IBM Cognos Analytics 11.1.

4 : Getting

Site feedback
Started with Exploration
Explore the key capabilities of IBM Cognos Analytics Data Exploration

Before you start

This tutorial is meant to be used with a VM designed to showcase IBM Cognos Analytics. If you do
not already have a running instance, please sign up for one.

Please log in (https://www.ibm.com/account/reg/us-en/signup?formid=urx-

36056&target=https://www.ibm.com/cloud/garage/dte/tutorial/ibm-cognos-analytics-1113-
getting-started-exploration-0?cm_mmc=Email_Newsletter-_-Developer_Ed%2BTech-_-
WW_WW-_-SkillsNetwork-Courses-IBMDeveloperSkillsNetwork-DV0130EN-SkillsNetwork-
20531073&cm_mmca1=000026UJ&cm_mmca2=10006555&cm_mmca3=M12345678&cvosrc=ema
_-Developer_Ed%2BTech-_-WW_WW-_-SkillsNetwork-Courses-IBMDeveloperSkillsNetwork-
DV0130EN-SkillsNetwork-
20531073&cm_mmca1=000026UJ&cm_mmca2=10006555&cm_mmca3=M12345678&cvosrc=ema
to be able to reserve an instance.

Privacy - Terms
Tutorial
Explore is a flexible workspace intended for ad-hoc data
discovery and exploration. You can explore your data or
investigate visualization from a dashboard or story. As you
work, contextual recommendations and advanced analytics
insights will be surfaced in a subtle way, so as not to
intrude. Take advantage of advanced analytics normally

Site feedback
available only in data science tools. You won’t be bothered
with positioning widgets or fancy formatting. Just work
away, and when you want to present your insights, simply
copy your visualizations to a story or dashboard.

This workshop introduces you to some of the newest

capabilities included in IBM Cognos Analytics, Data
Exploration, which utilizes augmented intelligence to
accelerate the time to value in finding insights hidden deep
within your data

In this tutorial, you will explore the following key

capabilities:

• Upload Data sources to Explore

• Discover relationships in your data
• Surface Insights to begin your analysis
• Identify patterns and key drivers

Get Started With Your Cognos Analytics Tutorial

This tutorial is meant to be used with a virtual machine
designed to showcase IBM Cognos Analytics. If you do not
already have a running instance, follow the steps described
below.

Reserve an Instance to Participate in the Cognos

Analytics Deep Dive

1 The first step is to reserve your instance to conduct the

Cognos Analytics deep dive. Click the log in link on this
page and then login with your IBMid:
  Privacy - Terms
  
If you do not have an IBMid you can easily create it from

Site feedback
the login page by clicking Create an account.
 

2 After you log in with your IBMid, complete the

registration form and click Reserve instance.

Privacy - Terms
 Site feedback
3 Check your mailbox and open the email from IBM
Digital Sales Support with the subject “Your IBM
demonstration is now ready”. The email includes the
link to your environment and a password to access this
environment.

Privacy - Terms
4 Click the link provided for full desktop access. Enter the

Site feedback
provided desktop password and click Submit to access
the VM.

Start the Virtual Machine VM for your Tutorial

1 Check the status of your image that is indicated in the

top-left corner. It could be Powered off, Suspended, or
Running.

2 If your image is already running, skip this step. If it’s

Suspended or Powered off, click the Play button to start
the VM and wait a few moments until the VM has booted
(you can see the spinning gear).

Privacy - Terms
 Site feedback
3 Once the VM is ready the computer monitor thumbnail
will have a light green background. Click the monitor to
start the VM.

4 The VM will open in a browser tab. You can use the

controls at the top of the screen to adjust the resolution.
Click Fit to Window button to fit the screen to the size of
the browser. Also, it is recommended to set the window
size of the browser to full screen using the Change
Resolution setting. Once done, you may close the panel.

5 As your environment is set up, you will see a welcome

screen. It may take up to 5 minutes to activate it. Your
environment is ready to use once you are able to close
the browser tab of the welcome page using the X on the
Privacy - Terms
browser tab.
 Site feedback
 

Start Cognos Analytics for your Tutorial

1 Once the Welcome browser tab is closed, the Cognos

login screen may be open. If not, from the desktop,
open Chrome. From the Bookmarks Toolbar, open the
Cognos Analytics bookmark to launch IBM Cognos
Analytics and bring up the login page.

2 Enter the sign-in credentials as follows:

• User ID: cogadmin

• Password: IBMDem0s!

Cognos Analytics User Interface

Privacy - Terms
The goal of the User Interface (UI) is to provide Users
with a streamlined way to get started using Cognos
Analytics and view content and activities pertinent to
them. 

1 The User Experience brings you directly into the

completely redesigned Cognos Analytics User Interface
(UI).  All Cognos Analytics Users begin their navigation

Site feedback
here. 

TECH TIP:  Not all Welcome Screen Getting Started tiles

may be available in your workshop instance.  The Welcome
screen getting started tiles presented are based on which
have been configured to run in the environment. 
 

2 Click on the Welcome Screen toggle button on the

upper left of the canvas to collapse the Welcome
Screen’s Getting Started content.

3 The canvas now shows the Recently used files in the

Recent section, if any, along with the File drop zone
where Users can easily upload their data files.

Privacy - Terms
 Site feedback
4 Once you begin working with content, the canvas will
update with your recently used items.  In your Cognos
Analytics instance, you may see recent content on the
canvas.

TECH TIP:  Detailed information regarding the UI

capabilities are shown in Appendix A for your reference. 

Business Use Case for this Workshop

You are a new Marketing analyst at a large Communications
company. Your company has just acquired a provider
named Telco that offers phone, DSL internet and Fiber
Optic Internet services. Telco’s customers subscribe to
these services and may have either phone or internet or
both services. Due to the recent acquisition, there is
concern that some Telco customers may cancel their
services.
In an increasingly competitive environment for these
services, you’ve been hired to find ways to reduce customer
churn. To begin, you need to understand your customer
base and identify which customers are most likely to churn.
This will enable Sales and Marketing to create promotional
programs targeted to those customers with a high
propensity to churn.
You’ve just received an email from the Marketing Director
asking for your assistance in analyzing data from the newly
acquired Telco.
Privacy - Terms
 Site feedback
Using the new Data Exploration capabilities in Cognos
Analytics, you can jump start your understanding of the
data and quickly uncover patterns and insights.

Uploading external data files

The ability for Business Users to leverage their
personal/external data for discovery dramatically
broadens the landscape of Users who can make new data
available for analysis. Users may upload an external data
file and immediately begin self-service data discovery, ad
hoc analysis and building dashboards.

For the first exercise, you will be using the file

“Telco_customer_churn.xlsx.” This file is located in a
folder on the desktop of your Cognos Analytics instance:
 Desktop>CA Workshop Data

1 To upload a file, you may either drag and drop this file
into the drop zone, or you may click Browse in the drop
zone to navigate to where the file is saved. For this
workshop, we will use the Drag and Drop option.
2 Open File Explorer and navigate to Desktop>CA
Workshop Data.
3 Drag the Telco_customer_churn.xlsx file over to the
Drop zone button on the canvas until the blue Quick
Launch menu appears at the bottom of the screen.
Privacy - Terms
 Site feedback
4 The Quick Launch panel will allow you to select your
intent on how you wish to interact with the data you are
uploading. Drop the file over the Exploration option.

Using this method, Cognos Analytics will automatically

upload the file to your My Content folder and launch the
Data Exploration capabilities for you to begin working with
your data.

5 As the file uploads, notice that under the Switcher

Menu, a series of status bars will be visible as the
upload process reads and analyzes the data being
brought in.

6 Once it completes, the status bar will update to show

the successful completion before closing.

Since you selected to upload the data and launch an

Exploration, the Exploration User Interface will immediately
open after the upload is complete. You’re now ready to Privacy - Terms
start to explore your data. Cognos Analytics analyzes the
data and identifies interesting items, then presents a
relationship diagram as the initial view.

Starting Cards
Every Exploration includes a Starting Card to begin

Site feedback
analysis.  Users can enter any data column to use as a
starting point, or can select from a list Cognos Analytics
identified

1 Select Churn for your Starting Card.

2 The Exploration will update to a Relationship Diagram

for your field of interest.

Data Relationships
The relationship diagram plots fields based on a
statistical evaluation of related items. The relationship
diagram is not a picture of the data model. However, the
model could be an influencing factor in the analysis. The
diagram builds around an identified “Field of Interest”.
Related fields branch off the Field of Interest and are
organized into circles, which represent concepts. Lines
connect the Field of Interest to concepts and represent
relationships.

1 In this example, Churn is the Field of interest. Privacy - Terms

 Site feedback
Churn is a measure field and you will see it as blue circle in
the center of the diagram. The lines show which fields are
related based on a statistical analysis, with the thickness
indicating the strength of the relationship.

2 To change the Field of Interest, click on the “X” in the

interest field. Cognos Analytics will render a list of
additional recommend Fields of interest for analysis.
Select the ‘Tenure’ field.

3 The diagram re-draws with ‘Tenure’ as the Field of

interest.

Privacy - Terms
 Site feedback
The Relationship Diagram shows interesting fields in your
data and how they are related to other fields in your data.
By default, 10 unique fields are shown, but you can choose
to show more or less using the Relationship strength slider.
 

4 Click on the arrow next to the Save icon on the

dashboard toolbar. Then click “Save As”. Navigate to My
Content. Save as “[Your Lastname] Telco Exploration”.

Privacy - Terms
Suggested Starting Points
You can continue your analysis by choosing a
visualization recommendation from Cognos Analytics. 
You can also select items in the relationship diagram to
generate visualizations to use as your own starting points
for analysis.

Site feedback
1 You’ll see in the right panel the visualizations Cognos
Analytics identified as addition areas of interest which
could be used in your analysis.  These are all generated
based on the field of interest in the relationship
diagram. 

2 In the diagram, use control key to multi-select Tenure,

Contract and Partner. As you select fields, suggested
starting points using those fields will be rendered, to the
left of the relationship diagram.

Privacy - Terms
 

3 Click the Column chart suggested visualization for

Tenure by Contract colored by Partner. This creates a
new “card” with the visualization in focus.

Site feedback
The Details panel to the right of the chart display
information about the visualization. Details provide
additional context about the data displayed in the chart
that might not be immediately obvious

Visualization Library
1 From the Exploration toolbar above the visualization,
click the down arrow next to Column to open the
Visualization library.

2 The Visualization library opens with Recommended

visualizations optimized to display the current data in
focus. Scroll through the Visualization Library. As you
scroll down through the library, you will see that
visualizations are organized into usage type. From the
Comparison section, select Heatmap.

  Privacy - Terms
 Site feedback
 

3 You can quickly see that customers on a 2-year contract

through a Partner have the greatest tenure as indicated
by darkest blue area on the heat map.

 
Privacy - Terms
4 Currently, you are looking at the Total Sum for Tenure,
but for this analysis, you want to look at average tenure.
From the right panel, select the Data Slots tab. Under
the Heat field, click the ellipses (…) next to Tenure.

Site feedback
5 Select Summarize.

 
Privacy - Terms
6 Select Average.

Site feedback
 

7 The visualization will update to reflect the new

aggregation. Click outside the Summarize List to close
it.

8 Hover over the darkest block to see that the Average

value for Tenure for Partners with 2-year contracts is
58.977.

Privacy - Terms
 Site feedback
 

9 Click the Details tab to the right of the chart to display

additional information about the visualization. Here you
can see the average value of Tenure for Partner and
Contract is 32.371. This is much lower than the 2-year
contracts with partners, which was 58.977 seen above.

10 Save your Exploration.

Related Visualizations
When a visualization is in focus in an exploration, the
system recommends related insights beyond what you
specifically requested. Based on Cognos Analytics’ data Privacy - Terms
analysis, these related visualizations might be of interest
to you as they were found to be related to the data items
currently in focus.

Related visualizations replace one of the data elements

in the visualization or add another data element to create
a new visualization. Related visualizations use a
combination of learned user interactions, statistics and

Site feedback
“Level of interest” to suggest useful next steps in your
analysis.

1 From the Exploration toolbar above the visualization,

click the down arrow next to Related to open the
Related visualizations.

2 The visualizations presented contain one or more of the

fields from the current card combined with other fields
from your data source. Related visualizations are
intended as suggested next steps in your exploration
journey. Click on the Show more button to see
additional Related visualizations.

3 Several Related visualizations are suggested for you to

use for your analysis. You can save these to your
Exploration content by clicking the “+” sign below the Privacy - Terms
 visualization. Click the “+” under Tenure by Churn
colored by Contract.

Site feedback
NOTE: The Related visualizations in your workshop
instance may differ from those shown in the example. You
may select any other recommended Related visualization
for this exercise.

4 The visualization has now been added to the Cards list

on the left.  The green dot indicates it is a newly added
exploration.

Privacy - Terms
 Site feedback

5 Save the Exploration.

The Assistant

Privacy - Terms
The Cognos Analytics Assistant provides
recommendations to help answer questions and provide
the User with quick insights into their data. Many times,
Users will have specific questions they are looking to
answer but may not be as familiar with the dataset or
exact data items they need to uncover the insight or
answer they are seeking.

Site feedback
From within an Exploration or Dashboard, the User can
type in natural language text to discover patterns and
relationships in the data, as well as generate
visualizations that can be added to an existing or new
dashboard, exploration or story. Simply enter text related
to the Users analytical intentions and an AI
conversational agent will respond with visualizations and
other information to satisfy the request.

1 Click the Assistant icon

 from the Navigation panel.

2 The Assistant panel opens. At the bottom of the panel,

in the Ask a question field, type “Show Tenure” and
click Enter.

Privacy - Terms
 Site feedback
3 Cognos Analytics identified Tenure in the
Telco_Customer_Churn.xlsx file you uploaded,and
identified interesting fields in the data set that relate to
Tenure.

Privacy - Terms
 Site feedback
 

TECH TIP: If Tenure is in any other data sources, those

would be listed as additional matching sources. Your
workshop environment may have more, fewer, or no other
matching sources. This will not impact the exercise in this
workshop.
 

Next, you would like to get a better understanding of tenure

by contract and the total charges. You can ask the Assistant
this question using natural language.
Privacy - Terms
4 In the Ask a question field, type ‘What is Tenure by
Contract and Total Charges?’

5 The Assistant will bring back a series of visualizations

related to your question. Scroll through the results using
the arrows until you find a visualization similar to the

Site feedback
example below.

TECH TIP: You can increase or decrease the width of the

assistant panel as needed by using the pull bar on the right
edge of the panel.

Privacy - Terms
 Site feedback
6 Hover over the Information icon in the upper right
corner of the visualization. Cognos Analytics provides
you with additional information on the data items from
your inquiry.

Privacy - Terms
 Site feedback
 

7 Click the visualization and drag it over to the canvas.

Privacy - Terms
At any time during your analysis, you can work with your
data and visualizations to modify them to suit your analysis.
For this visualization, you see the Tenure as a sum, but you
would like change this to the average as you did on the
heatmap. You would also like to change the visualization
type.

8 Click the Data Slots tab to open it.

Site feedback
 

9 Click the ellipses (…) next to Tenure.

 

10 Select Summarize and then select Average.

 

11 Click outside the Summarize list to close it.

 

12 The visualization has now updated to reflect average

Tenure.

Privacy - Terms
TECH TIP: Summarization may also be changed at the data
level from the Data Source panel, rather than each time you
use it in a visualization.

13 You may change the visualization style for your

Explorations. Select the down arrow next to Scatter and
scroll down to the Comparison section, then select
Column.

Site feedback
14 To customize your visualization further, you can modify
the data slot entries to refine your visualization.  In the
Data slots panel, drag Tenure from the Length slot, and
drop it on top of Total Charges in the Color slot when it
appears highlighted in blue as shown below.  This will
swap the data items into the opposite slots. 

15 Your data slots and visualization should now appear as

follows:

Privacy - Terms
 Site feedback
16 Hover over any of the bars to see the values.
 

17 Save the Exploration.

With these changes, it is easy to see that Two-year
contracts have the highest tenure and total changes. Also
of interest, is that the one-year contracts generate less
Total Charges than those that are month to month. Since
Telco contracts may be sold direct to consumers or through
partners, you would like to explore this further to compare
performance by sales channel.

Comparing Data
You can easily compare data across two visualizations.
When using the Compare feature, a summary of key
information and differences between the two
visualizations is generated.

1 Click the Data sources icon from the Navigation panel

to open the Data Sources panel.

Privacy - Terms
 Site feedback
2 From the Data sources panel, you can now view all the
data columns in the Telco Customer Churn data file.
Click on Partner and drag it over to the Local filters area
of the chart. Drop once the data slot is highlighted in
blue to indicate the drop zone.

TECH TIP: Depending on your screen resolution, you may

need to scroll to the bottom of the Data slots panel to
expose the ‘Local Filters’ area.
 

Privacy - Terms
3 The Partner Filter Window will render. For the first
chart, you want to filter the chart to include only the
data on for direct sales contracts, those where no
partners were involved. Click ‘Yes’ and then ‘OK’ to
apply the filter.

Site feedback
4 The Chart will update with the filter applied. Notice the
Local Filters will indicate the Filter is applied showing
that Partner Includes Yes. This will show the direct
sales channel, where Partners were involved.

Privacy - Terms
 Site feedback
 

5 Click on the Compare button from the Exploration

toolbar.

6 Cognos Analytics provides you with options on how to

compare your data.  Because you added a filter to the
current widget, it provides an option to Compare by
inverting the filter.  This option will provide you with two
widgets, one where Partner Includes Yes, and the other
where Partner Excludes Yes.  Click on the Invert filter
on Partner card.

Privacy - Terms
 Site feedback
 

7 Compare will take the two visualizations and place them

side by side, on a new Compare card. For more real
estate for the Comparison, you may collapse the panel
for Details and Data Slots panel. Click the ‘Hide Details
and Slots’ icon at the top of the comparison.

Privacy - Terms
 

8 The order of the charts is set by the filter placed on the

original widget.  For this workshop, you set the filter for
Partners to Yes, so Cognos Analytics created the
Compare card such that Chart A includes Partners

Site feedback
(Includes=Yes), while Chart B excludes Partners
(Excludes=Yes).

9 There is a summary table below the comparison values

for each chart and the combined values across charts.
These charts include two measures (Tenure and Total
Charges). The example below shows Total Charges. To
compare the values for another measure, use the down
arrow next to the Chart titles. Click the down arrow to
view other available measures.

Privacy - Terms
TECH TIP: You would need to select the new measure for
both charts for a “side by side” Comparison. You also may
compare different measures on each chart. For example,
you may want to compare revenue and expense or revenue
and headcount. Cognos Analytics will not display combined
values across different measures.

10 You may turn off the Comparison Stats by using the Hide

Site feedback
Comparison Stats button on the toolbar. Click the Hide
Comparison Stats button to view the update on the
canvas. Click the Hide Comparison Stats button again
to return the comparison to the canvas.

11 In order to easily compare values across the charts,

Cognos Analytics provides you with a Compare line.  To
use the compare line, you may click the drag handle
from either chart, left click, and move the line up or
down on the charts. 

12  Notice that when you move your mouse across the

charts, you have a side by side comparison indicator
called the Compare line.

Privacy - Terms
 Site feedback
By default, the Compare line is synced between the two
charts. You can un-sync the lines based on your settings for
Compare options.

13 Click the down arrow next to Sync. Here you see the
comparison options for the two charts. In addition to
setting the synchronization of the Compare line, you can
also un-sync the Axis Scale between the two charts.
Make no changes to the settings for this exercise. Click
outside of the Compare Options to close the options
list.

14 The Compare line can be turned off when not needed.

Click the Hide Compare line icon on the toolbar.

  Privacy - Terms
15 Save your Exploration.

Badges
Badges indicate that the system has identified
correlations to a target in your data. Badges are
displayed as green ovals with a number. The number

Site feedback
represents the number of correlations or correlation
combinations that were identified. Clicking on a badge
will provide additional information on the correlations.

Badges will appear on the x- or y-axis of a chart to

indicate that there are other data items correlated to the
item shown on the axis. If relationships to projected data
items are not shown on an axis (e.g. shown in a legend), a
badge will be shown to the right of the visualization title.

1 From Chart A, click the green badge on the Y-axis. This

will render identified drivers related to the measure on
the y-axis, Total Charges.

Privacy - Terms
 

2 On Chart A, click the green badge to the right of the

chart title. Any badges for information not on an x or y
axis will be shown here.

Site feedback
Here, you can see that Cognos Analytics Identified
correlations for data items that are not part of the
visualization but exist in the underlying data. Users can
select a correlation of interest to them to review the key
driver analysis.

Key Driver Analysis

A driver analysis visualization shows you the key drivers,
or predictors, for a target. The closer the driver is to the
right, the stronger that driver is. Cognos Analytics uses
sophisticated algorithms to deliver highly interpretable
insights that are based on complex modeling. Users do
not have to know which statistical tests to run on data.
Cognos Analytics picks the right tests for the data.
Privacy - Terms
1 From the list of correlations, select “Correlated with
Total Charges and Monthly Charges” to understand
more about how these fields are correlated.

Site feedback
 

2 Selecting a correlation from a badge list creates a new,

two-part card in your Exploration, containing a Driver
Analysis visualization and a preview of the selected
drivers.

Privacy - Terms
At the top of the screen is the key driver analysis in the
form of a “comet” visualization. This visualization identifies
and displays key drivers that factor into the data that was
identified as a target in the original visualization. Bubbles
are positioned along an axis of 1 to 100 to indicate the
strength of various drivers and driver combinations that
influence the target (at position 100). The proximity of the
bubble to the target indicates the predictive strength of the

Site feedback
driver. The closer to the target, the stronger the predictive
strength. The color of the dots indicates the number of
drivers. The size of the bubble indicates the number of
individual or combinations that were detected.
Visualizations below the comet give a preview of the driver
or driver combination that is highlighted along the axis.
 

TECH TIP: If you are familiar with the spiral visualization in

Cognos Analytics Dashboards, the Driver Analysis “Comet”
chart is like an unwound spiral. The target, in this case
Tenure, is at the far right of the diagram.

3 Single field drivers are represented in Blue, as indicated

in the legend. Hover over the blue bubble furthest to
the right. This blue bubble indicates that a single field,
Total Charges, is a driver of Tenure, with a strength of
73%.

4 Two fields, in combination, found to be drivers are

shown in Orange. The size of the bubble indicates the
number of combinations found to be drivers. Hover over
the largest Orange bubble to the see the various field
combinations found to be drivers.

Privacy - Terms
 

Site feedback
5 Click the Largest Orange Bubble. As bubbles are
selected, the visualization below the key driver analysis
will update. This brings into focus the combinations
found. A visualization is available for each for the
combinations identified and may be viewed using the
selection bar below the visualization.

TECH TIP: Depending on the fields identified in the driver

analysis, Cognos Analytics will generate different chart
types in the preview. These may be bar charts, heat maps,
bubble charts, or in the case of more than two drivers in
combination, a decision tree.

6 A Green bubble is a Combination of more than two fields

as drivers. Hover over the Green bubble (appears
“under” the large orange bubble on the axis) to see the
driver combinations.

Privacy - Terms
In this example, the Driver Analysis shows that the
combination of Total Charges, Monthly Charges and Device
Protection influence Tenure. However, you don’t yet know
whether it’s low charges or high charges or what type of
internet service drive tenure higher or lower.

Decision Trees

Site feedback
A Decision tree is used to highlight key insights in how
underlying data patterns drive the results of a chosen
target. Decision trees have a single target. If the target
field of the decision tree is continuous, for instance, a
measure such as Revenue, then the key insight indicators
highlight unusually high or low groups. If the target field
of the decision tree is categorical, for example, a product
or attribute, then the key insight is the most frequently
occurring category or categories of the target field within
the group.

1 Click the Green Bubble. Since the green bubble

represents more than two drivers, the visualization
preview renders a decision tree. Click the Add card
button to add the visualization to your Exploration.

TECH TIP: To interact with a visualization in Preview, you

can click on the visualization or use the Add Card option to
add it to your exploration to bring it into the full interactive
view.
Privacy - Terms
2 From the Navigation panel, open Explorations to see
your Exploration cards, if not already open. Select the
thumbnail of the Decision tree to open it on a new card
on the canvas.

Site feedback

Privacy - Terms
 Site feedback

Privacy - Terms
TECH TIP: Depending on the amount of data to render,
and/or your screen resolution, the decision tree may be too
large to optimize viewing. In these cases, a thumbnail of
the entire decision tree renders in the upper left corner. For
the User’s reference, a blue highlighted section indicates
the section of the decision tree currently in focus on the
canvas.

Site feedback
The box may be moved around to put focus on various
areas on the decision tree. The center scroll wheel on your
mouse will allow you to zoom in or out on the Decision tree
diagram on the canvas.

3 Click the Show details and slots icon on the toolbar to

open the panel and then click the data slot tab.

4 Currently, Tenure is set to the Target, but you are also

interested in Churn. From the Navigation panel, click
the Sources icon to open the data sources panel. Scroll
to the bottom of the data source list to select Churn.
Drag Churn over to the Target data slot to replace
Tenure.
Privacy - Terms
 Site feedback
 

5 The decision tree updates and you can now see the
patterns of drivers related to Churn where Partners were
used (Partners Includes Yes). To read the diagram, you
will start with the first driver listed from left to right,
Internet Service.

You can easily see that customers with Fiber Optic churn
more than those with DSL or no internet.  Of those, you can
see that customers with less than 16 months tenure are at
the greatest risk of Churn. But how does this compare to
customers who were not acquired through Partner
Contracts?  You can easily invert the filter to churn patterns
for non-Partner contracts. Privacy - Terms
 

6 Under Local filters, click on the three ellipses. Select

Filter and change the filter selection to No. Click OK.

7 The decision tree diagram updates to show the patterns

of churn for non-Partner contracts. Use the thumbnail or

Site feedback
your mouse scroll wheel to zoom in/out if needed. To
read the diagram, you will start with the first driver listed
from left to right, Tenure.

8 Question: What insight does this give you?

                                                                                                                     
                     

  Privacy - Terms
                                                                                                                     
                     

In reading the Decision tree diagram from left to right, you

see that the first driver identified is Tenure which has been

Site feedback
broken down into four statistically relevant groups. In the
diagram, you can immediately see what appears to be a
large difference in churn rates for customers that have less
than 10 months tenure versus those that have been
customers for 10 or more months.

9 Next, you’ll continue to read the decision tree diagram

moving to the right and focus on the second driver
identified.

Privacy - Terms
 Site feedback
10 Question: What insight does this give you?
                                                                                                                     
                     

                                                                                                                     
                     

Recall that customers have either phone service, internet

service or both services from Telco. The next driver
indicates that the statistically significant patterns driving
churn appear to be from those customers with Internet
service. Within this group, the decision tree makes it easy to
see that customers with Fiber Optic churn at a much higher
rate than those on DSL or without internet services.

Privacy - Terms
 Site feedback
11 Continue to read the decision tree diagram moving to
the right and focus on the last driver identified.

Privacy - Terms
 Site feedback
 

12 Question: What insight does this give you?

(HINT: Hover over the bars or use the data tray to view the
various Payment Methods)
                                                                                                                     
                     

                                                                                                                     
                     

Telco accepts payments from customers via mail,

electronic checks, bank transfers and credit cards. Those
paying by Electronic check have the highest degree of
churn.

Privacy - Terms
 Site feedback
Decision Rules
The decision tree diagram gives you a tremendous
amount of information and insight. But from all the
patterns and insights you discovered, how do you easily
prioritize actions? Cognos Analytics makes this easy by
providing you with the ability to easy slice and dice the
top and bottom results. It also lists out the rules in an
easy-to-read text format along with predictive strengths
for each.

1 Under Nodes, click the Top 5 target values and use the
pull-down arrow to set them to Yes.

Privacy - Terms
 Site feedback
The results for the top 5 patterns of customer churn are
presented. From here, you can see that those most likely to
churn are shorter tenured customers, either less than 3
months or those between 3 and 10 months, and those with
Fiber optic service are at the greatest risk of churning.
Your Exploration allowed you to identify specific decision
rules for customers with a high propensity to churn by
using the Decision tree visualization. Cognos Analytics will
provide you with the full list of Decision Rules and
predictive strengths for each in an easy to follow list that is
great for analysts looking to use these rules in their
predictive models.

2 Click on the Rules Tab. The visualization updates to

show the Decision rules in easy-to-understand text.
Note that the Churn visualization bar from the Decision
Tree is present along with the Decision Rules and record
count/percentages in easy to read text.

Privacy - Terms
 Site feedback
 

3 Use the down arrow next to Target Category and select

Yes.

Notice that the decision rules are listed in order of

predictive strength. The top two listed are the same as you
derived in the previous exercise using the Decision Tree
diagram.
This analysis has given you a tremendous amount of insight
into the customer base. You now have the insight you need
to propose the recommended steps to the Marketing team.

4 Save your Exploration.

Summary
Congratulations, you’ve completed your first Data
Exploration!  You can now bring your Explorations and
findings into Dashboards and Stories to communicate
Privacy - Terms
your findings to the Marketing Group.   To continue your
Analytics journey, please check out our other tutorials in
this Cognos Analytics 11.1 series:
• Dashboards, Visualizations and Exploration
• Data Preparation and Modeling
• Report Authoring
  Find these and more by visiting us at:
https://www.ibm.com/demos/collection/IBM-Cognos-

Site feedback
Analytics/ (https://www.ibm.com/demos/collection/IBM-
Cognos-Analytics/) You may download a .pdf version of
this workshop instruction here:  PDF | DTE HOL CA11.1.4
Getting Started with Data Exploration
(https://ibm.box.com/s/y5t3b6vun6ttd5357n01cdzobairwyyc)
How likely are you to recommend this tutorial to a friend or colleague?

1 2 3 4 5 6 7 8 9 10

How can we improve? (or any other comments)...

Submit

Privacy - Terms