It is very difficult to see the impact of product breadth via digital analytics. Very, very difficult.
Here's how a lot of this stuff works. Let's say you only offer one product line (Widgets). Last year, you had 100 customers purchase widgets. In the next twelve months, of the 100 customers who purchased widgets, 40 purchase again. Your job, as a marketer, is to find at least 60 new+reactivated buyers to keep widgets moving forward.
Somebody in Management, or somebody in your merchandising organization ... somebody decides it is time to sell Wudgets.
It is here, now, that things get interesting.
No longer do 40 of 100 widget buyers purchase widgets. Nope. Now, 25 of 100 widget buyers buy widgets. 10 of 100 widget buyers purchase widgets and wudgets. And, importantly, 10 of 100 widget buyers switch over to wudgets. Also interesting ... it becomes harder to find enough new+reactivated widget buyers to fuel the business ... in the short term, because now customers have to choose between widgets and wudgets. Finally, the wudgets product line brings in new customers as well.
The following year, the two product categories find and equilibrium. In most cases, one of the product lines becomes primary, the other secondary. In most cases, one of the product categories does the new customer heavy lifting, while the other product category siphons off customers from the category doing the heavy lifting.
It is the management of these dynamics that is so very interesting. When done well, the impact is multiplicative. When done poorly, the impact results in cannibalization of product categories.
Once you understand the dynamic, you can grow many categories by simply growing one category. When I worked at Nordstrom, the family and team intuitively knew how these dynamics worked - they didn't need my geeky math to tell them that Womens Shoes and Cosmetics brought the customer into the business, satisfying customers that then migrated into Accessories or Apparel or whatever.
Amazon takes product breadth to amazing extremes. They are an example of a multiplicative pattern.
Most retailers are going in the opposite direction ... the omnichannel movement results in significant inter-channel and inter-category cannibalization. Few of the experts understand this dynamic - but it is lying there in plain sight - for if the dynamic were multiplicative, retailers would experience overall 20% gains, year-over-year. Instead, online grows and retail shrinks - certain in-store product categories die as their demand is pushed online - and the slow death experienced there results in fewer customers buying from other in-store categories - resulting in smaller and smaller store formats. Again, too few of the experts understand this dynamic. Almost nobody measures it. And it lies there, in plain sight, waiting to be discovered.