If a company is trying to compare how toy sales perform through the year, then it might not want to use a straight comparison for January and December - the fact Christmas is at the back end of December is going to impact on sales then in a big way.
Or a convertable car dealership - a hot and sunny July is likely to see far more cars sold than a cold, wet October.

The "straight up numbers" are pretty meaning less for judging performance unless you do performs "tricks" on them - sales against money spent on marketing, sales as a percentage of sales for that month in previous years, how the rest of the market is going etc. You have to manipulate the figures to be able to compare performance.