[Video] Digital Analytics Meets Data Science

As I continued my exploration of R and statistics, I got interested in figuring out what I could do with “standard” digital analytics data. That resulted in this presentation at Superweek in 2019 that included three examples, all of which were accompanied by Shiny apps that participants could use to explore their own Google Analytics data:

  • “Time-normalizing” of web traffic–a straight-up munging of the traffic of a filtered set of web pages (such as blog posts) so that their launch dates were individually transformed to a “time zero”
  • Exploration of On-Site Search–demonstrating some basic text analysis, including unnesting of phrases, stemming, and LDA
  • Linear Regression with Categorical Variables–really, an explanation of how categorical/nominal variables get hot-encoded to dummy variables, and then how to interpret the results of a linear regression