This thesis investigates transparency, fairness, and data accuracy in the online advertising ecosystem through large-scale analysis of data from hundreds of thousands of internet users. The research addresses three key challenges: social faultlines in digital marketing, gender-based price discrimination, and location data accuracy in advertising campaigns.

The research aims to:

1) Develop a methodology to measure social faultlines using digital traces from online platforms;

2) Investigate gender-based price discrimination in online advertising;

3) Assess the accuracy of location data used in advertising campaigns.

Key Findings:

Social Faultlines – The research introduces a novel methodology to measure hypothetical societal dividing lines using digital traces from social media. By analyzing Facebook marketing data, the study demonstrates how to identify social faultlines at unprecedented scale, providing valuable insights for policymakers to address potential social conflicts.

Digital Marketing Pink Tax – Analysis reveals significant gender-based price discrimination in online advertising. Data from Facebook’s advertising platform shows advertisers often pay more to reach female consumers, particularly in developed countries, raising important questions about fairness in digital marketing practices.

Location Data Accuracy – The study uncovers that up to 90% of location data used for advertising is suspicious, significantly reducing campaign effectiveness. The research proposes a new algorithm to assess location data quality and identifies common sources of error in location-based targeting.

Contributions and Impact:

Key contributions include: First methodology to quantify societal faultlines at scale using social media data; Comprehensive evidence of gender-based economic discrimination in digital marketing; Pioneering algorithm to assess location data quality in advertising. These contributions enhance understanding of the online advertising ecosystem and propose practical solutions to improve transparency, fairness, and effectiveness in digital marketing.

Conclusions:

The findings highlight significant challenges in online advertising regarding data quality, transparency, and fairness. By addressing these issues, this research aims to foster a more equitable advertising environment. The methodologies developed offer valuable tools for advertisers, platforms, and regulators to improve practices and policies in digital marketing, balancing the unprecedented targeting capabilities with ethical considerations to ensure the ecosystem functions optimally for all stakeholders.

About Amir Mehrjoo

Amir is a PhD Student in the NETCOM Lab Research group at IMDEA Networks. This group focuses in 5G networks and SDN-based architectures. He earned his Bachelor’s degree in Mechanical Engineering at the University of Shiraz in Iran. After that, he moved to Madrid to achieve in the University Carlos III, a MSc in Business and Finance with a Marketing Specialization.

Supervisor de tesis: Dr. Rubén Cuevas Rumin, Universidad Carlos III de Madrid, Madrid, España

Universidad: Universidad Carlos III de Madrid

Programa de doctorado: Ingeniería Telemática

Miembros del tribunal:

• Presidenta: Andra Lutu, Senior Researcher. Telefonica Research. Madrid, España
• Secretaria: Patricia Callejo, Visiting Professor. University Carlos III of Madrid. Madrid. España
• Miembros del panel: Ignacio Castro Arribas, Lecturer. Queen Mary University of London. Reino Unido