I'm trying to wrap my head around the concept of advertising effectiveness and hit a wall when I tried to measure the quantitative effect of an advertising campaign on the sales tendency of a business. As far as I know, this is an econometrics question, since I have isolate the effect of said campaign from past and future campaigns, but also from distribution, macro effects, competitor activity and random effects. Can you point me to a paper or a textbook that deals with this kind of problem?
2
-
$\begingroup$ There's no way around it, for clean identification, you need to have an experiment with randomizing campaigns, for example based on region. $\endgroup$– FooBarMar 13, 2015 at 19:09
-
$\begingroup$ I'm well aware of that, the methodology used to make such experiment and the techniques and tets to detect and isolate the different variables that affect the sales series variance is what I am looking for :) @foobar $\endgroup$– toku_moMar 13, 2015 at 23:53
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
You might take a look at the Causal Impact R package and accompanying paper.
-
$\begingroup$ This seems to be the one, going to check if it holds in multivariate models. Thank you! $\endgroup$– toku_moMar 15, 2015 at 23:46