The manuscript presents a novel experimental methodology for studying the impact of online advertising, tracking, and targeting on consumer behavior. Existing approaches rely heavily on observational data, which suffer from self-selection and limited causal identification, or platform-specific experiments, which provide strong internal validity but cannot capture the broader consumer journey across the web. Laboratory and survey studies, while controlled, lack ecological validity and temporal depth. The methodology resented here addresses those limitations by deploying a client-server infrastructure that enables randomized interventions in advertising exposure, tracking, and argeting, while collecting rich, longitudinal data across multiple platforms and devices. The methodology integrates multiple captured behaviors (including browsing, advertising exposure, shopping and search, purchase confirmation, promotional emails) with periodic self-report surveys. Rigorous privacy and security safeguards ensure responsible handling of sensitive data. The manuscript illustrates the capabilities of the methodology with a demonstration experiment that assigns participants to full behavioral advertising exposure, ad-blocking, or anti-tracking conditions over three months, allowing tracing the effects of interventions on product search patterns, purchases, and satisfaction. Beyond consumer welfare, the methodology can support a wide range of inquiries into media consumption, vendor diversity, exposure to political or health-related advertising, interactions with large language models, and more.