I have a bunch of sales data of similar products and I want to know which product might be the best popular one.

However, for the same category of items, there are cheap ones and expensive ones, and the sales volume of the cheap ones are always higher than the expensive ones.

For instance, we have some ceiling lamps which cost 45\$, and some are 200\$. If the sales volumes for the 45\$ ones and 200$ are 200 and 180 respectively, even the cheap one is sold more than the expensive one. We know that the expensive one is actually more popular then the cheap one.

How can I model the popularity by this kind of information?


1 Answer 1


We know that the expensive one is actually more popular then the cheap one

What makes you think that? Apparently you have a reason for that, so start from there, find those attributes - is it high customer satisfaction? positive reviews? repurchase? low retention?

Gather those attributes together, evaluate them on all products to see what are the popular ones. It'd help to have a proper definition of popularity.

Sales volume can help too, give it a weight to balance the price difference.

Once you have all this in one table, you can start playing with prediction, clustering, trends etc.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.