AI Product Opportunity Finder

🧠 AI Product Opportunity Finder

Turn real customer reviews into decision-ready product insights. The AI Product Opportunity Finder scrapes reviews from product listings and uses an LLM to surface top strengths, repeated complaints, unmet needs, listing ideas, and (optionally) a competitor comparison across up to 5 products. Amazon is supported today — more review sources coming.

✨ Features

• 🔍 Real review extraction — pulls customer reviews embedded on each Amazon product page.
• 🧠 AI-powered insights — repeated strengths, repeated complaints, unmet needs, themes with sentiment, listing suggestions, and marketing angles.
• ⚔️ Competitor comparison — optional cross-product summary with common complaints across the market and differentiation opportunities.
• 🛡️ Graceful failure — one bad URL doesn't stop the run; failed products emit an error row.
• 🌍 Any output language — set outputLanguage to get summaries in your preferred language.

🛠️ How It Works

1. Paste Amazon product URLs – 1 to 5 /dp/ASIN URLs.
2. Run the actor – it scrapes each product, extracts reviews, and calls the LLM for analysis.
3. Read the insights – structured JSON rows land in the run's dataset, ready to download or integrate.

📥 Input

Field	Type	Default	Description
productUrls	string[] (1–5)	—	Amazon product URLs in /dp/ASIN format. Required.
maxReviewsPerProduct	integer (1–50)	20	Cap on reviews analyzed per product. Amazon embeds ~6–8 reviews per product page, so larger values rarely change the output.
sortBy	"recent" | "helpful"	"recent"	Review sort order on Amazon.
analysisMode	"product_insights" | "competitor_comparison"	"product_insights"	Set to competitor_comparison to also get a cross-product comparison row.
outputLanguage	string	"English"	Language for AI-generated summaries.

Example input:

{
    "productUrls": [
        "https://www.amazon.com/dp/B09B8V1LZ3"
    ],
    "maxReviewsPerProduct": 20,
    "analysisMode": "product_insights",
    "outputLanguage": "English"
}

📊 Sample Output Data

The actor pushes three row types to the dataset, distinguished by type: raw_review, product_insights, and (optionally) competitor_comparison. Example product_insights row:

{
    "type": "product_insights",
    "productId": "amazon:B09B8V1LZ3",
    "asin": "B09B8V1LZ3",
    "productTitle": "Amazon Echo Dot (newest model)",
    "brand": "Amazon",
    "source": "amazon",
    "reviewCountAnalyzed": 6,
    "overallSummary": "Customers appreciate the sound quality and ease of setup, but some feel it's overpriced and miss the audio jack.",
    "topStrengths": [
        "Clear and vibrant sound quality",
        "Easy and quick setup process",
        "Good integration with Alexa and smart home devices"
    ],
    "topComplaints": [
        "Perceived as overpriced for the features",
        "Missing 3.5mm audio jack"
    ],
    "unmetNeeds": [
        "Reintroduction of the audio jack",
        "More competitive pricing"
    ],
    "themes": [
        { "theme": "Sound Quality", "sentiment": "positive", "summary": "Customers are impressed with the sound clarity and volume." }
    ],
    "listingSuggestions": [
        "Highlight sound quality and setup ease in the product description"
    ],
    "marketingAngles": [
        "Perfect for music lovers and smart home enthusiasts"
    ]
}

💰 Pricing

This actor uses Pay Per Event (PPE) pricing with two chargeable events:

Event	When it fires	Price
review-scraped	Charged once per Amazon review successfully extracted and pushed to the dataset.	$1.00 / 1000 reviews
insights-generated	Charged once per successful AI output row — one per product, plus one more row if competitor_comparison mode is used with ≥2 products.	$20.00 / 1000 rows

A typical run of 3 products (~6 reviews each) costs roughly 18 × $0.001 + 3 × $0.02 ≈ $0.08. Failed products and fallback AI rows (emitted when the LLM call fails) are not charged.

Uncover your next product opportunity with AI Product Opportunity Finder today! 🚀