Clearly, the ability to do sums and averages with large amounts of business data is important. However, our previous examples have one significant limitation—the aggregated function operates on the entire table. Thus, we can see aggregation at one level, but if we want a more granular look at the data, we are forced to do a high degree of manual work breaking the data apart. For instance, we've seen an example of finding the average number of logins for all employees using the AVG() function. But, what if we wanted to see those averages on a different level? Say, the average number of logins differentiated between male and female employees? One way to do this is by utilizing two separate queries, using a WHERE clause to differentiate them. Examples of this are shown in the next two screenshots:
Thus, we see the average logins by employees, differentiated by gender. While this is serviceable in this example, what if our criteria for grouping was more complex? Let's say, for example...