Overview
This series explains why a digital advertising CPM is a reported outcome, not a causal explanation by itself. It traces nominal and effective bids through eligibility, ranking, reserves, payment, delivery, and measurement; then examines how targeting changes matching and competition. It separates the MRC's transparency framework from IBPA's formal model and counterfactual simulation, preserving the evidence limits on production and advertiser claims.
Learning outcomes
- A reported CPM is an outcome over a defined population, not a complete auction explanation.
- Cost diagnosis must separate bids, effective values, qualified competition, ranking, payment, and measurement.
- Targeting can improve matching while thinning competition, so its revenue effect is conditional.
- IBPA's gains are model-based and simulated; they do not establish production costs or universal advertiser benefits.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
What does a CPM actually tell you in a digital ad auction?
It tells you a reported cost per impression unit under a specified definition, population, time window, and billing treatment.
Why can the same nominal CPM produce different clearing prices?
The same nominal CPM can produce different clearing prices because eligibility, bid factors, reserves, payment rules, fees, discounts, and pacing can differ.
Can more targeting lower ad-auction revenue?
Yes. More granular targeting can lower revenue when improved matching is outweighed by fewer credible competing bidders; it can also raise revenue.
What do IBPA's reported revenue gains mean?
The reported 91% publisher-revenue gain is a counterfactual simulation relative to full-disclosure GSP in one retail-media category, not a production invoice or universal forecast.