-
How many of you are using statistics to better your work? How?
I feel like a lot of digital marketers tend to neglect their understanding of mathematics and statistics. I would like to know if and how you use concepts like standard deviation, regression, hypothesis testing etc. In your work.
-
4 Replies
-
I use stats a lot (specifically ANOVA) to report on how COVID numbers are outside of our normal conversion and impression distributions.
-
Hypothesis testing, whether it’s a split test or lift test, is definitely a piece of our marketing strategy.
Split tests help answer the day-to-day optimizations, like ‘do ads with images outside generate more clicks than ads with images inside?’, while lift tests help answer broader questions like, ‘are my ads driving incremental sales?’.
I’ve been learning about Media Mix Modeling and regression definitely plays a role. These types of models require a higher volume of data and years worth of information to be really good, and they’re only applicable for advertisers that meet those requirements.
-
Only at a shamefully rudimentary level. I’d love a marketing stats/experiment textbook to up my scientific applications.
-
I don’t actively do this, but our BI team has automated sheets for that. It’s a standard part of our monthly internal reporting/discussion. Nothing that we share with clients, but a quintessential part for our success.
I can only recommend anyone in PPC to brush up their mathematics. It’s helping a lot in the long run
Log in to reply.