Bibitor LLC *
Inventory Analysis Case
Study
Phase 2
*Note1: Bibitor, LLC is a fictitious company based on data created by the HUB of
Analytics Education @ www.hubae.org.
Introduction and
Case Overview
PwC 2
Objectives
• This case study is designed to enhance your understanding and knowledge of
data analytics, including leveraging analytics to make business decisions.
• By completing this case, you will obtain the following:
- An understanding of what data analytics and visualization are
- The ability to utilize analytic tools
- Comfort in leveraging data in making decisions
- Skills to select data components for analysis and illustrate results to
highlight data tracking and identify conditions for exception analysis or
business advisory strategy
PwC 3
Case Format
• You should plan to work in teams of 4 to 5 students to complete the case. This will give
you the opportunity to share the learning, knowledge, and experience.
• This case will be delivered in three phases. Each phase includes an instructional
component to learn how to manage the data within the tool and case study exercises
utilizing the skills.
- Phase 1: Introduction to Data Preparation
◦ Load data exported from company into SQLite to put it into a format for analysis and
create tables which can be used by other analysis tools.
- Phase 2: Introduction to Data Discovery and Visualization
◦ Using tables created in Phase 1, load data into Tableau to perform analysis then
create and interpret visualizations.
- Phase 3: Introduction to Statistical Analysis
◦ Analyze one set of independent/dependent variables from each of the two case
studies and create a tableau worksheet that uses linear regression to provide some
insight into the data.
PwC 4
Case Materials
• All the materials required to complete the case can be found at ((LOCATION)).
PwC 5
Case Background
• Bibitor, LLC* is a liquor store chain in the fictional state of Lincoln. It is a major
retailer with approximately 80 locations and total sales in excess of $450
million.
• Bibitor has asked the team to complete due diligence on their wine and spirits
business looking at data for their beginning and ending inventory, purchases
and sales for a
12 month period.
*Bibitor, LLC is a fictitious company based on data created by the HUB of Analytics Education @ www.hubae.org. The HUB of
Analytics Education materials are owned by Northeastern University. PricewaterhouseCoopers LLP is not responsible for any
errors or omissions in, or for the results obtained from the use of, the HUB of Analytics Education. The HUB of Analytics
Education materials are provided "as is", with no guarantee of completeness, accuracy, timeliness or of the results obtained
from the use of this information, and without warranty of any kind. In no event will PricewaterhouseCoopers LLP, or its partners,
principals, employees, or agents, be liable to you or anyone else for any decision made or action taken in reliance on the
information in the HUB of Analytics Education materials or for any consequential, special or similar damages, even if advised of
the possibility of such damages.
PwC 6
Getting Started – Phase 2 Pre-work
1. Download/Install Tableau and Tableau Reader
2. Review background information on Tableau and other Data
Discovery applications. Follow guidance in 3. below
3. In the Tableau application, open Help, then ... 1. Get Started - understanding
Dimensions & Measures 2. Analyze 3. Present 4. Watch training videos
4. Use Google as a resource and search for Tableau Cheat Sheets for handy
guidance
PwC 7
Phase 2 –
Introduction to
Data Discovery
and Visualization
PwC 8
Data Discovery and Visualization
• Now that the data is compiled into usable tables, the steps in this phase
will show you how to load these tables into a tool to create and
interpret visualizations.
PwC 9
Review of Pre-Work and Case Results
PwC 10
Data Pre-Processing
• Using the database from Phase 1, create a table with the following:
- For each unique record-- ‘Brand’ (e.g. Product SKU) by ‘Store’-- in table EndInvDec,
recalculate the ‘PurchasePrice’ assuming the costing method is moving-average cost.
- Assume that each transaction per table SalesDec reduces inventory on-hand
quantities and that each transaction in table increases inventory on-hand quantities.
- Your resulting table should include [A] a new column with the re-calculated purchase
price using moving-average cost based on actual transactions; the resulting table
should also include [B] a new field that calculates the difference between the
‘PurchasePrice’ per table EndInvDec and the recalculated column (column [A]).
- Assume that quantities & prices in table BegInvDec are as-of 12/1/2016.
- Hint: Sales & Purchases transaction tables include data from prior to December.
- Review Appendix3_Section2_1_SQLPrepTable.txt
PwC 11
Data Pre-Processing
PwC 12
Link Visualization File to Source Data
a. Export prep table from Exercise 1a above to excel, and save to local folder.
- Note: DO NOT USE SQL DB Export feature; simply copy/paste table
from query results into an excel file, and include your column names in
the top header row.
b. Open Tableau application. Click Connect to Text file navigate to prep table
saved locally in previous step.
PwC 13
Link Visualization File to Source Data
PwC 14
General Navigation
PwC 15
Become Familiar with Basic Functions
• Drag the ‘Store’ field to the
Columns breadcrumb, and
drag the ‘OnHand’ field to the
Rows breadcrumb.
PwC 16
Become Familiar with Basic Functions
• Tableau treated the
imported source data in
field ‘Store’ as an integer,
rather than a string.
• We will not be calculating
values within the ‘Store’
field. As such, click the
drop down on ‘Store’
within the Columns
breadcrumb, and
select ‘Dimension’.
PwC 17
Become Familiar with Basic Functions
• Try selecting each graph type to
see which type may fit the data
set best.
• For now, let’s use the ‘Packed
Bubbles’ graph type.
• How should we interpret this?
PwC 18
Become Familiar with Basic Functions
• Using Filters: Drag the
‘Store’ field to the
‘Filter’ card.
PwC 19
Become Familiar with Basic Functions
• Drill into the filter details and
select only the range of
stores from 1 to 5
PwC 20
Become Familiar with Basic Functions
• Using Marks: marks are features uses
to identify specific datapoints.
• In this example, marks are
represented by the size, color, and
number of circles.
• Drag the “Price” field to the ‘Color’
mark—the field should default to a
calculated value (sum(Price)).
• You can change the color by clicking
on the generic ‘Color’ box inside the
Marks card.
• Next, drag field ‘Difference’ to the
‘Size’ mark.
• How would you interpret this graph?
PwC 21
Become Familiar with Basic Functions
• Using Analytics: some standard
methods of analysis are built
into Tableau.
• Click and drag a ‘Cluster’ model into
the sheet.
• Drill into the cluster variables and
ensure the variables are ordered by
sum(Difference) , then by sum(Price).
• Values for respective individual
clusters will be re-colored to a
consistent color and arranged
together, based upon similar values for
the defined variable characteristics.
PwC 22
Become Familiar with Basic Functions
• Increase the number of
clusters (automatically
defaulted to 2) to 3 as an
example of how to
manipulate the clustering.
PwC 23
Become Familiar with Basic Functions
• Dashboards vs. Worksheets
– a dashboard can contain
multiple worksheets.
PwC 24
Case Studies
PwC 26
Instructions
• Create a basic dashboard using the SQL prep tables created in Phase 1
PwC 26
Case 1
• Executive leadership is looking for dashboard reporting that would be useful to
help identify and monitor vendor activity in order to focus efforts strategically
on key supplier relationships.
• Create (1) an aggregate table that includes all critical vendor billings and their
associated purchasing activity
• Create (2) separate tables to store key information, such as ‘top 10 vendors’ by
quantity purchased. Name all tables you create ‘c1_Prep_[table_name]’
PwC 27
Case 2
• You notice that there is inventory that is purchased but sits on the shelf for a
long period of time rather than being sold.
• As such, there is an opportunity to add value to the procurement process by
identifying trends in the timing inventory purchased versus its
corresponding sale.
• Consider key data points around purchase prices, seasonality, and vendor
information in creating an aggregate table for future visual &
statistical analysis.
• Name the table you create ‘c2_Prep_[table_name]’.
PwC 28
Next Steps
PwC 30
Phase 3 Pre-work
• Review background information on Linear Regression (you do not need to
download “R” when you read the article):
- https://www.hackerearth.com/practice/machine-learning/machine-learning-al
gorithms/beginners-guide-regression-analysis-plot-interpretations/tutorial/
• Review Basic Concepts
- Single-variable Linear Regression (Definition): a statistical method that
allows for the evaluation of the relationship between two continuous
variables, one predictor (input) variable –‘x’, and one response (output)
variable—‘y’.
PwC 30
Phase 3 Pre-work
• The key function is the same as the standard form of an algebraic
linear equation:
f(x) = b + a(x) = y
• Where x is the independent (predictor) variable, y, or f(x), is the dependent
(response) variable, b is a constant, and a is the slope of the line.
• Purpose: we are applying this function in order to predict the outcome of a
future event/input. The function approximates the line of ‘best fit’ given the
relationship of two variables based upon pre-existing/historical data. As such,
in reference to the following example graph, we would expect that new events
with input value = 4, would result in an output value of ~12.5.
PwC 31
Phase 3 Pre-work
Source:
https://www.google.com/search?q=basic+linear+equation+graph&rlz=1C1GGRV_enUS753U
S753&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjf29uDvJPVAhUKfiYKHQBvANIQ_AUI
CigB&biw=1533&bih=761#tbm=isch&q=basic+fitting+line+graph&imgrc=PG9RR9w1ofQR6M
PwC 32
Phase 3 Pre-work
• Deriving the Slope of the line of
best fit:
- Numerator: for each known data
point, calculate the sum of the
variances – where point is
each known event, and is the
calculated mean of the independent
variable values, and the is the
calculated mean of the dependent
variable values.
- Denominator: for each known
independent variable value, calculate
the sum of squares of variance –
where is each known
independent variable value and
is the calculated mean.
PwC 33
Phase 3 Pre-work
• Deriving the Constant of the
line of best fit:
- Where a is slope, is the
calculated mean of the
dependent variable values,
and is the calculated mean
of the independent variable
values.
PwC 34
Thank You.
