Researchers asking whether policy is affected by interest groups, social movement organizations, and other types of advocacy organizations usually expect the answer to be yes. More often than not, their expectations are disappointed. This paper considers why, considering whether the impact of advocacy is affected by which issues and polities are studied, how advocacy and policy are measured, the amount of advocacy, status quo bias, and the connection between advocacy and public opinion. Through a research synthesis of work on advocacy and policy, the paper finds that estimates of impact are probably affected by measurement, the paucity of advocacy on many issues, and whether public opinion is considered together with advocacy; but probably not by the selection of issues and polities or status quo bias. The findings lead to suggestions for improvements in research design that would enhance our understanding of the links between advocacy and policy.