SLIDE 1
Forecasting customer demand for packages in SIG Indonesia market for - - PowerPoint PPT Presentation
Forecasting customer demand for packages in SIG Indonesia market for - - PowerPoint PPT Presentation
Forecasting customer demand for packages in SIG Indonesia market for better customer sales promotion Team 2 Shang-Chi Tu, Beverly Lin, Ken Wu, Jimmy Wang Background Founded 1853, SIG is headquartered in Switzerland and offers services from a
SLIDE 2
SLIDE 3
Business goal
Through forecasting future monthly sales volume of packages, SIG could offer new sales promotion for different customers in Indonesia.
Forecasting goal
- Forecast customer demand for packages in the next 12 months on a customer and product type level
- SIG Dept. of Sales
- SIG’s Indonesia customers
- Set up suitable marketing strategy among
customers and seasons
- Salespeople workload rearrangement
- Over-forecast: resource-wasted
- Under-forecast: work overloaded
- Many other factors should be considered.
Client & Stakeholders Humanistic implication Challenge Opportunity
SLIDE 4
Data description
Source: from SIG Measurable: number of sale volume on a customer and product type level
Time period: Jan. 2009- Dec. 2018 Amount :120 records in each series 45 series in total Frequency of collecting data: Monthly Unit: Thousands of package sales volume
External data:
2018 Monthly Forecast provided by SIG Aggregation
SLIDE 5
Data pre-processing
Type A Type B Type C
SLIDE 6
Methods
We apply these methods to the repeat customer and chose the one with best RMSE(Regression).
Forecast next 12 months = 0 Naive Inactive Customer
(2018 ordering sum = 0)
New customer
(total ordering period < 24 months)
Repeat Customer
(total ordering period > 24 months)
Training period: 2009/1~2017/12 Validation period: 2018/1~2018/12
Partition
Naive Exponential smoothing (ets) Snaive Neural Network Regression
(with trend and seasonality)
Auto Arima
Ensemble using External Data 954104 PC010125A
SLIDE 7
Evaluation
Benchmark: Snaive Forecast by customer segmentation Ensemble: Our forecast+External Data
Inactive customer New customer Repeat Customer Total Inactive customer New customer Repeat Customer Total
RMSE 2699.6 2334.8 4030.4 2220.6 1066.1 2230.2 1152.7
SLIDE 8
Results
New customer
Validation Training
New customer
Actual Ensemble Naive
Validation Training
SLIDE 9
Repeat customer Repeat customer
Results
Lowest RMSE = 674.076 Highest RMSE = 9071.003
Validation Training Validation Training
Actual Ensemble Regression
SLIDE 10