Part 6. Media Investment and Optimization

People walking on a city street at night

Large advertisers spend millions, sometimes hundreds of millions, on media every year. For a $5B brand, even a modest media budget can easily exceed $100M across TV, search, social, retail media, and direct mail.

Even a small shift in how that budget is allocated — just one or two percent — can mean millions in business impact.

This is the core question in media measurement: not which ad drove a click, but where the next dollar should actually go.

Platform ROAS reports and multi-touch attribution (MTA) help with campaign management, but they are not built for budget allocation across channels, formats, or seasons.

In this part, I’ll walk through the modern measurement stack. Attribution explains what we observe in customer journeys. Media Mix Modeling (MMM) estimates how each channel contributes using aggregate data. Experiments provide the causal validation that ties it all together. When used together, these tools move media measurement from simple reporting to a system that actually supports budget decisions.

After this part, you will be able to:

Place attribution, MMM, and experiments in the measurement stack
Build and interpret a Media Mix Model
Allocate media budgets using response curves and experimental calibration
Calibrate MMM with experiments and run reliability checks before using the model for budget decisions