First of all many thanks to all the people that made me aware of the article, I’m touched that you thought of me to send it to. Yes it was of interest and I’ve read it more than a few times.
Apart from being a very thinly veiled plug for Charles Duhigg’s book The Power of Habit: Why We Do What We Do In Life and Business it’s a good pointer to where basket analysis should have gone, but ultimately didn’t.
What some people have failed to notice is that this all started in 2002 so there’s been 10 years of customer data analysis going on. In 2002 though the first main key of any exercise was formed, they started with a question.
“If we wanted to figure out if a customer is pregnant, even if she didn’t want us to know, can you do that? ”
The first immediate thing is there is a customer segment, pregnant mothers, nothing more nothing less. From there though the work starts, though in this instance I’m pretty convinved that you don’t need a maths nerd to figure this stuff out. You need domain experts, back to the people who know retail well.
Isolating the products that a pregnant mother would usually buy is the next step, as outlined in the article, this really is about product knowledge and knowing what people would instinctively go for.