (not niche products)
  • Categories should be accessible to budget-conscious Thai homeowners

Success Metric:

Your research is successful when you deliver an actionable report with:

3 underserved product categories clearly identified ✓ Competitor pricing data specific to Thailand market ✓ Mr. DIY pricing strategy for each category (20-30% below competitors) ✓ Market size estimates with confidence levels noted ✓ Verified findings (sources cited, assumptions stated)

Your Goal: Use AI-powered research tools to gather competitive intelligence in 30 minutes that would traditionally take days of manual research.


STEP 1: Choose Your Research Tool

You have TWO powerful research tool options. Choose based on your preference and working style.


OPTION A: ChatGPT Deep Research (Automated)

Best for: Fast, comprehensive multi-source research

How it works:

  1. You formulate a clear research question
  2. Launch “Deep Research” feature in ChatGPT
  3. AI autonomously searches multiple sources (5-10 minutes)
  4. Receive comprehensive report with citations

Pros:

  • ✅ Speed - get report in 5-10 minutes
  • ✅ Breadth - covers multiple sources automatically
  • ✅ Citations - automatic source references
  • ✅ Comprehensive - structured report format

Cons:

  • ❌ Less control over research path
  • ❌ Can’t intervene mid-research
  • ❌ May miss nuanced follow-ups

Required: ChatGPT Plus subscription


OPTION B: Claude Research (Conversational)

Best for: Iterative exploration with deep context and control

How it works:

  1. You formulate initial research prompt using ROLE+INPUT+REQ+OUTPUT
  2. Submit and review AI’s response
  3. Ask follow-up questions through multi-turn conversation (5-7 turns)
  4. Build comprehensive context iteratively
  5. Synthesize findings from entire conversation

Pros:

  • ✅ Control - guide research direction actively
  • ✅ Deep context - long conversation threads
  • ✅ Iterative - refine and dig deeper as you go
  • ✅ Accessible - works with free tier

Cons:

  • ❌ More manual work - requires active guidance
  • ❌ Time-intensive - must drive conversation
  • ❌ Synthesis needed - extract insights yourself

Required: Claude account (free tier works, Pro better for longer context)


Make Your Choice:

I choose: ☐ ChatGPT Deep Research ☐ Claude Research


Research Execution Paths


PATH A: ChatGPT Deep Research Users

STEP 2: Define Clear Research Question (3 min)

Formulate a specific, comprehensive research question that covers all your objectives.

Example research question:

"What are the pricing strategies and product category gaps of HomePro and Thai Watsadu
compared to budget retailers in Thailand's home improvement market? Identify 3 specific
product categories where these competitors are premium-positioned, include typical pricing
data, estimated market size for each category, and opportunities for a value-focused
competitor like Mr. DIY to capture market share with 20-30% lower pricing."

Research question tips:

  • Be specific about what information you need
  • Include both competitors by name (HomePro, Thai Watsadu)
  • Specify geographic scope (Thailand)
  • Request pricing data and market size estimates
  • Ask for actionable insights (gaps, opportunities)
  • Mention Mr. DIY’s value positioning for context

✏️ Personal Notes (No submission required - keep for your reference)

Your research question draft:

[Draft your research question here - this is for your own use]





STEP 3: Launch Deep Research (5 min)


⚠️ IMPORTANT: Start a NEW conversation in ChatGPT

Open ChatGPT (requires Plus subscription for Deep Research). Start a fresh conversation.


  1. Open ChatGPT (ensure you’re on Plus tier)
  2. Paste your research question
  3. Click “Deep Research” button/feature
  4. IMPORTANT: ChatGPT may ask follow-up clarifying questions before starting research
    • If it asks questions, provide specific answers to improve research accuracy
    • Example questions: “What specific price ranges?”, “Which regions in Thailand?”, “What time period?”
    • Answer these questions to help Deep Research deliver more precise results
  5. Wait while AI researches (typically 5-10 minutes)
  6. Watch progress as it searches multiple sources

During waiting time: Review verification checklist, prepare executive summary template


STEP 4: Review Research Report (10 min)

When Deep Research completes, carefully review the generated report.

What to check:

  • Sources & Citations: Are sources credible and cited properly?
  • Recency: Is information recent (within 6 months)?
  • Thailand-specific: Is data specific to Thailand market (not generic Asia/global)?
  • Product categories: Are 3 specific categories identified with pricing data?
  • Market size: Are market estimates included with confidence levels?
  • Gaps: What questions remain unanswered?

If gaps exist: Launch follow-up Deep Research on specific areas


STEP 5: Verify & Create Executive Summary (12 min)

Apply verification checklist and synthesize findings into actionable report.

Verification Checklist:

  • ☐ Sources cited and credible?
  • ☐ Findings recent and relevant?
  • ☐ Price comparisons fair (like-for-like products)?
  • ☐ Market size estimates have confidence levels?
  • ☐ Assumptions clearly stated?

Note: What are you confident about vs. what’s uncertain or estimated?


PATH B: Claude Research Users

STEP 2: Design Research Approach (3 min)

Use ROLE+INPUT+REQUIREMENTS+OUTPUT structure to formulate your initial research prompt.

Example initial prompt:

ROLE: You are a retail market analyst specializing in Thailand's home improvement
and DIY retail sector with deep knowledge of competitive dynamics and pricing strategies.

INPUT: I need competitive intelligence on HomePro (โฮมโปร) and Thai Watsadu (ไทยวัสดุ)
for Mr. DIY Thailand, a value-focused home improvement retailer with "Always Low Prices"
positioning. The goal is to identify 3 underserved product categories where Mr. DIY
can capture market share with 20-30% lower pricing than competitors.

REQUIREMENTS:
- Identify 3 specific product categories where HomePro/Thai Watsadu are premium-positioned
- Current pricing data for these categories (specific examples if available)
- Market size estimates for each category in Thailand
- Competitor gaps or vulnerabilities in these categories
- Evidence and sources where available
- Confidence levels for each major claim (high/medium/low)
- Focus on high-volume categories accessible to budget-conscious Thai homeowners

OUTPUT: Structured competitive analysis with clear sections for each product category,
pricing data, market size estimates, and strategic recommendations for Mr. DIY's
20-30% lower pricing approach. Include explicit confidence levels for all findings.

✏️ Personal Notes (No submission required - keep for your reference)

Your initial prompt draft:

[Draft your ROLE+INPUT+REQ+OUTPUT prompt here - this is for your own use]







STEP 3: Execute Iterative Research (15 min)


⚠️ IMPORTANT: Start a NEW conversation in Claude

Open Claude (free tier is sufficient). Start a completely fresh conversation for Deep Research.


Conduct research through multi-turn conversation. Build context progressively.

Conversation structure:

Turn 1: Submit initial prompt → Review response

Turns 2-7: Ask follow-up questions to dig deeper:

Example follow-up prompts:

  • “What specific pricing data supports the claim that HomePro is premium-positioned in [category]?”
  • “What is the estimated market size for [category] in Thailand specifically?”
  • “What evidence shows consumer demand for budget alternatives in [category]?”
  • “How does Thai Watsadu’s pricing compare to HomePro in [category]?”
  • “What are typical price points for [specific products] in these categories?”
  • “How confident are you in these market size estimates? What’s the evidence?”
  • “What sources support these pricing comparisons?”

Goal: Build comprehensive picture through 5-7 conversational turns


✏️ Personal Notes (No submission required - track your conversation for your own learning)

Your follow-up questions:

Turn 2: [Your question]

Turn 3: [Your question]

Turn 4: [Your question]

Turn 5: [Your question]

Turn 6: [Your question]

Turn 7: [Your question]

STEP 4: Synthesize & Verify (12 min)

Review entire conversation thread, extract key insights, verify, and create executive summary.

Synthesis process:

  1. Read through all conversation turns
  2. Extract key findings from across multiple turns
  3. Look for patterns and themes
  4. Identify contradictions or uncertainties
  5. Note strongest vs weakest claims

Verification Checklist:

  • ☐ Sources mentioned (if any)?
  • ☐ Findings recent and relevant?
  • ☐ Assumptions clearly stated?
  • ☐ Confidence levels noted or inferable?
  • ☐ Critical claims cross-checked through multiple turns?

Executive Summary Template

For BOTH tool users

✏️ Personal Notes (No submission required - you can capture these insights mentally or save AI outputs for yourself)

This template helps you organize your thinking. You’re not required to hand this in - it’s for your learning and potential future use in your work.


1. Three Underserved Product Categories

Category 1: [Category Name]

Competitor pricing: [Price range from HomePro/Thai Watsadu]

Mr. DIY recommended pricing: [20-30% lower pricing strategy]

Market size estimate: [Thailand market size with confidence level]

Gap/Opportunity: [Why this category is underserved]

Category 2: [Category Name]

Competitor pricing:

Mr. DIY recommended pricing:

Market size estimate:

Gap/Opportunity:

Category 3: [Category Name]

Competitor pricing:

Mr. DIY recommended pricing:

Market size estimate:

Gap/Opportunity:

2. Competitive Positioning Summary

HomePro:

Pricing approach:

Key strengths:

Vulnerabilities Mr. DIY can exploit:

Thai Watsadu:

Pricing approach:

Key strengths:

Vulnerabilities Mr. DIY can exploit:

3. Strategic Recommendations

Based on competitive intelligence, what should Mr. DIY do?

1. [Recommendation for Category 1 with pricing strategy]


2. [Recommendation for Category 2 with pricing strategy]


3. [Recommendation for Category 3 with pricing strategy]


4. Confidence & Caveats

High confidence claims (verified, recent data):

-

-

Medium confidence claims (logical inference, some evidence):

-

-

Low confidence / Assumptions (limited data, needs verification):

-

-

Reflection Questions

✏️ Personal Notes (No formal submission required - reflect mentally or jot notes for yourself)

These questions help you internalize what you’ve learned. You don’t need to write formal answers unless you want to keep them for your own reference.


1. Research Tool Experience

Which tool did you use? (ChatGPT Deep Research or Claude Research)

☐ ChatGPT Deep Research ☐ Claude Research

What worked well about your chosen tool?

(Optional notes for yourself)



What were the limitations?

(Optional notes for yourself)



Would you use this tool again for similar competitive intelligence tasks?

(Optional notes for yourself)




2. Verification Insights

How confident are you in your findings?

Consider:

  • Were your sources credible and recent?
  • Did you find Thailand-specific data or mostly generic?
  • What assumptions did you have to make?
  • What would you verify before presenting to management?

Your assessment:

(Optional notes for yourself)




3. Business Application

How would you use these insights in your actual work at Mr. DIY?

Consider:

  • Which findings are most actionable immediately?
  • What additional research would you need before making decisions?
  • How does this compare to traditional competitive intelligence methods?

Your insight:

(Optional notes for yourself)




Key Takeaways

🎯 PRINCIPLE LEARNED: AI-powered research dramatically accelerates business intelligence gathering

Deep Research Tools Transform Competitive Intelligence

  • ChatGPT Deep Research: Breadth and speed (5-10 min comprehensive report)

    • Automated multi-source research
    • Automatic citations and structured reports
    • Best for comprehensive overviews
  • Claude Research: Depth and iterative exploration (more control, more work)

    • Conversational guidance allows precise direction
    • Deep context building across multiple turns
    • Best for targeted deep-dives

Both Tools Require Verification

  • Always cross-reference critical claims
  • Note confidence levels explicitly (high/medium/low)
  • Validate with experts when making major decisions
  • AI assists research, humans verify and decide

Choose Tool Based on Task

  • Need comprehensive overview fast? → ChatGPT Deep Research
  • Need deep exploration with control? → Claude Research
  • Need both perspectives? → Use both! Start with Deep Research for overview, then use Claude for deep-dive

Competitive Intelligence Informs Strategy

Research findings can guide:

  • Pricing strategy decisions
  • Product category selection
  • Store expansion planning
  • Marketing and promotional tactics
  • Competitive positioning

Reusable Prompts

ChatGPT Deep Research Question Template:

"What are the [SPECIFIC ASPECT] of [COMPETITOR NAMES] compared to [YOUR POSITIONING]
retailers in [GEOGRAPHIC MARKET]? Identify [NUMBER] specific [OPPORTUNITIES/CATEGORIES/GAPS]
where [COMPETITORS] are [POSITIONED HOW], include [DATA TYPE 1], [DATA TYPE 2],
and opportunities for a [YOUR POSITIONING] competitor like [YOUR COMPANY] to
[YOUR OBJECTIVE]."

Claude Research Initial Prompt Template:

ROLE: You are a [INDUSTRY] market analyst specializing in [GEOGRAPHIC MARKET]
with deep knowledge of [SPECIFIC DOMAIN].

INPUT: I need competitive intelligence on [COMPETITORS] for [YOUR COMPANY],
a [POSITIONING] with [KEY DIFFERENTIATOR]. The goal is to [RESEARCH OBJECTIVE].

REQUIREMENTS:

- [Requirement 1: specific data needed]
- [Requirement 2: analysis type needed]
- [Requirement 3: market scope]
- [Requirement 4: evidence and sources]
- Confidence levels for major claims

OUTPUT: Structured analysis with [OUTPUT FORMAT] and explicit confidence levels.

Follow-Up Question Framework:

EVIDENCE PROBING:
"What specific data supports the claim that [CLAIM]?"

CONFIDENCE CHECKING:
"How confident are you in [FINDING]? What's the evidence?"

MARKET SIZING:
"What is the estimated market size for [CATEGORY] in [GEOGRAPHY]?"

COMPARATIVE ANALYSIS:
"How does [COMPETITOR A] compare to [COMPETITOR B] in [DIMENSION]?"

GAP IDENTIFICATION:
"What evidence shows consumer demand for [ALTERNATIVE] in [CATEGORY]?"

Quality Checklist

Before finishing this exercise, verify:

  • ☐ I identified 3 specific product categories with pricing data
  • ☐ I included competitor pricing and Mr. DIY recommendations (20-30% lower)
  • ☐ I estimated market size for each category with confidence levels
  • ☐ I verified sources and noted what’s confirmed vs. assumed
  • ☐ My recommendations are actionable for Q1 2026 expansion planning
  • ☐ I completed the executive summary template with all sections

Need Help?

Not getting Thailand-specific data?

Try these approaches:

  • Emphasize “Thailand market specifically” in your question
  • Ask explicitly: “Focus on Thailand market, not generic Asia/global data”
  • Use Thai competitor names with Thai script: HomePro (โฮมโปร)
  • Request verification: “Is this data specific to Thailand?”

Deep Research taking too long (15+ minutes)?

  • Wait if it’s 5-10 minutes (normal processing time)
  • If 15+ minutes, cancel and reformulate with a narrower question
  • Consider switching to Claude Research if time is critical

TA is here to help! Raise your hand if you need guidance.


Ready? Start your competitive intelligence research!