Marketing mix model with marketing campaigns that are not consistent?

Question

So we got sales data for six months which has 4 different marketing type campaigns running. One of them is in-store which runs for 3 months only. Another campaign is social which has the least spend but runs for 6 months. Other 2 run for 2 & 3 months respectively. If you run linear regression it gives social campaign as significant. How I consider the time range of these campaigns and the spend for each towards my mix marketing model.

My question here is how do you come up with a mix marketing model which considers the factors such as the cost for each marketing campaign and the total time duration when it was run.

Answer 1

Normally when you run a Marketing Mix Model, you are trying to find out the impact of a given $ spend on marketing channel onto your target variable (usually sales).

The main thing here lies in having granular cost information for your marketing campaigns, if you do not have that, you can try to allocate the cost over the period it ran.

You should essentially have a column for each campaign containing the spend for each time period. You can as a second step introduce lagged terms to get a better sense of how past spend as affected your current sales.