After reading the notes on the power distribution, I read further into the topic by looking at one the of references. What fascinated me about it, was the practicality of it in the business world, obviously walmart would pick a sales distribution like the normal that would die off quickly rather than, a power distribution and limit their sales to certain threshold. My question is how would they pick this point and know when to limit it? Trial and error, some sort of advanced statistics? Why not buy it from someone who knows?

It would be very profitable for the companies that actually observe the entire demand distribution to sell their market data, even if overall it is much smaller, it is much more relevant as it provides data that other larger retailers can not possibly have access to. Instead of picking the threshold at which to stop providing goods and determining by actual sales if it is correct, why not use some else’s more complete data and save money. Netflix does not have much market collusion with Walmart, unless walmart expands a lot more online. It would be very profitable for them to sell such data, without any individual user information to a large retailer.

In addition, it probably is possible to use the pareto distribution for some minimum value, and actually fit data more practically for this application.