Dear Management Analytics Consultant:
Here's a problem. I have a segment of customers who love buying Men's Footwear. It's their preferred category. However, when I slice and dice the data, I observe that the category the Men's Footwear buyer is most likely to purchase next is Women's Dresses. This kind of makes sense, because a third of our business is Women's Dresses.
So how should I target this customer? Does this customer get Men's Footwear, other Men's Categories, or Women's Dresses?
This isn't an "optimization" problem. This is a "marketing strategy" problem. You are a marketer. It's your job to have a strategy.
Here's a tip. In the vast majority of my projects, the most valuable customers (after equalizing for recency/frequency/monetary/channel-preference) are those who buy from multiple merchandise categories.
This means you need to have a multi-pronged strategy for communicating to this customer. Let's use email as an example. This might be a "personalization strategy" for this customer, via email marketing:
- Monday = Primary Category (Mens Footwear).
- Tuesday = Primary Division (Mens).
- Wednesday = Most Likely Response Category (Womens Dresses).
- Thursday = Weekly Promotional Strategy.
- Friday = New Merchandise Friday.
This means that Monday/Tuesday/Wednesday campaigns exploit "merchandise personalization" based on prior purchasing behavior.