Skip to content
Discussion options

You must be logged in to vote

You can see how Meridian models the carryover effect here. It is modeled using the Adstock function. At a high-level, a media channel has a composite "impression stock" at a given time, say, t. And this stock is a weighted-average of impressions from time t going back L periods (where L is the chosen max_lag parameter). The default value for this max_lag parameter is 8 weeks.

Within the Adstock function, the effect of media in previous periods decays at rate "alpha" (the geometric decay rate, as described here). Media from periods further back in time than L periods do not, by definition, have an effect on the KPI/outcome in the current period.

You can use functions such as the one here t…

Replies: 1 comment

Comment options

You must be logged in to vote
0 replies
Answer selected by yongqiangzhangzazzle
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
question Further information is requested
2 participants