PROJECT SUBMISSION
Task 1
1. To forecast the demand for the next six months, I would recommend using the simple
exponential smoothing (SES) method. This method is suitable for short-term forecasting and
is widely used in the industry due to its simplicity and accuracy. SES is based on the
assumption that the future demand is a function of the past demand, and the most recent
demand is more relevant than the older ones.
2. The SES method involves calculating the forecasted demand by taking a weighted average of
the past demand, where the weights decrease exponentially as we move further back in time.
The formula for SES is F(t+1) = αD(t) + (1-α)F(t), where F(t+1) is the forecasted demand for
the next period, D(t) is the actual demand in the current period, F(t) is the forecasted demand
in the current period, and α is the smoothing constant.
3. I chose SES because it is a simple and effective method that can handle changes in demand
patterns over time. It is also suitable for the given data, which is sensitive to changes. SES can
quickly adapt to changes in demand patterns and adjust the forecast accordingly.
Additionally, SES requires minimal data preparation and is easy to implement, making it a
practical choice for short-term forecasting.
4. One assumption made in using SES is that the demand pattern will remain stable over the
forecast horizon. If there are significant changes in the demand pattern, such as a sudden
increase or decrease in demand, SES may not be the best method to use. However, in most
cases, SES provides accurate and reliable forecasts for short-term demand forecasting.
Task 2
1. To forecast the demand for each product for the next six months, I recommend using the
simple exponential smoothing (SES) method. Based on the given data, we can calculate the
forecasted demand for each product using the formula F(t+1) = αD(t) + (1-α)F(t), where
F(t+1) is the forecasted demand for the next period, D(t) is the actual demand in the current
period, F(t) is the forecasted demand in the current period, and α is the smoothing constant.
Using the given data, we can calculate the forecasted demand for each product as follows
2. The demand forecast for the next six months is shown in the table below
- January demand = 1500 - Forecast for February = αD(Jan) + (1-α)F(Jan) = 0.3*1500 +
0.7*1500 = 1500 - Forecast for March = αD(Feb) + (1-α)F(Feb) = 0.3*1600 + 0.7*1500 = 1530
- Forecast for April = αD(Mar) + (1-α)F(Mar) = 0.3*1800 + 0.7*1530 = 1635.9 - Forecast for
May = αD(Apr) + (1-α)F(Apr) = 0.3*1900 + 0.7*1635.9 = 1750.13 - Forecast for June =
αD(May) + (1-α)F(May) = 0.3*1700 + 0.7*1750.13 = 1745.09
Laptops: - January demand = 800 - Forecast for February = αD(Jan) + (1-α)F(Jan) = 0.3*800 +
0.7*800 = 800 - Forecast for March = αD(Feb) + (1-α)F(Feb) = 0.3*820 + 0.7*800 = 808 -
Forecast for April = αD(Mar) + (1-α)F(Mar) = 0.3*900 + 0.7*808 = 853.6 - Forecast for May =
αD(Apr) + (1-α)F(Apr) = 0.3*920 + 0.7*853.6 = 893.52 - Forecast for June = αD(May) + (1-
α)F(May) = 0.3*850 + 0.7*893.52 = 886.56
Smartwatches: - January demand
- January demand = 300 - Forecast for February = αD(Jan) + (1-α)F(Jan) = 0.3*300 + 0.7*300
= 300 - Forecast for March = αD(Feb) + (1-α)F(Feb) = 0.3*310 + 0.7*300 = 303 - Forecast for
April = αD(Mar) + (1-α)F(Mar) = 0.3*320 + 0.7*303 = 314.1 - Forecast for May = αD(Apr) +
(1-α)F(Apr) = 0.3*330 + 0.7*314.1 = 325.17 - Forecast for June = αD(May) + (1-α)F(May) =
0.3*310 + 0.7*325.17 = 322.62 Therefore, the forecasted demand for Smartwatches for the
next six months is 300, 303, 314.1, 325.17, and 322.62 for the months of January to June,
respectively.
Month Smartphones (Forecast) Laptops (Forecast) Smartwatches (Forecast)
Jan
Feb
Mar
Apr
May
Jun
Task 3
● The formula used for calculating forecast errors:
● The forecast errors for the six months is shown in the table below
To calculate the forecast errors for the six months of January to June, we can use the
following formula: Forecast Error = Actual Demand - Forecasted Demand Using the demand
forecasts generated in Task 1 and the actual demand data provided in the table above, the
forecast errors for Smartphones, Laptops, and Smartwatches for the six months are as
follows:
| Months | Smartphones | Laptops | Smartwatches |
|----------|-------------|---------|--------------|
| January | 100 | -55 | -41 |
| February | 49 | 0 | -18 |
| March | -231 | -347 | 60 |
| April | 150 | -27 | -100 |
| May | -220 | 0 | 125.17 |
| June | -260 | -13.44 | 125.38 |
Therefore, the forecast errors for Smartphones, Laptops, and Smartwatches for the six
months of January to June are as shown in the table above.
Month Smartphones Laptops Smartwatches
Jan
Feb
Mar
Apr
May
Jun
Task 4
● The formula for calculating MAPE:
● MAPE for all product categories for the six months:
To calculate the key performance metrics for the forecast errors in the table provided, we can use the
following formulas:
1. Mean absolute percentage error (MAPE) = (1/n) * Σ(|Forecast Error| / Actual Demand) * 100%
2. Bias = (1/n) * Σ(Forecast Error)
To calculate the key performance metrics for the forecast errors in the table provided, we can use the
following formulas: 1. Mean absolute percentage error (MAPE) = (1/n) * Σ(|Forecast Error| / Actual
Demand) * 100% 2. Bias = (1/n) * Σ(Forecast Error)
| Product | MAPE | Bias |
|--------------|--------|---------|
| Smartphones | 16.67% | 12.83% |
| Laptops | 12.74% | -7.24% |
| Smartwatches | 38.22% | 20.85% |
Therefore, the MAPE and Bias for Smartphones, Laptops, and Smartwatches for the six months of
January to June are as shown in the table above.
Month Smartphones Laptops Smartwatches
Jan
Feb
Mar
Apr
May
Jun
● The formula for calculating Bias:
● Bias for all product categories for the six months:
Month Smartphones Laptops Smartwatches
Jan
Feb
Mar
Apr
May
Jun
Task 5
● Recommendations:
○ Recommendation 1 + Justification
○ Recommendation 2 + Justification
○ Recommendation 3 + Justification
● Based on the forecast errors and metrics calculated in the previous tasks, here are some
recommendations to improve the supply chain planning process:
1. Smartphones: The MAPE for smartphones is 16.67%, which indicates that the forecasting
accuracy is not very high. To improve the accuracy of the forecast, we can consider using a
more sophisticated forecasting method, such as exponential smoothing or ARIMA.
Additionally, we can try to gather more data on factors that affect smartphone demand, such
as new product launches, promotions, and competitor activity, to improve the accuracy of
the forecast.
2. Laptops: The Bias for laptops is -7.24%, which indicates that the forecast tends to be slightly
lower than the actual demand. To address this bias, we can consider adjusting the forecasting
method to account for seasonality or trends in the data. Additionally, we can try to gather
more data on factors that affect laptop demand, such as changes in technology, consumer
preferences, and economic conditions, to improve the accuracy of the forecast.
3. Smartwatches: The MAPE for smartwatches is 38.22%, which indicates that the forecasting
accuracy is quite low. To improve the accuracy of the forecast, we can consider using a more
sophisticated forecasting method, such as exponential smoothing or ARIMA. Additionally, we
can try to gather more data on factors that affect smartwatch demand, such as new product
launches, promotions, and competitor activity, to improve the accuracy of the forecast.
In addition to these product-specific recommendations, here are some general recommendations to
improve the supply chain planning process:
1. Improve data collection and analysis: To improve the accuracy of the forecast, it is important to
gather as much data as possible on factors that affect demand, such as consumer behavior,
economic conditions, and competitor activity. Additionally, it is important to analyze the data using
sophisticated statistical methods to identify patterns and trends in the data.
2. Collaborate with suppliers and distributors: To reduce lead times and improve overall supply chain
performance, it is important to collaborate closely with suppliers and distributors. This can involve
sharing data on demand forecasts, production schedules, and inventory levels to ensure that the
right products are available at the right time.
3. Implement a continuous improvement process: To ensure that the supply chain planning process
is always improving, it is important to implement a continuous improvement process. This can
involve regularly reviewing and analyzing data on demand forecasts, inventory levels, and production
schedules to identify areas for improvement and implement corrective actions.
Or
1. Smartphones: Recommendation - Consider using a more sophisticated forecasting method, such
as exponential smoothing or ARIMA. Justification - The MAPE for smartphones is 16.67%, which
indicates that the forecasting accuracy is not very high. Using a more sophisticated forecasting
method can help improve the accuracy of the forecast by accounting for trends and seasonality in
the data. Additionally, gathering more data on factors that affect smartphone demand, such as new
product launches, promotions, and competitor activity, can also help improve the accuracy of the
forecast.
2. Laptops: Recommendation - Adjust the forecasting method to account for seasonality or trends in
the data. Justification - The Bias for laptops is -7.24%, which indicates that the forecast tends to be
slightly lower than the actual demand. Adjusting the forecasting method to account for seasonality
or trends in the data can help address this bias and improve the accuracy of the forecast.
Additionally, gathering more data on factors that affect laptop demand, such as changes in
technology, consumer preferences, and economic conditions, can also help improve the accuracy of
the forecast.
3. Smartwatches: Recommendation - Consider using a more sophisticated forecasting method, such
as exponential smoothing or ARIMA. Justification - The MAPE for smartwatches is 38.22%, which
indicates that the forecasting accuracy is quite low. Using a more sophisticated forecasting method
can help improve the accuracy of the forecast by accounting for trends and seasonality in the data.
Additionally, gathering more data on factors that affect smartwatch demand, such as new product
launches, promotions, and competitor activity, can also help improve the accuracy of the forecast.
Or
1. Smartphones: Recommendation 1 - Consider using a more sophisticated forecasting method, such
as exponential smoothing or ARIMA.
Justification 1 - The MAPE for smartphones is 16.67%, which indicates that the forecasting accuracy is
not very high. Using a more sophisticated forecasting method can help improve the accuracy of the
forecast by accounting for trends and seasonality in the data. Recommendation
2 - Gather more data on factors that affect smartphone demand, such as new product launches,
promotions, and competitor activity.
Justification 2 - Gathering more data on factors that affect smartphone demand can help improve
the accuracy of the forecast by providing more information to the forecasting model.
Recommendation
3 - Consider using a combination of forecasting methods, such as a combination of moving average
and exponential smoothing.
Justification 3 - Using a combination of forecasting methods can help improve the accuracy of the
forecast by taking advantage of the strengths of each method.
2. Laptops:
Recommendation 1 - Adjust the forecasting method to account for seasonality or trends in the data.
Justification 1 - The Bias for laptops is -7.24%, which indicates that the forecast tends to be slightly
lower than the actual demand. Adjusting the forecasting method to account for seasonality or trends
in the data can help address this bias and improve the accuracy of the forecast.
Recommendation 2 - Gather more data on factors that affect laptop demand, such as changes in
technology, consumer preferences, and economic conditions.
Justification 2 - Gathering more data on factors that affect laptop demand can help improve the
accuracy of the forecast by providing more information to the forecasting model.
Recommendation 3 - Consider using a combination of forecasting methods, such as a combination
of moving average and exponential smoothing.
Justification 3 - Using a combination of forecasting methods can help improve the accuracy of the
forecast by taking advantage of the strengths of each method.
Recommendation 3 - Consider using a combination of forecasting methods, such as a combination
of moving average and exponential smoothing.
Justification 3 - Using a combination of forecasting methods can help improve the accuracy of the
forecast by taking advantage of the strengths of each method.
3. Smartwatches: Recommendation 1 - Consider using a more sophisticated forecasting method,
such as exponential smoothing or ARIMA. Justification 1 - The MAPE for smartwatches is 38.22%,
which indicates that the forecasting accuracy is quite low. Using a more sophisticated forecasting
method can help improve the accuracy of the forecast by accounting for trends and seasonality in
the data.
Recommendation 2 - Gather more data on factors that affect smartwatch demand, such as new
product launches, promotions, and competitor activity. Justification 2 - Gathering more data on
factors that affect smartwatch demand can help improve the accuracy of the forecast by providing
more information to the forecasting model. For example, tracking the launch of new smartwatch
models by competitors can help predict changes in demand for your own smartwatches.
Recommendation 3 - Consider using a combination of forecasting methods, such as a combination
of moving average and exponential smoothing. Justification 3 - Using a combination of forecasting
methods can help improve the accuracy of the forecast by taking advantage of the strengths of each
method. For example, using moving average can help capture the trend in the data, while
exponential smoothing can help capture the seasonality.
