I feel like i might be massively over complicating things here:
At first i thought i could do it with a pivot table but that didn't work out how i expected, then i thought i could get around it by doing a sumproduct but that didn't work out either..
Background
Imagine you have 1000s of customers some of which will have an iphone, ipad, apple tv, some will have samsung phone, ipad, sony tv, ps4, some will have pens and paper only.. I need to find out the propensity to have X product if they already have Y product (in real life this is across thousands of different products).
See attached which gives an example of the data and supporting information.
DATA LOOKS LIKE:
CUST / PROD A B C D E F G H I J
JOE BLOGGS 1 A E H
JOE BLOGGS 2 B C D E F G H I J
JOE BLOGGS 3 B D F H
JOE BLOGGS 4 I
JOE BLOGGS 5 D I
JOE BLOGGS 6 B F H
JOE BLOGGS 7 C F H J
JOE BLOGGS 8 B E H
JOE BLOGGS 9 A D G
JOE BLOGGS 10 C F H
I basically need to forget about the cust and plot out the prod letter relationships with each other if that makes sense?
i'm not totally sure on the output but i envisaged something like the below:
I feel like i need to re-format the data first of all but i'm not sure the best way to go about it.
PVP A B
A 2
B 4
C 1
D 1 2
E 1 2
F 3
G 1 1
H 1 4
I 1
J 1
Apologies for the terrible explanation, struggling to verbalise this hopefully the attachment should help.
Any help would be greatly appreciated!!
Bookmarks