Question

Example of a sales forecast for Amazon.

Answer 1

So Amazon mainly does forecasting on both qualitative and quantitative parameters out of this if you have a deeper look into the qualitative parameters we find that regular sales service sales surveys and consumer feedback is taken from emails and prospective visitors of their website why the qualitative technique is to some extent used in understanding the perception of the buyers about the products and the listings the main technique that Amazon uses in order to accurately forecast the financial parameters of the sales like the revenues is the quantitative techniques known ARIMA.
Autoregressive integrated moving average method
In this method the moving averages taken of a sequence of previous period based on the sales and revenue data and they are regressed on a moving an of a sequence of previous period based on the sales and revenue data and they are regressed With respect to any other dependent variable in order to forecast the expected sales for the next year.

The variable on the LHS are the predictors in the model.

And the RHS variables are independent variables on which the sales depends,

Like a1 = seasonality

Yn= historical sales.

a2= customer feedback received earlier

The moderation is done using a moderator variable like e1 . Moderation is basically done to eliminate the negative risks in having forecast error.

Like an item having higher sales return may have a probability of Lower future sales

So if we have,

a1 = 1.5 ( peak sales season)

Previous period sales = 60mill

a2 feedback factor = 0.90 (90% positive feedback)

And e1 with (-0.05) 5% sales return,

We have forecasted sales

ARIMA Yf =(1.5x60)x0.90-( 0.05x60 )=