There are many studies on just this thing. Primarily examine mass communications journals if you're interested in scholarly research. But the two most common:

There are many studies on just this thing. Primarily examine mass communications journals if you're interested in scholarly research. But the two most common: 1) Focus groups. Take a group of people, have them fill out a questionnaire, watch a group of ads, then answer the questionnaire. These have become more effective over time as the result of methodological research on research bias. In the "early days" people were given a set of questions, watched THE AD, and then answered the same set of questions. It turns out, though, that those questions 'primed the pump' and skewed results when compared to a more neutral setting where there are many questions about many products and then many ads. 2) Market research. Take two representative regions, deploy one advert in one location and another at the other, compare and contrast. Again, this has improved over time where it used to be "nothing vs. something", but that really showed the effectiveness of advertising, not the result of a specific campaign.