PROJECT SUBMISSION
Task 1
● Method of forecasting:
For Short Term demand forecasting, particularly for next six months, a
suitable method for Tech Hub Electronics to consider would be Time Series
Forecasting, specially using a method like Exponential Smoothing or ARIMA
(Autoregressive Integrated Moving Average)
● Justification for the choice of method:
Time Series Data: The Provided data represents historical sales data for the past
year, making it a time series dataset. Time series forecasting methods are well
-suited for predicting future values based on past observations.
Seasonality & Trends: By analyzing the data, we can observe that there might be
some seasonality and trends in the demand for each product category. For
Instance, smartphones and laptops show an increasing trend, and there might be
monthly or quarterly seasonality. Time Series methods can capture and
incorporate such patterns into the forecasts.
Short-Term Focus:
TechHub Electronics is interested in forecasting demand for the next six months.
Time series methods are effective for short-term forecasting, as they take into
account recent historical data and provide forecasts for a specific time horizon.
Assumption:
One key assumption in using time series forecasting is that historical patterns and
relationships in the data will continue into the future. This assumption may not hold
if there are significant changes in market conditions, consumer preferences, or
external factors.
To address this, the company should regularly assess the model’s accuracy and be
prepared to update the forecast if substantial changes occur.
Additionally, it’s crucial to monitor factors like product launches, marketing
campaigns, and economic conditions that can influence demand and incorporate
them into the forecasting process when necessary.
Task 2
● The formula used for forecasting demand:
Ft + 1 = α * Dt + (1 – α) * Ft
● To forecast the demand for each product for the next six months using a
simple exponential smoothing method with α = 0.3, following steps was
used:
● Calculate the forecast for January (Month 13) as the average of previous
year’s January demand:
Forecast for January = (1 – α) * Demand for January + α * Forecast for
December Forecast for January = (1- 0.3) *1500 +0.3 *2300
Forecast for January = 0.7 * 1500 + 0.3
*2300 Forecast for January = 1050 +
690
Forecast for January = 1740
This process was continued for the remaining months (March to June) using
the previous forecasted value and actual demand for each month. Repeat this
process for each product category (Smartphones, Laptops, and
Smartwatches).
● The demand forecast for the next six months is shown in the table below
Month Smartphones (Forecast) Laptops (Forecast) Smartwatches (Forecast)
Jan 1740 941 361
Feb 1642 872 345
Mar 1644 876 346
Apr 1592 838 332
May 1607 843 335
Jun 1598 845 337
Task 3
● The formula used for calculating forecast
errors: Forecast Formula = [ Actual Demand –
Forecast Demand]
● The forecast errors for the six months are shown in the table below
Month Smartphones Laptops Smartwatches
Jan 140 186 102
Feb 7 7 53
Mar 75 1 96
Apr 158 85 118
May 313 157 135
Jun 442 255 141
Task 4
● The formula for calculating MAPE:
Mean Absolute Percentage Error is the measure of accuracy of a forecasting
method. It measures the average absolute percentage difference between the
forecasted values and the actual values.
MAPE = (1/n) *∑ (| (Actual Demand – Forecast Demand)/ Actual Demand|) *100
Where,
n is the number of months (is this case ,6.)
Actual Demand is the actual demand for a given month.
Forecast Demand is the forecasted demand for the
same month.
● MAPE for all product categories for the six months:
Month Smartphones Laptops Smartwatches
Jan 8.75% 24.83% 39.38%
Feb 0.43% 0.80% 18.15%
Mar 4.24% 0.11% 38.40%
Apr 8.32% 9.24% 35.76%
May 16.14% 15.57% 43.55%
Jun 27.64% 30.24% 41.97%
● The formula for calculating Bias:
Bias measures the overall tendency of forecasts to be consistently higher than lower
than actual values. It is the sum of forecast errors divided by the number of months.
Bias = (1/n) *∑ [Actual Demand – Forecast Demand]
Where,
n is the number of months (is this case ,6.)
Actual Demand is the actual demand for a given month.
Forecast Demand is the forecasted demand for the
same month.
● Bias for all product categories for the six months:
Month Smartphones Laptops Smartwatches
Jan -140 -186 -102
Feb -7 -7 -53
Mar -75 1 -96
Apr 158 85 -118
May 313 157 -135
Jun 442 255 -141
These Bias values represent the tendency of the forecasts to be consistently higher
or lower than actual demand for each product category over the six-month period.
Positive values Indicate an overestimation of demand, while negative values
indicate an underestimation of demand. These metrics provide insights into the
direction and magnitude of forecast errors for each month and product category.
Task 5
● Recommendations:
Analyzing the forecast errors and key performance metrics (MAPE and Bias)
provides insights into the the supply chain’s performance in forecasting and
meeting actual demand. Based on this analysis, here are recommendations for
improving the supply chain planning process for each product category:
SMARTPHONES
MAPE Analysis
o The MAPE for each smartphone over the past 6 months indicates that
the forecasting accuracy has room for improvement, with significant
errors in March and June.
Bias Analysis
o The Bias Value Is generally negative, indicating a tendency to
underestimate demand for smartphones.
Recommendation for Smartphones:
Enhance Demand Sensing: Implement a more responsive demand sensing
system that can quickly capture changes in consumer preferences and
market trends. This can include real- time data analytics and social media
monitoring to detect emerging patterns.
Collaborate with suppliers: Collaborate closely with suppliers to reduce lead
times for critical components used in smartphone production. Establishing a
more agile and responsive supply chain can help meet sudden increase in
Demand.
Advanced Forecasting Techniques: Explore advanced forecasting techniques
such as machine learning algorithms, incorporating more variables like
promotional activities and competitor data to improve forecast accuracy.
LAPTOPS
MAPE Analysis
o The MAPE for laptops indicates relatively good forecasting accuracy
for most months.
Bias Analysis
o The Bias Value is close to Zero, suggesting that forecasts are relatively
unbiased, but there is room for improvement.
Recommendation for Laptops:
Implement Demand Collaboration: Collaborate closely with retailers and
distributors to share demand and Inventory Data. This can help optimize the
supply chain and reduce the risk of overstocking or understocking laptops.
Inventory Buffer: Maintain a small Inventory buffer for laptop components
that are prone to supply chain disruptions. This can help reduce lead times
and ensure a more consistent supply.
Continuous Improvement: Invest in continuous improvement initiatives
within the supply chain, focusing on demand forecasting techniques and
supply chain efficiency. Regularly, review and refine forecasting models to
adapt to changing market conditions.
SMARTWATCHES
MAPE Analysis
o The MAPE for smartwatches indicates significant forecast errors,
especially in the later months (May and June)
Bias Analysis
o The Bias Value is negative, indicating a tendency to underestimate
demand for smartwatches.
Recommendation for Smartwatches:
Market Segmentation: Implement market segmentation strategies to better
understand different customer segments and their preferences for
smartwatches. This can lead to more accurate demand forecasts.
Collaboration with Retailers: Collaborate closely with retailers and e-
commerce partners to better align production with actual demand.
Implement vendor-managed Inventory (VMI) systems to ensure more
accurate replenishment.
Product Diversification: Diversify the smartwatch product range to offer
options catering to various consumer preferences. This can help reduce the
risk of forecasting errors by providing a wider product portfolio.
In Summary, TechHub Electronics can improve its supply chain planning process by
enhancing demand sensing, collaborating with suppliers and retailers,
implementing advanced forecasting techniques, maintaining inventory buffers, and
focusing on continuous improvement.
Differentiated strategies are needed for each product category to address
specific challenges and leverage opportunities for better supply chain
performance in a rapidly changing market.
********************************************************************
