top of page
Market research.png

AI Market Research with Gemini and Claude

 
 

1. Intro to workflow

2. Competitor and Market Analysis with Gemini

3. Insight Generation with Claude

4. Gap Map and Market Brief​​

Intro to the Workflow

 

​When to Use it:

  • Handed a new project and know almost nothing about the space

  • Starting a side project and need to understand the competitive landscape

  • Client or PM needs to know by tomorrow what competitors are doing and what users are saying

  • You have a kickoff tomorrow and need to look like you've done your homework

What you'll have at the end of this

  • A structured competitive landscape covering every major player in your space.

  • The exact words people use when they're frustrated or delighted, pulled from reviews and forums.

  • A set of insight statements that connect user frustration to design opportunities.

  • A gap map: what every product has, what everyone does poorly, and what nobody does at all.

image.png

🧠 Why not just Claude?

  • Claude has a knowledge cutoff: It knows what it was trained on, which has a specific end date (It cannot tell you what launched three months ago).

  • It doesn't know what users are complaining about this week: It can't pull current App Store reviews or live Reddit threads.

  • Recency matters enormously: A competitive landscape built on 12-month-old data can point you in exactly the wrong direction.

🧠 The role of Claude

  • Take the Gemini output and find patterns across it.

  • Separate product-specific issues from category-wide problems.

  • Extract insight statements 📝: what user behavior means for design.

  • Build the gap map and other deliverables 📊: what's missing, done poorly, or nonexistent.

image.png

🌐 Why not just Gemini?

  • Gemini is not optimized for turning data into design insight: It surfaces patterns less reliably than Claude when working with large, messy datasets.

  • The brief, the insight statements, and the gap map, Claude does these better.

🌐 Role of Gemini

  • Search the live web for current user reviews, forum threads, and recent launches.

  • Pull the exact words people use when frustrated or delighted.

  • Find recent competitor moves: pricing changes, feature releases, rebrands.

  • Research Assistant: You're using it as a research assistant with live internet access, not as an analyst.

  • Not a decision-maker: Don't ask Gemini to draw conclusions; just gather.

Step 1: Competitor & Market Analysis with Gemini 🌐

 

In this step, Gemini serves as your real-time research assistant, scraping crowdsourced platforms for current user sentiment and recent market activity.

 

Prompt 1: User Sentiment (Run in a New Chat)

 

Setting: Thinking mode turned on. Do not use Deep Research.

I am researching the [Product Category] market. I want to understand what real users think about existing products right now. Search App Store reviews, Play Store reviews, G2, Capterra, and Product Hunt comments for these products: [List 4-5 Specific Competitors]. 

 

For each product, tell me:
1. The top 3 things users love, use their exact words.
2. The top 3 things users complain most about, use their exact words.
3. Specific features, functionality, or workflows users wish existed.
4. Any complaints that have spiked recently.

Source each finding. I need to know where it came from. I want real user language, not your summary of it.

Prompt 2: User Sentiment (Run in a New Chat)

 

Setting: Deep Research mode turned on.

I am researching the [Product Category] market. I want to understand what real users think about existing products right now. Search App Store reviews, Play Store reviews, G2, Capterra, and Product Hunt comments for these products: [List 4-5 Specific Competitors]. 

 

For each product, tell me:
1. The top 3 things users love, use their exact words.
2. The top 3 things users complain most about, use their exact words.
3. Specific features, functionality, or workflows users wish existed.
4. Any complaints that have spiked recently.

Source each finding. I need to know where it came from. I want real user language, not your summary of it.

Action Step:

Copy the exact text responses from both Gemini sessions and save them as separate files in a single folder. 📁

Step 2: Insight Generation with Claude 🧠

 

In this step, Claude takes over as your strategic analyst to find recurring patterns across the raw research and translate emotional user feedback into direct design implications.

 

Setup:

Create a new Project in Claude named "AI Market Research" and upload your saved Gemini files.

 

Verification:

Run a quick phrase to confirm: "I added 4 documents (2 prompts and 2 outputs). Read and confirm you see them. Do not provide an output, just confirmation."

Prompt 3: Synthesize the Competitive Landscape

 

In the attached documents of this project is the raw research data about the [Product Category] market. Please review all documents, both prompts, and the outputs from those prompts. 

My product is [Insert 1-sentence description of your product concept and who it is for].

Synthesize this into a structured, competitive landscape with the following:
1. A 2 to 3-sentence profile of each major player in the space: what they do well, do poorly, and who they are really for.
2. The top 5 complaints that appear across multiple products (category-wide problems, not just one product's issues).
3. The top 3 things users love in this category (what any new product must match to be considered).
4. Any pattern in these products that is not serving users well.

Use real user language from the research where useful. Specific observations only, no generic observations.

Prompt 4: Extracting Insight Statements

 

Based on the synthesis above, generate 6 to 8 insight statements about users in this market. Use this exact format:

[User Type] [does/believes/feels X] because [Underlying Reason], which means [Design Implication].

Requirements for each insight:
- Grounded in two pieces of evidence from the research (flag which evidence supports each one).
- The "because" must explain behavior and not just restate it.
- The "which means" must be specific enough to influence a design decision and not just "make something easier".
- No generic insights.

Step 3: Gap Map & Market Brief 📊

 

In this final step, we use Claude to turn those synthesized insights into a structured competitive matrix (Gap Map) and a highly focused strategic document (Market Brief) to align your team.

 

Prompt 5: The Gap Map

 

Based on everything above, build a gap map for the [Product Category] space using three tiers:

1. Table Stakes: What every product has; required to be considered (no differentiation value).
2. Done by Everyone, but Done Poorly: Things all competitors attempt, but none execute well. This is an execution opportunity.
3. Nobody Does This: Genuine gaps where user research suggests users want this, but no product delivers it.

For tiers 2 and 3, specify exactly what the gap is and what evidence points toward it. For tier 3, analyze why no one has done this yet (technical constraints, category issues, or just overlooked?). 

End with a one-sentence statement: "The single biggest opportunity in this space right now, based on all of the above, is..." Ensure this opportunity informs us as designers, not simply serves as a technical/backend solution.

Prompt 6: The Market Brief

 

Using everything we have built (the competitive landscape, insight statements, and gap map), write a highly opinionated Market Brief. It must be readable in 5 minutes and structured as follows:

1. The Market (one paragraph): What is happening, who the players are, and where it is heading.
2. What Users Actually Need: The top 3 core insights in plain language (not the formatted statements).
3. Where the Opportunity Is: Specific and opinionated direction drawn from the gap map.
4. What a Product Needs to Do to Win: Focus on positioning and experience quality, not a raw feature list.
5. The Opportunity in One Sentence: A clear strategic anchor specific enough to filter future design decisions.

bottom of page