Member Only Access
Summary
Randomized experiments are widely viewed as the gold standard for measuring advertising incrementality, yet advertisers run them far less often than theory would predict. This MSI working paper develops a game-theoretic model of online advertising to explain why. By explicitly modeling the fixed and opportunity costs of experimentation—and the strategic role of ad platforms in setting reserve prices—the study shows how platforms may rationally discourage experimentation even when learning ad effectiveness could improve allocation efficiency. The findings offer a new economic explanation for persistently low experimentation rates in digital advertising markets.
