Abstract
Marketers for decades have been using Marketing Mix Modeling (MMM), which is performed market by market and the results derived at a national level by attributing more ROAS to those media over-delivering audiences (or investing the most money per capita) in the same market/weeks as those in which the greatest sales increases occurred. This procedure is intensive in terms of assembling the right data, and data errors are known to occur. A simpler procedure would be to do the same type of analysis at a national level, extending the length of time measured so as to maintain a large...