Talk:Demand sensing

Old talk

This article clearly explains the limits of traditionnal states that Demand Sensing uses different data and different mathmatical techniques. It would be helpful to have a description of these data more precise than promotion, launch and stocks.

Morevover, it would be helpful to understand which new mathematical techniques are involved.

Lastly, a reference to an independent study should be added to the statement that DS improves forecast accuracy by up to 30%. CedricHT (talk) 07:16, 3 October 2014 (UTC)[reply]

Written like a marketing advert

There appears to be very little substance backing the bold claims of "demand sensing" compared to the still-very-much-active area of plain "time-series forecasting", in particular, the complete lack of any published peer-reviewed algorithms begs the question of the solidity of any of the claims.

Joannes Vermorel (talk) 16:06, 28 August 2019 (UTC)[reply]

I removed most of the article as unrefferenced and of little direct relevance to the subject. I did some copyvio search and see that most of it was piecewise copied from texts by businesses which are in this business, i.e., it seems you are right, this is a snake oil advertorial. Staszek Lem (talk) 18:47, 28 August 2019 (UTC)[reply]