LinkedIn Ad Library Analytics Report Generator

Some might not know this, but you can research advertisers on Linkedin via the publicly available Linkedin ad library. It is quite a powerful resource if looking to research competitors or get inspiration. I highly recommend surfing through various advertisers as they are informative. However, you can apply LLM to analyze various advertisers and get a comprehensive report to take it up a notch. I find this useful when starting a new job, and I want to review what the company has been doing on LinkedIn.

This is a LinkedIn Ad Library analysis workflow that consists of two main steps:

1. Get all paginated results - A Python code step that fetches and aggregates LinkedIn ad data

2. LLM - An LLM step that analyzes the aggregated data and generates a comprehensive report

The workflow takes the following inputs:

• Advertiser Name

• From Month (1-12)

• From Year (YYYY)

• To Month (1-12)

• To Year (YYYY)

Tool(s) used:

The workflow operates as follows:

1. The Get all paginated results step:

   • Makes authenticated API calls to LinkedIn's Ad Library

   • Handles pagination to get all results

   • Manages rate limiting and retries

   • Aggregates all ad data into a structured JSON format

   • Includes error handling and proper response formatting

2. The LLM step:

   • Takes the aggregated data from the previous step

   • Analyzes ad types, geographic distribution, targeting, and impressions

   • Generates a detailed markdown report with key metrics and insights

   • Uses GPT-4o as the model with a temperature of 0.3 for consistent, factual analysis

Video Walk-thru