Product Integration

How It Works

1. Before article generation, Publish Owl makes an LLM call with web search enabled
2. The AI researches real products matching your keyword
3. Product data (names, prices, ratings, links, pros/cons) is collected
4. This data is injected into the article generation prompt as the {'{{products}}'} variable

Settings

Requirement: AI Discovery requires web search to be enabled on at least one LLM workflow step. Without web search, the AI cannot look up real product data from the web.

Pre-Research

Research products ahead of time for faster, more reliable results during article generation. Select keywords from your list, research products in bulk, then review and edit the results before running your workflow. Pre-researched product data is cached and reused when the job runs, skipping the research step entirely.

Product Images

Three options for product images with AI Discovery:

• None: No product images
• Scrape OG Image: Fetches the Open Graph image from the product page URL
• Screenshot Product Page: Takes a screenshot of the product page for a visual preview

Tip: AI Discovery is the easiest way to add products — no API keys or manual data entry required. Just enable it and the AI handles the research automatically.

Single Article Mode

When running a single article, add products manually with the following fields:

• Name: Product name/title
• Price: Product price
• Rating: Star rating
• URL: Product page link
• Image: Product image URL
• Description: Short product description
• Pros: List of advantages
• Cons: List of disadvantages

Bulk Mode (CSV Variables)

For bulk article generation, use CSV variable columns to supply product data. Name your columns with the product_ prefix and reference them with indexed syntax in your prompts:

CSV columns: keyword, product_name, product_url, product_price

Prompt: Write a review featuring these products:
- {'{{product_1_name}}'} ({'{{product_1_price}}'}) - {'{{product_1_url}}'}
- {'{{product_2_name}}'} ({'{{product_2_price}}'}) - {'{{product_2_url}}'}

Step 1: Get Amazon PA API Access

1. Sign up for the Amazon Associates Program
2. Once approved, go to Tools → Product Advertising API in your Associates dashboard
3. Request API access (may require meeting sales thresholds depending on your marketplace)
4. Generate your Access Key and Secret Key

Step 2: Configure in Publish Owl

1. Go to Settings from the top menu
2. Scroll to the Amazon Product Advertising API section
3. Enter your credentials:
   • Access Key ID: Your AWS-style access key (starts with "AKIA...")
   • Secret Access Key: Your secret key (keep this secure!)
   • Partner Tag: Your Amazon Associates tracking ID (e.g., yourstore-20)
   • Marketplace: Select your Amazon marketplace (US, UK, DE, etc.)
4. Click Test Connection to verify your credentials work
5. Click Save Credentials to store them securely

Supported Marketplaces (21): United States, Canada, Mexico, Brazil, United Kingdom, Germany, France, Spain, Italy, Netherlands, Sweden, Poland, Belgium, Turkey, UAE, Saudi Arabia, Egypt, India, Japan, Australia, and Singapore.

Important: Your credentials are encrypted before being stored. The Settings page displays your usage stats including requests used today and your daily limit status.

Step 1: Enable Amazon Products

1. In your workflow editor, click + Add Amazon Products
2. The Amazon Product Integration panel will appear

Step 2: Configure Search Settings

Step 3: Add Your Tracking ID (Optional)

Enter your Amazon Associates tracking ID (e.g., yourstore-20) to earn commissions from product links.

How it works: When a job runs, products are fetched based on the keyword and made available to all workflow steps via the {'{{products}}'} variable. Your AI workflow steps can then reference this product data in their output.

Pro Tip: Fetch More, Let AI Curate

Set max products to 15-20 instead of the default 10, then let your AI workflow narrow down to the best 5-10 products. This works especially well when you have a Research Step enabled—it can analyze reviews, ratings, and product details to pick truly relevant products while filtering out:

• Low-rated products (under 4 stars)
• Near-duplicates or variations of the same item
• Products that don't match the article's focus
• Items with too few reviews to be trustworthy

This approach gives your final article a more curated, researched feel—like a human actually compared options before recommending them.

Option 1: From Column (ASIN/URL Lookup)

Option 2: Keyword Search

Standard Mode (All Product Sources)

In your workflow step prompts, reference the products variable. This works the same regardless of which product source you're using:

Write a buying guide for {'{{keyword}}'}.

Use the following products as the basis for your recommendations:
{'{{products}}'}

For each product, include the title, price, rating, and a brief description
of why it's worth considering. Include product links.

Template Mode (pSEO) — Amazon Variables

These variables are automatically available in your content template once Amazon integration is enabled:

Variable	Description
{'{{product_title}}'}	Product name
{'{{product_price}}'}	Formatted price (e.g., "$29.99")
{'{{product_rating}}'}	Star rating (e.g., 4.5)
{'{{product_reviewCount}}'}	Number of reviews
{'{{product_imageUrl}}'}	Product image URL
{'{{product_affiliateUrl}}'}	Affiliate link (includes your tracking ID)
{'{{product_asin}}'}	Amazon ASIN
{'{{product_features}}'}	Product features list

For multiple products, use numbered variables: {'{{product1_title}}'}, {'{{product2_title}}'}, etc.

Single Product per Article

Add columns for your product data and reference them with {'{{column_name}}'} in your prompts:

CSV columns: keyword, product_name, product_price, product_url, product_box_html

Prompt: Write a review of {'{{product_name}}'} ({'{{product_price}}'}).
Include this product box in the introduction:
{'{{product_box_html}}'}

This works well for pre-built WordPress block HTML, affiliate links, or any custom product markup.

Product Roundups (Multiple Products per Article)

When multiple CSV rows share the same keyword, they are automatically grouped into a single job. All rows' column data is collected and available as:

• {'{{products}}'} - A formatted text block listing all products and their fields
• {'{{product_1_product_name}}'}, {'{{product_2_product_price}}'}, etc. - Indexed per-product access (1-based)

keyword,        group,       product_name,     product_price, product_url
best headphones, headphones, Sony WH-1000XM5,  $348,          https://...
best headphones, headphones, AirPods Max,       $549,          https://...
best headphones, headphones, Bose QC Ultra,     $429,          https://...

Add a group column to control grouping explicitly when the same keyword should produce separate articles.

Tip: CSV-based product data and Amazon product data can be used together. The {'{{products}}'} variable combines data from both sources when both are available.

Example 1: AI Discovery — Buying Guide

Let AI research products automatically — no API keys or manual data needed.

• Product Source: AI Discovery
• Max Products: 8
• Search Instructions: "Focus on highly-rated options with good value"
• Include: Pricing, Pros & Cons, Ratings, Links

You are a product review expert. Write a comprehensive buying guide for "{'{{keyword}}'}".

Use these researched products as the basis for your recommendations:
{'{{products}}'}

Structure your article as follows:
1. Introduction explaining what to look for
2. Top picks with pros/cons for each
3. Comparison table
4. Buying advice and recommendations

Use the actual prices and ratings provided. Make the content helpful and unbiased.

Example 2: Amazon PA API — Affiliate Roundup

Fetch real-time Amazon product data with built-in affiliate tracking.

• Product Source: Amazon PA API
• Keyword Template: {'{{keyword}}'} (search exact keyword)
• Exclude Words: best, top, 2026, review
• Max Products: 10
• Tracking ID: mysite-20

You are a product review expert. Write a comprehensive buying guide for "{'{{keyword}}'}".

Here are the top products from Amazon to feature:
{'{{products}}'}

Structure your article as follows:
1. Introduction explaining what to look for
2. Top picks with pros/cons for each
3. Comparison table
4. Buying advice and recommendations

Include affiliate links for each product. Use the actual prices and ratings provided.
Make the content helpful and unbiased.

Sample keywords: "best wireless headphones 2026", "top laptops for students", "best running shoes review". With the exclude words configured, these become searches for "wireless headphones", "laptops for students", and "running shoes".

Example 3: Manual Products — Curated List

Full control over exactly which products appear, with data you provide.

• Product Source: Manual Products
• Products: Added by hand or via CSV columns with product_ prefix

Write a curated gift guide for "{'{{keyword}}'}".

Feature these hand-picked products:
{'{{products}}'}

For each product, explain why it makes a great gift and who
it's best suited for. Include prices and purchase links.

• Rate Limits: Publish Owl enforces 1 request per second and a soft cap of 2,000 requests per day to comply with Amazon's API terms.
• Backoff: If Amazon returns a rate limit error (429), there's an automatic 90-second backoff before retrying.
• Fresh Data: Product data is always fetched fresh from Amazon to ensure accurate pricing and availability.

All Product Sources

• Incomplete Data: Some products may not return all fields (e.g., missing price or image). Your prompts should handle incomplete data gracefully.
• Product Count: The actual number of products returned may be fewer than "Max Products" if not enough matching products are found.

AI Discovery

• Web Search Required: AI Discovery requires web search to be enabled on at least one LLM workflow step.
• Result Variability: AI-researched products may vary between runs since web search results change over time. Use Pre-Research to lock in specific products.

Amazon PA API

• Search Results: Products are returned in Amazon's relevance order — results may vary over time.
• Marketplace Lock: Your credentials are tied to a specific marketplace. To use a different Amazon marketplace, update the marketplace setting in Settings.