Key Takeaways
- Helium10 Cerebro is the best reverse-ASIN keyword research tool for Amazon sellers
- With MCP (Model Context Protocol), you can run Cerebro analysis in Claude or ChatGPT — no manual exports
- Setup takes 1-2 hours (one-time), then saves 40+ minutes per keyword research session
- This tutorial covers: MCP setup, 5 starter prompts, and before/after workflow comparison
- Bottom line: Cerebro + MCP replaces hours of spreadsheet work with one 5-minute prompt
What is Helium10 Cerebro?
Cerebro is Helium10's reverse-ASIN keyword research tool. You enter a competitor's ASIN, and Cerebro shows you every keyword they rank for — with search volume, rank position, competition score, and more.
Why it's powerful:
- Reverse-engineer competitors' keyword strategy
- Find hidden long-tail keywords they rank for
- Identify keyword gaps (keywords they rank for, you don't)
- Validate demand before launching a product
The old way (manual):
- Enter ASIN in Cerebro web dashboard
- Filter results (search volume, rank, competition)
- Export to CSV
- Open Excel, build pivot tables
- Cross-reference with your own keyword ranks
- Manually flag opportunities Time: 45-90 minutes per analysis
The new way (with MCP):
- Open Claude Desktop or ChatGPT
- Run prompt: "Use Helium10 MCP to analyze competitor ASIN B0XXXXXXXX, show top 50 keywords by search volume where competition score <60"
- AI returns ranked list in seconds
- Follow-up: "Which of these keywords don't I rank for yet?" Time: 3-5 minutes per analysis
That's what this tutorial teaches — how to use Cerebro through AI instead of manually.
Prerequisites
Before you start, you need:
- Helium10 Diamond plan ($250/mo) — MCP requires Diamond or higher
- Claude Desktop (free) or ChatGPT Plus ($20/mo) — pick one
- Competitor ASINs to analyze (you probably have these already)
- Your own ASINs (for keyword gap analysis)
Don't have Helium10 Diamond yet? Compare plans here (affiliate link - we earn a commission if you sign up).
Don't have Claude Desktop? Download free from Anthropic (Mac, Windows, Linux).
Step 1: Set Up Helium10 MCP (One-Time Setup)
Time required: 1-2 hours (one-time only)
Option A: Follow Helium10's Official Guide
Helium10 published an official MCP setup guide: Getting Started with Helium 10 MCP
Steps:
- Log into Helium10 Diamond
- Navigate to Settings → API & Integrations
- Enable MCP access
- Copy your MCP OAuth credentials
- Add credentials to Claude Desktop config file
Estimated time: 30-60 minutes if following official docs
Option B: Use Our Step-by-Step Course
Our Helium10 MCP Mastery course includes:
- Annotated screenshots for every step
- Troubleshooting guide (10 most common setup errors)
- Verification prompts to confirm setup works
- Video walkthrough (Tier 2: Workflow Kit, $199)
Why this is faster: Official docs assume technical knowledge. Our course is written for sellers, not developers.
Estimated time: 15-30 minutes with our guide
Verification Test
After setup, run this prompt in Claude or ChatGPT:
Use Helium10 MCP to list my products and show which one had the highest revenue last month.
If it works: You'll see your product list with revenue data.
If it fails: Check our troubleshooting guide (5 most common errors + fixes).
Step 2: Your First Cerebro Keyword Research Prompt
Once MCP is connected, you can run Cerebro analysis directly in Claude/ChatGPT.
Basic Prompt Template
Use Helium10 MCP to find keywords for ASIN [COMPETITOR_ASIN]
in the [US/UK/etc.] marketplace.
Show:
- Top 50 keywords by search volume
- Only keywords with search volume >1,000
- Only keywords with competition score <60
- Rank position for each keyword
Format as a table sorted by search volume (highest first).
Replace:
[COMPETITOR_ASIN]→ Actual ASIN (e.g.,B0ABCD1234)[US/UK/etc.]→ Marketplace (US, UK, CA, etc.)
Example: Yoga Mat Keyword Research
Prompt:
Use Helium10 MCP to find keywords for ASIN B0BSHF7LZS
(Best Seller in Yoga Mats) in the US marketplace.
Show:
- Top 30 keywords by search volume
- Only keywords with search volume >3,000
- Competition score for each keyword
- Rank position for this ASIN
Format as a table.
What you'll get:
- Table with 30 keywords (e.g., "yoga mat", "thick exercise mat", "yoga mat non slip")
- Search volume (e.g., 150,000, 45,000, 12,000)
- Competition score (0-100, lower is easier to rank)
- Current rank position for that ASIN
Time: ~30 seconds (vs. 45 minutes manually)
Step 3: Advanced Cerebro Prompts (Copy-Paste Ready)
Once you've tested the basic prompt, try these 5 advanced workflows:
Prompt 1: Keyword Gap Analysis
Use case: Find keywords your competitor ranks for that you don't.
Use Helium10 MCP:
1. Find keywords for competitor ASIN [COMPETITOR_ASIN]
2. Find keywords for my ASIN [YOUR_ASIN]
3. Show keywords where:
- Competitor ranks in top 20
- I rank below position 50 (or don't rank)
- Search volume >2,000
Format as: Keyword | Search Volume | Their Rank | My Rank | Opportunity Score
Why this works: Finds quick wins — keywords with proven demand where you can improve rank.
Prompt 2: Long-Tail Keyword Discovery
Use case: Find 3-5 word phrases (easier to rank, still decent volume).
Use Helium10 MCP to find keywords for ASIN [COMPETITOR_ASIN].
Filter for:
- 3+ words (long-tail phrases)
- Search volume 1,000–10,000
- Competition score <50
Show top 30, sorted by search volume.
Why this works: Long-tail keywords are easier to rank for and often have better conversion rates (more specific intent).
Prompt 3: Seasonal Keyword Check
Use case: Identify if keywords are seasonal (e.g., "yoga mat" spikes in January for New Year's resolutions).
Use Helium10 MCP to analyze keyword "yoga mat".
Show:
- 12-month search volume trend
- Peak month and low month
- Average monthly searches
Then recommend: Is this keyword seasonal or evergreen?
Why this works: Helps you avoid launching products with hidden seasonality.
Prompt 4: Competitor Keyword Overlap
Use case: Compare 2-3 competitors to find keywords all of them rank for (= high intent).
Use Helium10 MCP to find keywords where ALL of these ASINs rank in top 20:
- ASIN 1: [COMPETITOR_A]
- ASIN 2: [COMPETITOR_B]
- ASIN 3: [COMPETITOR_C]
Show:
- Keyword
- Search volume
- Each ASIN's rank position
Only include keywords with search volume >5,000.
Why this works: If all your competitors rank for the same keyword, it's probably a money keyword for that niche.
Prompt 5: PPC Keyword Prioritization
Use case: Which keywords from Cerebro should you add to PPC campaigns?
Use Helium10 MCP to find keywords for ASIN [COMPETITOR_ASIN].
Filter for:
- Search volume 5,000–50,000 (not too broad, not too niche)
- Competition score 40–70 (competitive but achievable)
- 2–4 words (specific intent)
Show top 20, explain why each is good for PPC.
Why this works: Narrows Cerebro's 1,000+ keyword output to 20 actionable PPC targets.
Before vs. After: Real Workflow Comparison
Before MCP (Manual Cerebro Workflow)
Task: Find keyword gaps between my ASIN and top competitor.
Steps:
- Log into Helium10 Cerebro (web dashboard)
- Enter competitor ASIN → Run analysis
- Apply filters (search volume >2K, competition <60)
- Export to CSV (wait for download)
- Open my ASIN in Keyword Tracker
- Export my keyword ranks to CSV
- Open Excel, import both CSVs
- Use VLOOKUP to cross-reference
- Filter for keywords competitor ranks top 20, I rank 50+
- Sort by search volume
- Manually review and flag top 20 opportunities
- Copy to Google Sheet for team
Time: 60-90 minutes
Error-prone steps: CSV exports, VLOOKUP formulas, manual filtering
After MCP (AI Workflow)
Task: Same — find keyword gaps.
Steps:
- Open Claude Desktop
- Run this prompt:
Use Helium10 MCP:
1. Find keywords for competitor ASIN B0XXXXXXXX
2. Find keywords for my ASIN B0YYYYYYYY
3. Show keywords where they rank top 20, I rank 50+ or don't rank
4. Filter for search volume >2,000
5. Show top 20, sorted by search volume
Format as table with: Keyword | Volume | Their Rank | My Rank | Gap
- AI returns table in 30 seconds
- Follow-up: "Which 5 of these should I prioritize for PPC?"
- AI explains why (intent, competition, volume)
Time: 3-5 minutes
Error-prone steps: None — AI handles all logic
Time saved: 55-85 minutes per analysis
Common Mistakes (And How to Avoid Them)
Mistake 1: Not Specifying Marketplace
Bad prompt:
Use Helium10 MCP to find keywords for ASIN B0XXXXXXXX.
Problem: Defaults to US marketplace. If your product is UK-based, results are useless.
Good prompt:
Use Helium10 MCP to find keywords for ASIN B0XXXXXXXX in the UK marketplace.
Mistake 2: Asking for Too Many Keywords at Once
Bad prompt:
Show me ALL keywords for this ASIN.
Problem: Cerebro returns 1,000+ keywords. AI output gets truncated or overwhelming.
Good prompt:
Show top 50 keywords by search volume where competition <60.
Mistake 3: Using SKUs Instead of ASINs
Bad prompt:
Find keywords for SKU ABC-123.
Problem: MCP tools use ASINs, not SKUs. Prompt fails.
Good prompt:
First, show me the ASIN for my SKU ABC-123.
Then find keywords for that ASIN.
AI will look up ASIN, then run Cerebro.
Mistake 4: Not Asking for Interpretation
Mediocre prompt:
Find keywords for ASIN B0XXXXXXXX.
Problem: You get raw data. No insights.
Better prompt:
Find keywords for ASIN B0XXXXXXXX.
Then explain:
- Which 5 keywords have best opportunity (volume + low competition)
- Which keywords are too competitive to target
- Any surprising keywords I should investigate
AI provides analysis, not just data.
When to Use Cerebro Manually vs. via MCP
Use Cerebro Web Dashboard When:
- You need visual charts (search volume trends over time)
- You're exploring a new niche (browsing, not searching)
- You want to export data for offline analysis
Use Cerebro via MCP When:
- You know what you're looking for (keyword gaps, competitor analysis)
- You run the same analysis weekly (prompt library saves time)
- You want AI to interpret results, not just display them
- You're analyzing 5+ ASINs in one session (much faster with prompts)
Bottom line: MCP is faster for recurring workflows. Dashboard is better for exploratory research.
Next Steps: Build Your Prompt Library
Once you've tested these 5 prompts, save them in a document (or use our prompt library template).
Recommended workflow:
- Monday: Keyword rank check (track top 20 keywords)
- Wednesday: Competitor gap analysis (any new keywords they're ranking for?)
- Friday: Search term review for PPC (which new keywords converting?)
Time per week with MCP: <30 minutes total
Time per week manually: 5-8 hours
Annual time savings: 250-400 hours
At $50/hr, that's $12,500-$20,000/year in value from learning MCP.
Frequently Asked Questions
Yes. MCP access requires Helium10 Diamond ($250/mo) or higher. Platinum ($99/mo) does not include MCP.
Related Reading
- Amazon FBA Training in 2026: Why Tool Mastery Wins — How MCP training replaced theory courses
- Best Amazon PPC Automation Tools — Compare automation options
- Helium10 MCP Mastery Guide — Complete course from $99
Browse more Amazon strategy on the Lucrivo blog or compare all MCP courses.
