Hello everyone. In this video,
I'd like to give an overview of
the course on Demand Analytics.
Demand analytics is the art and science of
applying data analytics to
demand planning and forecasting,
which is critical to a firm's goal
of increasing revenue and reducing cost.
Demand planning and forecasting is
a cross-functional activity and involves supply chain,
marketing, and finance people.
It is closely linked to
the firm's strategic and financial objectives,
and often serves as the starting point of the sales and
operations planning and the integrated business planning.
Demand analytics is one of the most sought-after
skills in supply chain management and marketing.
As reported by supply chain insights,
jobs like demand and supply planning and data
scientist top the list of positions in greatest demand.
Demand planning and forecasting is especially important,
but challenging in the new product introduction,
where demand can be hardly predictable,
with forecast errors as high as 200 percent.
Yet, accurate forecast is needed for sourcing,
production, distribution, and advertising planning.
Analytics can help to make accurate forecast for
the demand and assess
the impact of various factors on the demand.
Upon completion of this course,
you will be able to;
first, build data-driven analytics models
to predict the demand.
Second, assess the impact
of various factors such as time,
seasonality, price, and other environmental
factors on the demand.
Finally, understand how
demand planning and forecasting is
linked to a company's strategic goals
and the overall business planning.
To achieve these objectives,
this course covers the following topics in four weeks.
In week 1, you will learn
the general principles of
demand planning and forecasting in practice.
In week 2, we'll cover
data visualization and predicting the trend.
In week 3, you will learn
model validation and improvement techniques,
and how to predict the impact of
price and other environmental factors.
In week 4, you will learn to model
categorical variables such as seasonality,
and how to create and test forecasts.
The course is designed so that over these weeks,
you will learn more and more techniques to
improve the model and make it sharper,
better, and more accurate.
At the end, you should be able to build a model to make
accurate forecasts and quantify
the impact of various factors on demand.
Time commitment is about 2.5 hours a week,
and we will use Excel and
the Data Analysis Add-In as the Software.
More specifically, in week 2,
you'll learn how to pre-process the data and visualize
the data by line and scatter charts, such as this one.
Then you'll build a simple linear regression model
to predict the trend.
You will understand the parameter estimation,
model building, and interpretation, and how to
test the significance of the model.
In week 3, you will learn
model validation and improvement techniques
to detect hidden patterns in errors of the model,
such as this one, which is called Residuals.
Based on the patterns identified,
you will build a multiple regression model
of both time and price,
to significantly improve the fit of the model.
Although the model is much improved,
it's errors still have
a periodic pattern as shown by this chart.
So in week 4, we'll introduce seasonality
by modeling and formatting
seasons as categorical variables.
The resulting model with time, price,
and seasonality, accurately captured
the variations in the demand data.
Using the model, you can
make a forecast for the whole year of
2013 as shown by the red dots on this chart.
Testing the forecasts on the first three months of 2013,
for which the actual sales data is available,
you will find that the forecast error
measured by the mean absolute percentage deviation,
MAPD, is 8.9 percent,
which is very accurate for
a relatively new product in cookware.
This introductory course on Demand Analytics,
is designed for two general audience.
First, those of you who are exploring
a career in Supply Chain Management and Marketing,
especially in demand planning,
sales and operations planning,
and integrated business planning.
Second, those of you who are
fascinated by the potential of
Predictive Analytics and like to learn
its applications in
supply chain management and marketing.
This course is designed for
beginners with no prior experiences.
However, it will be helpful if you have
some knowledge or experiences in the following areas,
such as general business acumen,
Supply Chain Analytics Essentials,
and Supply Chain Planning.
