Look at the items from classes prior to 2014 ... those items generated $13.5 million in 2017.
Meanwhile, items from the classes of 2016/2017 only generated $12.7 million in 2017.
In most of my projects, there is what I call a "Merchandising Imbalance". There is a class of historical items that generate a ton of demand ... and new items generate a ton of demand ... with items from recent years being "killed off" too soon.
This creates all sorts of productivity problems that marketers get blamed for. When you don't have enough good items from 2014 - 2016, the "brand" appears to bifurcate ... it looks both old and chaotic at the same time. Existing customers are bored with seeing the same thing month after month for a decade while seeing a bunch of new merchandise that will quickly be discontinued. Boredom and chaos lead to lower conversion rates, and lower conversion rates yield blame ... for the marketer.
You MUST measure this stuff ... you don't want to get blamed for merchandising sins from 2-4 years ago, do you?
P.S.: If you are bored by this stuff, read about Under Armour getting hacked (click here). I'm increasingly confident all of this digital stuff is going to turn out just fine ...