Product Research Workflows

A 48-node research system built from three interconnected sub-workflows that turn raw market data into actionable product intelligence. This is what product research looks like when you stop guessing and start building systems. Each sub-workflow handles a distinct research dimension -- keyword analysis, product evaluation, and Amazon marketplace intelligence -- and all three can be triggered independently or chained together for comprehensive market analysis.

Nodes	Category	Complexity	Key Integrations
48	Research & Analysis	Advanced	DataForSEO, Mistral, Anthropic, SerpAPI, Brave Search, Google Sheets

Prerequisites

• DataForSEO API credentials (login and password) for real search volume data
• Mistral API key for rapid-fire analysis tasks (6 nodes)
• Anthropic API key for nuanced evaluation tasks (3 nodes)
• SerpAPI API key for Amazon marketplace queries
• Brave Search API key for web-based product research
• Google Sheets OAuth2 credentials for persistent research database storage
• Knowledge level: Intermediate to Advanced -- requires understanding of webhook-triggered workflows, dual-model AI architecture, and SEO/product research concepts

Estimated Setup Time

45-60 minutes including DataForSEO account setup, API credential configuration, webhook endpoint testing, and Google Sheets template preparation.

What It Does

This workflow system provides three distinct research pipelines, each accessible via its own webhook endpoint. The first pipeline pulls real search volume data from DataForSEO and runs it through AI analysis to identify high-opportunity keywords. The second uses AI agents armed with Brave Search to evaluate specific products against market conditions. The third queries Amazon directly via SerpAPI and layers AI-powered insights on top of the raw results.

What makes this system powerful is the dual-model AI architecture. Mistral handles rapid-fire analysis tasks where speed matters, while Anthropic Claude steps in for nuanced evaluation that requires deeper reasoning. Each model is deployed where its strengths align with the task, not as an afterthought.

All three pipelines write structured results to Google Sheets, building a persistent research database that grows more valuable with every execution. The webhook triggers mean any external tool -- a Slack bot, a custom dashboard, or another n8n workflow -- can kick off research on demand.

Workflow Diagram

Node Breakdown

Triggers

• Manual Trigger x2
• Webhook Trigger x3

AI & Language Models

• AI Agent x3
• Mistral x6
• Anthropic x3

Research & Data

• DataForSEO
• Brave Search
• SerpAPI (Amazon)

Storage & Logic

• Google Sheets x10+
• Code x3

Step-by-Step Flow

1. Webhook Entry -- One of three webhook endpoints receives a research request. Each endpoint corresponds to a distinct pipeline: keyword research, product evaluation, or Amazon analysis.

2. Pipeline 1: Keyword Research -- DataForSEO pulls real search volume, CPC, and competition data for the target keywords. Mistral analyzes the raw data to identify high-opportunity terms with strong volume-to-competition ratios.

3. Pipeline 2: Product Evaluation -- An AI Agent equipped with Brave Search investigates the product from multiple angles -- market fit, competition density, pricing dynamics, and customer sentiment. Anthropic Claude provides the final evaluation with nuanced scoring.

4. Pipeline 3: Amazon Intelligence -- SerpAPI queries the Amazon marketplace for the target product category. The results are parsed, scored, and analyzed by Mistral to surface pricing patterns, review trends, and competitive gaps.

5. AI Synthesis -- Each pipeline produces structured output through its respective AI model. Mistral handles rapid analysis, Anthropic handles nuanced evaluation, and together they provide complementary perspectives.

6. Data Storage -- Results from every execution are appended to Google Sheets with timestamps, source attribution, and confidence scores. This builds a persistent research database over time.

7. Response -- The webhook returns a structured JSON response with key findings, enabling downstream systems to act on the intelligence immediately.

What Sets It Apart

DataForSEO for real search volume data. Unlike workflows that rely on AI to estimate search volume, this system pulls actual search data from DataForSEO. Real numbers, not hallucinated estimates. This single integration elevates the entire research output from "interesting guesses" to "data-backed decisions."

Three distinct webhook-triggered pipelines. Each research dimension is independently callable. You can run keyword research on Monday, product evaluation on Wednesday, and Amazon analysis on Friday -- or trigger all three simultaneously for a complete market picture.

Mistral plus Anthropic dual-model architecture. This system strategically deploys two different AI models based on task requirements. Mistral handles high-throughput analysis where speed matters. Anthropic Claude handles nuanced evaluation where depth of reasoning matters. The result is faster and smarter than either model alone.

Cost Considerations

• DataForSEO: Pay-per-request pricing; keyword data costs ~$0.05-0.10 per batch of keywords. Minimum deposit required
• Mistral: Cost-effective for high-throughput analysis; ~$0.10-0.50 per pipeline execution
• Anthropic: Used for deeper evaluation; ~$0.20-1.00 per pipeline execution depending on research depth
• SerpAPI: Free tier includes 100 searches/month; paid plans from $50/month
• Brave Search: Free tier (2,000 requests/month); paid plans from $5/month
• Google Sheets: Free with Google account
• Estimated cost per research run: $0.50-3.00 depending on which pipeline(s) are triggered

Ideal For

• E-commerce entrepreneurs validating product ideas before investing inventory dollars
• Product researchers building systematic evaluation frameworks
• Amazon FBA sellers seeking data-driven product selection
• Dropshippers identifying profitable niches with real market data

Tip: Key Techniques

Dual-Model AI Strategy -- Deploying Mistral for speed-critical analysis and Anthropic for depth-critical evaluation. Match the model to the task rather than using a single model for everything.

Webhook-Triggered Sub-Workflows -- Each research pipeline operates as an independent service with its own endpoint. This modular architecture allows external systems to trigger specific research on demand.

Persistent Research Database -- Every execution appends structured results to Google Sheets, building a cumulative research asset that grows more valuable over time. Historical data enables trend analysis and retrospective validation.

Download

The workflow.json file in this folder can be imported directly into your n8n instance via Settings > Import Workflow.