Data-Driven Business Decisions

CHAPTER 12

What Is My Brand Recognition? Will It Sell? Analyzing Counts and Proportions

12.0. Introduction: What Is The Issue?

How do the purchasing habits of male and female customers differ? The standard way to look at this would be to compare the mean purchase of males and females using the methods of the previous chapter. The mean, though, does not tell you much about extremely high and low purchasers and how this might relate to gender. One way to look at this would be to divide the data into different levels of purchase—for instance the highest 10%, the next highest 10%, and down to the lowest 10%—and then look at the numbers of males and females in each group. The key numbers, then, would be the percentages of males and females. Gender is not a number, though. It is an attribute.

There are many business variables that are non-numeric. Rather, the variable is something that is true or false, that is present or absent, and that requires a yes or no answer. Gender is one example but there are many others. Can customers name your brand without prompting? Are they satisfied? Do they prefer Coke to Pepsi? Was the item defective? Did the seller of a car secure the servicing contract? Did my product sell at this price? Do you agree with the proposition? This chapter is about analyzing this kind of non-numeric “yes–no” data.

The analysis of “yes–no” data is slightly different from the analysis of a numeric outcome such as revenue or sales. For a start, it does not immediately make ...