In: Operations Management
A convenience store recently started to carry a new brand of soft drink. Management is interested in estimating future sales volume to determine whether it should continue to carry the new brand or replace it with another brand. The following table provides the number of cans sold per week. Use both the trend projection with regression and the exponential smoothing (let
alphaαequals=0.40.4
with an initial forecast for week 1 of
581581)
methods to forecast demand for week
1313.
Compare these methods by using the mean absolute deviation and mean absolute percent error performance criteria. Does your analysis suggest that sales are trending and if so, by how much?
Period |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
Sales |
581581 |
599599 |
642642 |
747747 |
674674 |
625625 |
729729 |
712712 |
754754 |
728728 |
671671 |
745745 |
(i) Obtain the trend projection with regression forecast.
Obtain the trend projection with regression forecast.
The forecast for week 13 is
(Enter your response rounded to the nearest whole number.)
Specify the mean absolute deviation (MAD) and mean absolute percent error (MAPE). (Enter your responses rounded to two decimal places.)
MAD |
MAPE |
(ii) Obtain the exponential smoothing forecast.
The forecast for week
13 is
(Enter your response rounded to the nearest whole number.)
Specify the mean absolute deviation (MAD) and mean absolute percent error (MAPE). (Enter your responses rounded to two decimal places.)
MAD |
MAPE |
Based on MAD, the best method is
trend projection with regressiontrend projection with regression
.
Based on MAPE, the best method is
trend projection with regressiontrend projection with regression
.
Answer i)
Trend with regression:
n |
At |
Ft |
et = At-Ft |
|et| |
|At-Ft|/At |
Period |
Sales |
Forecast |
Errors |
Absolute |
Absolute |
1 |
581581 |
620941 |
-39360 |
39360 |
6.77% |
2 |
599599 |
632516 |
-32917 |
32917 |
5.49% |
3 |
642642 |
644091 |
-1449 |
1449 |
0.23% |
4 |
747747 |
655666 |
92081 |
92081 |
12.31% |
5 |
674674 |
667241 |
7433 |
7433 |
1.10% |
6 |
625625 |
678816 |
-53191 |
53191 |
8.50% |
7 |
729729 |
690391 |
39338 |
39338 |
5.39% |
8 |
712712 |
701966 |
10746 |
10746 |
1.51% |
9 |
754754 |
713541 |
41213 |
41213 |
5.46% |
10 |
728728 |
725116 |
3612 |
3612 |
0.50% |
11 |
671671 |
736691 |
-65020 |
65020 |
9.68% |
12 |
745745 |
748266 |
-2521 |
2521 |
0.34% |
13 |
759841 |
||||
MAD |
MAPE |
||||
32406.75 |
4.77% |
Trend
In MS Excel,
Select the sales data
Click on Insert and then “Line”
On the line graph, do right click and select “Add Trendline”
Then from the menu, click on the “Display Equation on chart” checkbox.
y = 11575x + 609366
where,
Y = Ft |
x = t |
Using it,
Forecast for period 1:
y = 11575x + 609366 = 11575*1 + 609366
Forecast for period 13:
y = 11575x + 609366 = y = 11575*13 + 609366 = 759841
MAD (Mean Absolute Deviation) = Sum of all Absolute Errors / Total number of period = 388881/12 = 32406.75
MAPE (Mean Absolute Percentage Error) = (Sum of all Absolute % error / Total number of period) =57.27% / 12 = 4.77%
Answer ii)
Exponential Smoothing:
n |
At |
Ft |
et = At-Ft |
|et| |
|At-Ft|/At |
Period |
Sales |
Forecast |
Errors |
Absolute |
Absolute |
1 |
581581 |
581581 |
0 |
0 |
0.00% |
2 |
599599 |
581581 |
18018 |
18018 |
3.01% |
3 |
642642 |
588788 |
53854 |
53854 |
8.38% |
4 |
747747 |
610330 |
137417 |
137417 |
18.38% |
5 |
674674 |
665297 |
9377 |
9377 |
1.39% |
6 |
625625 |
669048 |
-43423 |
43423 |
6.94% |
7 |
729729 |
651679 |
78050 |
78050 |
10.70% |
8 |
712712 |
682899 |
29813 |
29813 |
4.18% |
9 |
754754 |
694824 |
59930 |
59930 |
7.94% |
10 |
728728 |
718796 |
9932 |
9932 |
1.36% |
11 |
671671 |
722769 |
-51098 |
51098 |
7.61% |
12 |
745745 |
702330 |
43415 |
43415 |
5.82% |
13 |
719696 |
||||
MAD |
MAPE |
||||
44527.32 |
6.31% |
Exponential smoothing Forecast formula:
F (t+1)= Ft+α(At-Ft)
F (t+1) is Forecast for the current period
Actual Sales (At)
Forecast Sales (Ft)
Smoothing constant = α = 0.4
Forecast for period 3 = 581581+0.4*(599599-581581) = 588788
Forecast for period 13 = 702330+0.4*(745745-702330) = 719696
MAD (Mean Absolute Deviation) = Sum of all Absolute Errors / Total number of period = 534328/12 = 44527.32
MAPE (Mean Absolute Percentage Error) = (Sum of all Absolute % error / Total number of period) =75.70% / 12 = 6.31%