Readings for 9/11/18

For class today, we had to read chapters 1 and 11 of Marketing Analytics by Mike Grigsby. Chapter 1 was an introduction to marketing analytics and market science, while chapter 11 focused on A/B testing and just market tests in general.

Chapter 1 is the author introducing market science, and his take on it with his beliefs of why it is so important. One thing I found very interesting is in one of the first paragraphs, Grigsby says, "Any ‘marketing’ that is not about consumer behaviour (understanding it, incentivizing it, changing it, etc) is probably heading down the wrong road"(Grigsby). This stuck with me because as a marketing student that makes a lot of sense to me and every business's marketing should be based on those behaviors. Another reason I found interest in that quote is because it led me to my first question which is, if it is super obvious why marketing science is more efficient and leads to more answers, why is not every business using it at this point? If chances of making a wrong decision can be diminished or decreased, then I feel that every business that promotes and markets themselves should be using it. 

Another thing I found interesting about the assigned readings were the authors opinion about testing in chapter 11. The author makes it clear that he himself isn't a big tester, and doesn't always like the results they give. He goes more in depth to say that sometimes the test ends up getting "dirtied" where the test consumers receive stimuli they weren't supposed to. Grigsby describes the sometimes often testing error and says that the number one rule is only one thing can be different in measuring test vs control. This leads to my next question, which is if testing could go bad at any point and happens somewhat often, why would it still continue to be a go-to idea for any business? I didnt seem to understand where if their could be common error, why companies would still use it. 

The final thing I found interesting on chapter 11 was the recap on A/B testing. We've already talked about this in past classes, and the book touches up on it and makes it even more clear on how it works. I was interested that the point of A/B testing is to only change is that the test cell has treatment while the other cell doesn't. I found interest in this because that information then gets tested and if their is a significant difference between information then it becomes worth it to test at a larger population. My final question to this weeks readings is chapter 11 touches on both lift measures, and A/B testing, which one is better and why? This was unclear to me after reading through because it seems like the author favors A/B testing more, but why?