15. method described in Section 1.1. As we saw in that section, it is helpful to have data on sales promotions
when pursuing this approach.
jaké výsledky bude mít průmysl?
Since we do not have such data, we tried an alternative fixed-effects modeling approach. That is, we
we assume that the short-term and seasonal lags are the same across all makes and that the differences
in sales volume by make can be captured by an additive fixed effect.
Chevrolet Chevrolet
Sales
!20
300000
300000
Google Trends
!24
250000
250000
!28
Sales
Sales
!32
200000
200000
!36
150000
150000
!40
2004 2005 2006 2007 2008 2004 2005 2006 2007 2008
Time Time
Toyota Toyota
45
Sales
35
Fitted
200000
200000
25
Sales
Sales
15
160000
160000
5
!5
120000
120000
!15
2004 2005 2006 2007 2008 2004 2005 2006 2007 2008
Time Time
(a) Sales vs. Google Trends (b) Actual & Fitted Sales
Figure 2.3: Sales and Google Trends for Top 2 Makes, Chevrolet & Toyota
36. Poučení na závěr
Pokud se naučíme kolektivní vědomí internetu
poslouchat a dobře se ho ptát, získáme tím neustálý
proud inovací pro náš bussiness, ať už je jakýkoli.