DIY Business Intelligence Tools Cut Costs for Facebook Google Ads

Published: 01/04/2026
BI System Advertising Data Low Cost
This paper proposes a low-cost approach to building a self-built BI system for integrating Facebook and Google advertising data. By leveraging Deferred Deeplink and S2S integration technologies, advertising campaign information is obtained and combined with a self-built BI system for data analysis, enabling ad performance tracking and optimization. The paper also discusses data attribution and duplicate installations, proposing corresponding solutions to address these challenges. This approach offers a cost-effective way to gain insights from advertising data and improve campaign effectiveness.
DIY Business Intelligence Tools Cut Costs for Facebook Google Ads

As marketing costs continue to rise, small and medium-sized businesses along with independent advertisers face the pressing challenge of effectively tracking and analyzing Facebook and Google advertising data with limited resources. This article presents a low-cost approach to building an in-house Business Intelligence (BI) system, enabling basic data statistics and analysis without relying on third-party Mobile Measurement Platforms (MMPs).

Core Technical Logic

The solution focuses on integrating advertising data from Facebook and Google into a custom BI system through technical means. The approach consists of two key components:

1. Facebook Ad Data Collection: Deferred Deeplink Technology

Deferred Deeplink directs users to specific pages or actions after app installation. This technique enables the collection of campaign, ad set, and ad information from when users initially clicked the advertisement.

Implementation Process:

  • Ad Launch Phase: Record each campaign, ad set, and ad ID along with their names during Facebook ad creation.
  • Deeplink Generation: Create unique deeplinks containing parameters like sourceid=123456789 , where sourceid serves as a random unique identifier.
  • Client-Side Data Collection: After ad click and app installation, the client retrieves device identifiers (IDFA or Android ID) along with the deeplink's sourceid.
  • Server-Side Data Association: The client transmits device information and sourceid to the server, which correlates this data with previously recorded campaign details.

Implementation Options:

  • API Solution: Automate ad deployment and data recording through Facebook Marketing API for technical teams.
  • Excel Solution: Manually maintain an Excel spreadsheet encoding campaign details into numerical identifiers for basic tracking.

2. Google Ad Data Collection: Server-to-Server (S2S) Integration

Google's S2S integration allows direct data retrieval from Google Ads servers without third-party MMP dependency.

Integration Process:

  • Obtain an MCC Developer Token
  • Create a Link ID to associate Google Ads accounts with the BI system
  • Apply for Google Ads API access
  • Develop a data receiver following Google's S2S documentation

Technical Considerations:

  • Requires API development and data processing expertise
  • Developer token and API approval may involve significant lead time
  • MCC account information must be accurately provided during Link ID creation

Data Attribution and Reinstallation Challenges

Custom BI systems must address data attribution and reinstallation issues without advanced MMP capabilities.

1. Data Attribution Approaches:

  • Last-Click Attribution: Credits the final clicked ad before installation
  • Click Timing Analysis: Evaluates time intervals between clicks and installations
  • Referral Analysis: Leverages Android's referral mechanism for source identification

2. Reinstallation Handling:

  • Ignore Reinstalls: Treat reinstalls as new installations
  • Differentiate New vs. Reinstalls: Track device IDs to distinguish between first-time and returning users
  • Practical Compromises: Accept reasonable data inaccuracies given the cost-saving objectives

Custom BI System Design

An effective BI system should prioritize data organization, analysis, and presentation with these characteristics:

  • Fast loading speeds for large datasets
  • Intuitive user interface for non-technical users
  • Practical analytical functions for core metrics

Key Functional Requirements:

  1. Channel/Campaign Data Visualization:
    • Bar charts displaying installs, activations, and revenue by channel
    • Customizable date ranges for temporal analysis
    • Drill-down capabilities to ad set and ad levels
  2. Event Data Analysis:
    • Definition of key events (registrations, purchases, onboarding completion)
    • Conversion rate calculations by campaign
    • Funnel visualization to identify conversion bottlenecks
  3. Retention Analysis:
    • Retention rate calculations (day 1, day 7, day 30)
    • Retention curve visualization
    • Cohort analysis (paid vs. non-paying users)

Technical Challenges and Considerations

Building a robust BI system requires careful planning around data architecture and infrastructure:

  • Data Storage: Choose between relational (MySQL, PostgreSQL) or NoSQL (MongoDB, Cassandra) databases based on data structure needs
  • Data Analysis: Implement SQL for basic queries or MapReduce for large-scale processing
  • Data Visualization: Utilize tools like Tableau, Power BI, or Echarts for comprehensive reporting

Additional Considerations:

  • Implement data security measures including encryption and access controls
  • Establish data quality monitoring to identify and correct errors
  • Plan for continuous system optimization based on evolving needs

Conclusion

Custom BI systems offer a budget-friendly alternative for small businesses and independent advertisers to analyze advertising performance. While lacking the sophistication of commercial MMPs, these solutions provide sufficient functionality for optimizing ad strategies. Successful implementation requires balancing technical capabilities with practical business needs while maintaining focus on core analytical requirements.

Related Topics