## Need a methodology to predict car sales from various showrooms in a district

Data Assumptions:
The sales on a particular day is always 60
There are 3 showrooms in the district
There are other factors too(I have omitted the others for a simple understanding of the crux of the problem)

Data availability:
Daily sales of all showroom in the district (Data is available for 1 year(not shown in the example)
Types of cars available in each showroom on a daily basis(eg: Mercedes, Ford, etc)
Popularity of each type of car on a daily basis

Problem:
Let us presume that the sales depends on just 2 factors:
• Trend/popularity of the showroom(internal factors)
• Availability of popular cars in the showroom(external factors)

Questions I would like answers to:
• Most analytical tools remove insignificant variables. I do NOT understand the reason for this, since small variations(insignificant)/outliers are also responsible for the change in the output variable.
• Would I be using the same model equation to predict all 3 showroom sales OR would each showroom have different models(Since each showroom might react in a different fashion due to some intrinsic factors which are not captured in data)
o If I were to use different models, there is a chance that my overall sales on a particular day surpasses the possible sales(60).
• Please suggest a methodology for the same
Date Showroom ID Sales
1-Jan 1 10
1-Jan 2 35
1-Jan 3 15
2-Jan 1 25
2-Jan 2 16
2-Jan 3 19
3-Jan 1 32
3-Jan 2 7
3-Jan 3 21
4-Jan 1 ?
4-Jan 2 ?
4-Jan 3 ?