Channel analytics based on surveys
As the name suggests, in this approach, we typically design and conduct a market research surveys in which a uniform mix of candidates are chosen to take the survey. The advantage with this method is that we can control the representation of certain sections of the population by oversampling or under sampling techniques.
However, this approach heavily relies on assumptions around the absence of any kind of bias in the survey questions and sample. Also, since this data is self-reported, there might not be a concrete method to verify the actual truth.