Client Background

  • Client: A leading ad tech firm in India
  • Industry Type: Digital Advertisements
  • Products & Services: Monetized websites running ads through global platforms like Google AdX, AdSense, and Ezoic
  • Organization Size: 70+

The Problem

The client ran multiple monetized platforms and global ad campaigns through Google AdX, AdSense, and Ezoic. Each platform had its own reporting tools, fragmented across different websites and campaigns.

Calculating Return on Ad Spend (ROAS) and other performance metrics required heavy manual work, involving logins to multiple dashboards, data downloads, and tedious aggregation. This created:

  • High operational overhead
  • Delayed reporting
  • Inconsistent tracking of performance metrics
  • No unified or real-time view of global ad performance

Our Solution

We engineered an automated and centralized ad performance tracking system that included:

  • API Integration: Established authenticated access to Google AdX, AdSense, Ezoic, and Google Ads APIs.
  • ETL Pipeline: Built a Python-based ETL pipeline that fetched, normalized, and loaded ad campaign data into a PostgreSQL database.
  • Scheduled Updates: Enabled automated data refresh every 6 hours to ensure near real-time insights.
  • Deployment: Dockerized the entire ETL solution and deployed it on a DigitalOcean droplet.
  • Monitoring and Alerts: Integrated Slack notifications for ETL job status (success/failure) along with in-depth logging and error handling.

Solution Architecture

  Data Sources: Google AdX, Google AdSense, Google Ads API, Ezoic

  ETL Engine: Custom Python ETL tool

  Database Layer: PostgreSQL on DigitalOcean

  Scheduler: Cron inside Docker container for timed runs

  Deployment Infrastructure: DigitalOcean droplet running Docker

  Monitoring & Alerts: Slack integration with job status logging and error alerts

Deliverables

  API integrations with Google AdX, AdSense, Ezoic, and Google Ads

  Python-based ETL pipeline to ingest ad campaign performance metrics

  Normalized PostgreSQL database schema for unified data structure

  Scheduled data updates (6-hour intervals)

  Dockerized deployment on DigitalOcean

  Slack alerts for operational transparency and failure detection

Tech Stack

  • Tools used

DigitalOcean, Docker, Slack, Postman

  • Language/techniques used

Python, REST APIs, ETL orchestration, Cron scheduling

  • Models used
  • N/A
  • Skills used
  • API integration, ETL development, Dockerization, Monitoring, AdTech Analytics
  • Databases used
  • PostgreSQL
  • Web Cloud Servers used
  • DigitalOcean (Droplet hosting ETL and DB)

What are the technical Challenges Faced during Project Execution

  Navigating different authentication mechanisms and response formats of multiple ad platforms

  Unifying disparate metrics from different platforms into a single data model

  Ensuring consistency and freshness across time zones and scheduling conflicts

  Implementing monitoring and fault tolerance in long-running ETL jobs

How the Technical Challenges were Solved

  Developed a standardized API wrapper layer in Python for all platforms

  Mapped each platform’s metrics to a common schema for unified reporting

  Used Docker + Cron to schedule updates without OS-level dependency

  Built Slack-based logging with detailed job tracking and real-time alerts for transparency

  Introduced robust error handling and retry mechanisms within the ETL pipeline

Business Impact

  95% reduction in manual time spent on aggregating ROAS and ad performance

  Daily operational efficiency and faster decision-making from fresh, centralized data

  Scalable ETL setup that can incorporate new ad platforms in the future

  Increased data accuracy and stakeholder confidence in marketing KPIs

Project Snapshots

Backend System/Confidential code

RELATED ARTICLESMORE FROM AUTHOR