From the course: Marketing Attribution and Mix Modeling
Unlock the full course today
Join today to access over 23,400 courses taught by industry experts.
Data-driven attribution and machine learning
From the course: Marketing Attribution and Mix Modeling
Data-driven attribution and machine learning
- [Instructor] Data-driven models are the most sophisticated, but also the most robust and accurate models available. Most other attribution models require human intervention to calculate and update. Data-driven models work in the background automatically, and therefore can be less biased. We take the example of a user who read some of your blog posts and then signed up for email lists, got an email and came through to the website, but they didn't want to purchase straight away, and a few days later they saw an ad, clicked on the ad and purchased. On the last click, 100% of the credit would go to the ad, even though the email and the blog also played a big role. With data-driven, the credit would be shared based on the uplift that that channel drives and that's figured out algorithmically using machine learning. So it might find that the ad actually deserved a decent amount of credit, but also the blog deserved a lot of…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
Last-click attribution: The default model1m 54s
-
(Locked)
Time decay and conversion lags2m 42s
-
(Locked)
Linear attribution: Treating all touches equally2m 14s
-
(Locked)
First-click models: From awareness to acquisition2m 41s
-
(Locked)
Position-based models and assigning credit2m 14s
-
(Locked)
Data-driven attribution and machine learning2m 4s
-
(Locked)
Click windows and view-through conversions3m 37s
-
-
-
-