Predictive Analytics DemoScript

8/20/2019 Predictive Analytics DemoScript

http://slidepdf.com/reader/full/predictive-analytics-demoscript 1/39

DOCUMENT CLASSIFICATION: INTERNAL AND PARTNER

General Information: Industry Name – Retail

Application – SAP Predictive Analysis

Country or Global - Global

A th I817248

DEMO SCRIPTRetail

Reward Customer Loyaltywith SAP Predictive Analysisand SAP Business Objects

Dashboards



© COPYRIGHT 2014 SAP AG. ALL RIGHTS RESERVED.

No part of this publication may be reproduced or transmitted in any form or for any purpose without the express

permission of SAP AG. The information contained herein may be changed without prior notice. Some softwareproducts marketed by SAP AG and its distributors contain proprietary software components of other software vendors.

Microsoft, Windows, Excel, Outlook, and PowerPoint are registered trademarks of Microsoft Corporation.

IBM, DB2, DB2 Universal Database, System i, System i5, System p, System p5, System x, System z, System z10,

System z9, z10, z9, iSeries, pSeries, xSeries, zSeries, eServer, z/VM, z/OS, i5/OS, S/390, OS/390, OS/400, AS/400,

S/390 Parallel Enterprise Server, PowerVM, Power Architecture, POWER6+, POWER6, POWER5+, POWER5,

POWER, OpenPower, PowerPC, BatchPipes, BladeCenter, System Storage, GPFS, HACMP, RETAIN, DB2 Connect,

RACF, Redbooks, OS/2, Parallel Sysplex, MVS/ESA, AIX, Intelligent Miner, WebSphere, Netfinity, Tivoli and Informix

are trademarks or registered trademarks of IBM Corporation. Linux is the registered trademark of Linus Torvalds in the

U.S. and other countries.

Adobe, the Adobe logo, Acrobat, PostScript, and Reader are either trademarks or registered trademarks of Adobe

Systems Incorporated in the United States and/or other countries.

Oracle is a registered trademark of Oracle Corporation.

UNIX, X/Open, OSF/1, and Motif are registered trademarks of the Open Group.

Citrix, ICA, Program Neighborhood, MetaFrame, WinFrame, VideoFrame, and MultiWin are trademarks or registered

trademarks of Citrix Systems, Inc.

HTML, XML, XHTML and W3C are trademarks or registered trademarks of W3C®, World Wide Web Consortium,

Massachusetts Institute of Technology.

Java is a registered trademark of Sun Microsystems, Inc.

JavaScript is a registered trademark of Sun Microsystems, Inc., used under license for technology invented and

implemented by Netscape.

SAP, R/3, xApps, xApp, SAP NetWeaver, Duet, PartnerEdge, ByDesign, SAP Business ByDesign, and other SAP

products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of

SAP AG in Germany and in several other countries all over the world. All other product and service names mentioned

are the trademarks of their respective companies. Data contained in this document serves informational purposes only.

N ti l d t ifi ti



TABLE OF CONTENTS

1. Demo Script Overview .........................................................................................................4

1.1.

Demo Description .................................................................................................................................... 4

1.2. Intended audience ................................................................................................................................... 4

1.3. Protagonists ............................................................................................................................................ 4

1.4 Key Messages and Value Proposition ..................................................................................................... 4

2. Technical Requirements ......................................................................................................5

2.1. Prerequisites/Restrictions ........................................................................................................................ 5

2.2. System Access Information ..................................................................................................................... 5

2.2.1. System Landscape ..................................................................................................................... 5

2.2.2. System Access ........................................................................................................................... 5

2.2.3. Users .......................................................................................................................................... 5

2.2.4.

Languages Supported ................................................................................................................ 6

2.3. Release for Used Components ............................................................................................................... 6

3. Demo Script .........................................................................................................................7

3.1. Story Flow ............................................................................................................................................... 7

3.2 Step-By Step Guide................................................................................................................................. 8



DEMO SCRIPT OVERVIEW

1.1. DEMO DESCRIPTION

This demo shows how a national retailer is able to quickly evaluate customer and sales data to retain key customers byrewarding their loyalty. The retailer uses SAP Predictive Analysis to create customer segmentations, identify the mostloyal and profitable customers, and develop targeted promotions..

1.2. INTENDED AUDIENCE

Please select for which audience this demo script is intended by selecting the appropriate boxes below.

Retailers

CPG

1.3. PROTAGONISTS

Marketing Analyst/Managers

1.4. KEY MESSAGES AND VALUE PROPOSITION

SAP Predictive Analysis:

• Determine hidden purchase relationships between products

• Optimize customer segmentation

• Visualize large amounts of data quickly• Improve up-sell and cross-sell success with stronger relationship definition• Identify traffic drivers and affinity products from transaction history



TECHNICAL REQUIREMENTS

2.1 PREREQUISITES/RESTRICTIONS

The software licensing requirements are needed to implement this solution are:SAP BusinessObjects Enterprise 4.0 SAP BusinessObjects Dashboard Design (Xcelsius)SAP BusinessObjects Web IntelligenceSAP Predictive AnalysisSAP JAM

Please download and install the latest SAP BusinessObjects Mobile App:

Launch App Store and search “SAP BusinessObjects Mobile” and you will see the App Install the app

2.2 SYSTEM ACCESS INFORMATION

2.2.1 System Landscape

System Landscape, this Process is built in

SAP Demo Cloud



2.2.3 Languages Supported

Please select the standard languages in which this demo is available (i.e. translated into).

English PortugueseGerman Japanese

French Simplified Chinese

Spanish Korean

2.3 RELEASE FOR USED COMPONENTS

Software Component Release Support Pack

SAP Predictive Analysis1.0.9

SAP BusinessObjects Mobile4.4



DEMO SCRIPT

3.1 STORY

1) It is close to quarter end, Jack Lee, Operations Director at BestRun Retailer, notices from his Dashboard thatsales will not meet target would like to run some promotions to increase sales. They would like to identify themost profitable customers and their purchase preference, in order to reward their loyalty and promote cross-sell opportunities.

2) Lead Analyst, Mike starts with high level visualization of where their customers are and their demographicinformation.

3) To gain more insight, Mike uses Clustering to perform customer segmentation. He discovers a small cluster ofcustomers that had made relatively high amount and item in purchases. They are also younger, more affluentcustomers.

4) Mike zooms in this specific group of customers, he also notices that most of them made last purchase only 1week ago. This looks like the loyal, profitable customers they want to target. Mike also discovers that thesecustomers make purchases online.

5) Mike exports the segmentation result to a CSV file.

6) Mike now needs to figure out what items can be offered in the promotion to this group of customers. He

conducts an association run for market analysis. Mike concludes that focusing on the cross-sell relationship

between these items would lead to the a win-win situation for both Best Run and its loyal customers.

7) Best Run sends special promotional ‘online’ offers to this group of customers. The promotion is a success,

along with other actions that management has taken, Best Run closes out the quarter above target. In

addition, due to analysis done in SAP Predictive Analysis, Best Run is now able to track their top customers ‘s

KPI on a ongoing basis. Best Run is also able to create individual view of their most loyal customers’ activities

and metrics. This dashboard can also viewed on a mobile device. Team members can collaborate using SAP

Jam to quickly create personalized offers.



©SAP AG 2014 / SCENARIO ID: 8860

3.2 STEP-BY STEP GUIDE

3.2.1 Install SAP BusinessObjects Mobile App on iPad

Step 1.0 Download and install SAP BusinessObjects Mobile App on iPad

What To Do What to Say (include necessary Screenshots)

Launch SAP BusinessObjects Mobile app oniPad and connect to the BI platform, refer to 2.1.

User name: smithjo

Password: welcome




3.2.2 Demo

What to say

It is close to quarter end, It is close to quarter end, Jack Lee, Operations Director at Best Run, reviews his Sales Performance Dashboard. He

notices that Total Sales is not going to meet target.

Though HANA is not directly used for tstoring the back end data displayed on these dashboard, you can easily talk here about what HANA can bring

to your customers/prospects

When used as a transactional or Datawarehousing platform, SAP HANA provides on-the-fly analysis for all combinations of data. You can make

immediate decisions using real-time operational analytics whether the data comes from your SAP® applications, third-party solutions, or custom

applications.

Big Data Warehousing:

SAP HANA unleashes the potential of Big Data with its ability to handle large volumes and a variety of structured and unstructured data in real-time.

Build a data warehouse integrated with Hadoop or easily migrate your existing SAP NetWeaver Business Warehouse (BW) to SAP HA NA and make

decisions within the window of opportunity.


Save swf files customerloyalty.swf and customer Analtycis.swf to your local drive.

First open the customer Loyaltyswf file.

Present the dashboard, firstshowcase its functionality suchas displaying a secondarychart.

Click on the ‘*’ next to PromoSales

Jack is able to view metrics such as sales, inventory, top products on the Sales Performance dashboard. Thisparticular dashboard is develped in LAVA (Lightweighted Applied Visual Analytics), the newest approach toanalytics visualization developed by SAP. It is simple and systematic. Jack is able to view a secondary chart bytapping on the ‘*’ when he would like to get additional information.




Then point at the sign nextto Total Sales chart. And pointout the downward trendpresented the blue line which isactual sales

Jack notices the red alert signals next to the Total sales and net sales metrics. It seems that sales for this quarteris not meeting target.




The same dashboard can beviewed on the SAPBusinessObjects Mobile App oniPad.

Open Mobile BI followinstructions described in 2.2.2system access portion of thescript.

logon with below account:

smithjo/welcome

Tap on

Lava Dashboard works on mobile devices too. When Jack is outside of his office, he can easily review the samedashboard on his iPad and make the same observations.




Tap on Retail from the list under‘Server’.

And tap on




What to say

Jack realizes that they need to run some promotions to increase sales.

One of the action items is to identify the most profitable customers and their purchase preference, in order to reward their loyalty and promote cross-sell

opportunities. Best Run will utilize SAP Predictive Analysis.

Use also this opportunity to discuss SAP HANA advantages for PA:

SAP HANA is a fully ACID compliant, in-memory, columnar, massively parallel processing platform that provides common database for online transactional

processing (OLTP) and online analytical processing (OLAP), eliminating redundancy and latency. It also unifi es end-to-end data processing with search,

text, geospatial and predictive analytics on top of the in-memory columnar foundation, further extending the benefit of redundancy and latency elimination toall data processing workloads. SAP HANA provides a powerful suite of predictive, spatial, and text analytics librairies that can run across multiple data

sources. Visualize new opportunities and gain deep insights with unified analysis of all of data types.

SAP HANA also enables real-time prescriptive analytics by reducing layers of processing and accelerating complex computations required for scheduling

and simulations.


Start SAP Demo Cloud CSA(Green Phoenix) Showroom and

Click on Desktop.

You should see a Predictive Analysis icon once remotedesktop starts.

Save Excel file first to thedesktop drive.Note: You can obtain the Excelfile from the same place in SAP

Demo Store where you downloadthis demo script.

At next screen, click on ‘NewDocument’

Mike, Lead Analyst of BestRun, starts his analysis by opening SAP Predictive Analysis and beginning a “NewDocument”.




Click on

Select Retail.xlsx from your localdrive and select sheet:Customers.

And click on Create




Click on “Prepare ” tab

Move to “Facets” view of the data

Mike uses data from customer table, which contains customer demographic information gathered from loyalty cardprogram and transcations information. He can view the “Facets” view of our data, which provides the frequencies of thedifferent values for our variables.

Click on icon next to“Income” Column.

Select sort from the drop downmenu and choose Sort dimensiondescending

In Facet view, Mike can directly sort to view the values of each varialbe. For instance, he click on the sort button forIncome and notices that the highest income of the customers is $307064 and there is one customer with that amount ofincome.




Click on “Visualize” tab

Click on next to Dimension“Region”

Select “create a geographichierarchy” and click on “ByNames”

Predictive Analysis has visualization and graphing capabilities built right into the modelling tool, so there is no need tomove between programs when Mike conducts his analysis. Mike would like to perform some high level analysis first.He starts with a Geo chart to gain a high level view of the customers.




Click on Confirm.

Click on OK




Choose Geo Choropleth Chart

Drag and drop the Region fromHierarchies to the Geographyportion of the chart.

Drag Customer_Sum fromMeasures to Value portion of thechart.

Mike is able to tell right away that California and Midwest states have the highest number of customers. And he alsonotices the states with very light colors, these are the states with fewer customers.




Change to “Geo Pie chart”

Drag Time Last Purchase Madeto Overlay Data

Mike can get an additional layer view of the data quickly by adding Time last purchase made to the same chart.




Click on share tab.

Select the visualization and clickon “Send visualization by mail”

Mike can share this visualization in one click via email with his colleagues.




Click on “Predict” tab

Drag and drop R-K-Meansalgorithm to the workspace

The Geo map chart provides a high level overview; but does not provide the in depth analysis needed to identify the mostloyal customers. To conduct advanced analysis, Mike moves to the Designer area of SAP Predictive Analysis to set upthe workflow needed for his analysis.

First, Mike would like to run a Clustering Algorithem to segment the customers.




Double click on R-K-Means iconto Configure Properties.

First Check

Then uncheck these twovariables:

Customer numberPostal code

Set cluster number to 8

And click on

Click on on top right of window to run analysis.He starts the analysis and clicks Yes to proceed to view the result. And then proceed to Visualization area when he receives the run success window.




Click on icon thebottom corner of the Result view.

Customers are segmented into 8 groups. Mike notices clusters 1, 2, 6 and 8, these clusters all have high density, whichmeans customers within each of these groups share very similar characteristics.




Select the scatter matrix chart

in the result view.

Mike observes that cluster 1 consists only of the younger age group.




Click on

Select the component

From the top of the window.

Select bubble chart.

Click on the down arrow next to

the following measures to select Average for each measure.o # of purchases last 6

monthso Ageo CreditScoreo Total purchase lifetimeo Total items purchased

last 6 months

Mike now switches to Visualize mode to get a better understanding of the segmentation result.He would like to know the purchasing pattern for each cluster. He starts with a bubble chart.

Mike notices the big bubble that is apart from the rest of bubbles. Based on the legend, this bubble represents cluster 1customers. It seems that cluster 1 customers have the highest lifetime purchase amount, and highest number of totalitems purchased in the last 6 months.

Note: Cluster number is assigned randomly when you run th is a lgor i thm. It might not be cluster #2 when you

run th is in PA . Please note the cluster wi th the least numb er of records; th is is the cluster that wi l l be used in

the fo l lowing steps.




Then set up the bubble chart asshown on right.

Select “Combined column linechart with 2Y- Axes”

Next, Mike introduces some demographic variables to the analysis and discovers that cluster 1 customers are a younger

group of customers with high credit scores. This is definitely the type of customers that retailers would like to target.

And enter setting as below:




And enter setting as below:

Select “Tag cloud” for your

visualization now.

Enter setting as below:

Next, Mike conducts analysis on this cluster customer alone. First, he creates a tag cloud chart to check their buying

frequency and most recent buying period. It turns out most of them made a purchase pretty recently, ranging from 1week ago to 6 month ago. And they purchased relatively high amount of popular items –as indicated by the colorintensity of the words. It seems that Mike has found the loyal customers he is searching for. However, Mike alsonoticed that more customers in this cluster made last purchase 1 month ago, so they could be at the verge ofabandonment.




Click on “Add filter” at the top ofthe window.

Select cluster number for addingfilter from the drop down menu.

Choose cluster number 1 andclick on “OK”

Click on “Pie Chart” Now that Mike has found the loyal customers he is searching for he needs to understand how they purchase Mike uses




Click on Pie Chart

Enter setting as below:

Leave ClusterNumber filter infounchanged.

Now that Mike has found the loyal customers he is searching for, he needs to understand how they purchase. Mike usesa pie chart and discovers that 77% of them shopped online. This fits the younger, affluent customers’ shopping pattern.

Click on TabMike would like to export the segmentation result to a file. He chooses to export as a CSV file. This output assignscluster number to each customer record.

Drag and drop CSV writer from




Drag and drop CSV writer fromthe data writers tab on the rightside of the window.

Double click on the CSV writer toconfigure its properties.

Enter path and file name youwant the export to be.

Click on Done.

Select Run to execute theanalysis

When the execution statusprompt window appears click on“OK”




Click on tab

Click on “New Data Set”

Choose the same file and selectthe Transactions sheet.

Click on “Create”.

Click on tab.

Drag and drop “R- Apriori” fromthe algorithms tab to theworkspace

Now Mike needs to decide what items can be used to reward this group of loyal customers and address the potentialabandonment issue. He conducts a R-Apriori Association analysis for basket analysis. The apriori algorithm will output a

set of rules representing cross-sell or up-sell relationships. First, Mike selects the item column to be “product”, as he

wants to start by conducting his analysis at the product level. Next, Mike selects transaction ID column to be “trans”.

“Support” is the minimum percentage of t ransactions in our data that need to contain items from the categories that we

are analyzing. If we are looking at the purchase relationship between categories A and B with a support level of 5%, then

greater than 5% of the purchases must contain products from categories A & B. Mike sets this value at “0.05”.

“Confidence” is the liklihood that a customer will purchase from category B given they have purchased A, in essence what

percent of transactions that contain A also contain B. Mike sets this value at “0.1”. Now he can save and run his market

basket analysis.




D bl li k th l ith t




Double click on the algorithm toconfigure its properties.

Configure the settings as shownin the figure.

Click on “Done”.

Click on “Run” to execute thel i




analysis.

When the execution statusprompt window appears click on“OK”

Select the “Association chart”from the result view.

Mike will get a notification that his analysis has been successfully executed, and he will be able to see the relationshipsbetween his purchases.

Now he can evaluate the results of our analysis with the provided purchase relationships. The “PreRule” productcategories on the far left are the traffic driver products. The “PostRule” product categories are affinity products, basketbuilders that are most often purchased with something else. The “Lift” value on the far r ight is the one number used toevaluate these relationships as cross-sell or up-sell opportunities, the higher the better. Generally a “Lift” value over 1 isconsidered a viable opportunity.

The basket analysis results suggests that customers usually buy collection pencil skirt when they shop for chiffon tops

and suede pumps. Mike concludes that focusing on the cross-sell relationship between these items would lead to the a

win-win situation for both Best Run and its loyal customers.

Since Mike has exported the segmentation result already, in the output table, each record is assigned a cluster number.

He can easily identify cluster #1 customer and email them an ‘online special offer of ‘buy 1 chiffon top, get 1 collection

pencil skirt 75% off’.




Open the Sales Performancedashboard again Click on

The promotion is a success, along with other actions that management has taken, Best Run closes out the quarter abovetarget In addition due to analysis done in SAP Predictive Analysis Best Run is now able to track their top customers ‘s




dashboard again. Click onScenario 3 tab on the top rightcorner.

Click on ALL next to CUSTOMERpoint out the upward trend inTotal Sales chart.

target. In addition, due to analysis done in SAP Predictive Analysis, Best Run is now able to track their top customers sKPI on a ongoing basis.

Click on TOP In addition, due to analysis done in SAP Predictive Analysis, Best Run is now able to track their top customers ‘s KPI ona ongoing basis. They have created a Top filter specifically to track metrics related to this group of customers.




a ongoing basis. They have created a Top filter specifically to track metrics related to this group of customers.

The same dashboard can beaccessed on iPad by followingthe instructions outlined earlierfor

The same dashboard is aviliable on iPad for consumption out of office.

Open Customer Analytics swf file to showcase the various

Best Run is also able to create individual view of their most loyal customers’ activities and metrics. And take immediateaction to retain/reward these customers. Team members can collaborate using SAP Jam to quickly create personalized




metrics tracked for the individualcustomer. Also mention that SAPJam can be utilize forcollaboration.

g q y poffers.

The same dashboard can beaccessed on iPad by followingthe instructions outlined early.In this case, tap on

The exact same dashboard can be viewed on an iPad.

Predictive Analytics DemoScript

Documents

Transcript of Predictive Analytics DemoScript