0504 -SAP Data Conversion at Kellogg

[ 2 Go-Lives, 0 Data Defects

Wynter Johnson, Daniel Bunting, Emil Beloglavec

[ Learning Points

Data Conversion is a business process, not just a technical

exercise
Data conversion is a collaborative process that requires
coordination across all teams
The business owns the data
Development and testing of conversion components should be
an iterative process
Data Services is a good tool to help facilitate conversion

Real Experience. Real Advantage. 2

[ Best Practices
Develop a set of data guiding principles and follow closely
Set detailed and specific expectations for responsibility and
timing of data related tasks (up front)
Develop a disciplined documentation process for key design
documents
Data Relevancy rules, Conversion Functional Designs, Mappings
Measure data conversion loads against predefined criteria
Set goals for each mock load before loads start
Measure overall progress against having business ready data, not
just being able to technically load to SAP
Document risks and issues as soon as they are known
Risks and Issues are not bad things. They are tools to help solve
problems

Real Experience. Real Advantage. 3

[ Kellogg
Our Vision
To enrich and delight the world through foods and brands that matter.

Our Purpose
Nourishing families so they can flourish and thrive.

Company
Established in Battle Creek, MI in1905. Today we have 50+ plants on all but one
continent with 31,000+ employees and $13.2 billion annual revenues.

Leading Brands
Our brands, which are produced in 18 countries and marketed in more than
180 countries, include Cheez-It, Coco Pops, Corn Flakes, Eggo, Frosted Flakes, Kashi,
Keebler, Kelloggs, Mini-Wheats, Pop-Tarts, Rice Krispies, Special K and many others.

About our Company

Please visit www.kelloggcompany.com for more information about our company and our
brands.

Real Experience. Real Advantage.

[ Accenture
Accenture is a global management consulting, technology services and
outsourcing company. Our mission is to help our clients become high-
performance businesses and governments.

Business Overview
$25.5 billion in annual revenues (FY ended August 31, 2011)
Over 244,000 employees
Offices and operations in more than 200 cities in 54 countries
Accenture serves approximately 4,000 clients worldwide

Accentures SAP Practice

34,000+ SAP-skilled professionalsmore than any other systems integrator
SAP Consulting Services delivered to 1,300 clients last FY - circa 1 in 3
Accenture clients overall
80+ SAP industry (vertical) assets & 20+ SAP cross-industry (horizontal)
assets
Won 20 SAP Pinnacle Awardsmore than any other SI

Real Experience. Real Advantage.

[ Presenters
Wynter Johnson
Kellogg - Data Conversion Lead

Dan Bunting
Accenture Data Team Lead

Emil Beloglavec
Kellogg - Business Objects Information Management

Real Experience. Real Advantage.

[ SAP Upgrade
What Upgrading our enterprise systems platform
Simplifying, standardizing, and automating

Why Improve integration and access of information

Better leverage SAP capabilities
Create the foundation for the future
Sustain and support recent business changes

Why now Building for the future is a priority

Real Experience. Real Advantage.

[ Components of Data Migration

Real Experience. Real Advantage.

[ Continuous Improvement

Discover

SAP BusinessObjects
Remediate Understand
Information Steward

Monitor Define

Real Experience. Real Advantage. 9

[ Information Management Value Triangle
Gaining value from data starts with solid data designs and properly converted data. Data
management and governance must be in place to ensure data integrity. Business Intelligence
can then provide the information necessary to make fact based real time decisions.

Real Experience. Real Advantage. 10

[ Data Conversion - Effort
It is important to recognize that data conversion is a business process and it should not be
treated as a technical exercise.
Conversion Effort
5%
10%
5%
Extraction

Technical
Transformation
Loading
80%
Business Activities

Data Identification
Data Conversion Design
Data Cleansing
Data Construction
Data Validation

Taking a holistic approach to data conversion is necessary to avoid common data related
issues and overruns due to improper planning.
Real Experience. Real Advantage. 11
 [Data Conversion Holistic View
Language Strategy Successful data conversion needs to account
Relevancy Criteria/Rules
Functional Design hand offs for all aspects involved in the end to end
On/off site Considerations and
Planning conversion process.

Scope Management
Workplanning and Estimating
Security Strategy Resource Planning
Validation Strategy Status & Progress Reporting
Cleansing Strategy Issue & Risk Management
Deployment Strategy Cross team integration coordination
Environment Strategy Project Management Integration across Symphony
End to end PM, Integration within Symphony
Thought Leadership
Conversion Positions/Approaches, Leading Practices
Technical Functional Deployment

Migration Preparation Conversion Analysis and Planning and

and Execution Functional Design coordination of BU
activities
Set-up migration tools
Establish/maintain legacy application
Establish BU Engagament Plan
connections
Coordinate BU activities
Write conversion routines
BU/ Regional Team
Configure Migration Tools
Apply Legacy Knowledge
Run Migration Tools
Support Functional Designs
Generate Cleansing / validationReports
Develop Functional Designs Support Mappings
Develop Functional Mappings Collect, cleanse, and validate data
Analyze Data & Reports
Develop Validation Processes
Integrate with Process Teams
Integrate with Data Network

Real Experience. Real Advantage. 12

[Conversion Team Structure & Key Interaction Points
Data Conversion Team

DC Leadership Team Data Lead

Data Conversion Manager Cleansing Manager

Cleansing Coordinator(s)
DC Technical Team DC Functional Team
Data
Regional Cleansing
Cut-over R3
lead
RQT Tech. Team RQT Fnx. Team
Data
Regional Cleansing
Cut-over leadR2
MFG Tech. Team MFG Fnx. Team
Data
Regional Cleansing
Cut-over R1
lead
M&S Tech. Team M&S Fnx. Team
RQT Data Cleaners FI Data Cleaners
FI Tech. Team FI Fnx. Team
MFG Data Cleaners PRO Data Cleaners
PRO Tech. Team PRO Fnx. Team
M&S Data Cleaners SC&L Data Cleaners
SC&L Tech. Team SC&L Fnx. Team Kellogg Business SMEs

Tech.Teams Process Teams Local Users 13

Real Experience. Real Advantage.

[ Waves

Real Experience. Real Advantage.

[ Mock Loads

Mock Load 1 Mock Load 2 Go-live Production

Mock Load n Simulation Load

Data conversion is made up of a number of Mock Loads 3-5 generally. We are planning 3
Mock loads for Release 1.
Mock 1 load is to test process and does not include a all the data, nor is it cleansed or
transformed.
Mock 2 load is to begin testing a cleansed and transformed data using a large data set.
Mock 3 load is all the data and a large percentage of the cleansed and transformed data.
Go-Live Simulation is a test of the complete process.

Real Experience. Real Advantage.

[ Mock Load Phases

Conversion Load Post Load Defect

Pre-Load
Design / Data Validation Management
Validation
Development

Real Experience. Real Advantage.

 [ Change Management
A disciplined approach to controlling change and monitoring the rate of change is
critical to the success of an ERP program.

Go-Live Go-Live
Amount of Change

Failed Project Missed Test Cycles

Unforeseen Affects

Risk of Change
Reactionary Process
Hastily Made Decisions

Successful Project

Project Timeline Project Timeline

Any change can have a ripple effect on other areas and the risk profile for any
change increases as Go-Live approaches.

Real Experience. Real Advantage. 17

[Solution Change Process Overview
A solution change process should be implemented for any change that affects the
solution after design is complete.

Implement when solution design is fairly stable:

Configuration complete, object design fairly far along, prior to completion of mock
loads (ideally implement when 2 are left)
Required for changes to:
Requirements, Processes, Configuration, Development Objects, Roles, Transactions,
Security, Data
Only changes that significantly affect solution integrity should be allowed:
Solution wont function correctly, transactions will fail, significant manual effort will be
required, change will save $xxx,xxx

All changes must follow the change process regardless of how minor they may
seem (1 minor change may not be a risk but the aggregation of 100 minor changes
is).
Real Experience. Real Advantage. 18
[ Technology - Toolset
SharePoint (2 sites)
MS Excel
HP Quality Center
Data Services
WinShuttle

Real Experience. Real Advantage.

[ Data Services

Unix or Wintel
Physical HW versus Virtual HW
Supporting Database for Transformations
SW Version 3.2 versus 4
Environments DEV, QA, PROD

Real Experience. Real Advantage.

[ Landscape

Two separate Data Center facilities

DS Designer (remote desktop server)
Staging area UNIX shared drive
SAMBA
Access ECC/BW, DS, Remote Desktop Server for Developer

Real Experience. Real Advantage.

[ Data Flow
Extract
Flat files
Excel Files
SAP Table
ABAP
BW OpenHub
Transformation
Load
IDOCs
LSMW (Legacy System Migration Workbench)
ABAP
BW InfoSource/DataSource

Real Experience. Real Advantage.

[ Data Services Architecture
Release 4.X
Release 3.X
Management
Console Designer

Web Application Server

Access Address Real Time

Job Server
Server Server Clients
Data Services

DS
Dictionaries
Repositories
&
Directories
Database Server

Real Experience. Real Advantage. 23

[ Challenges
Data Services Designer
Excel spreadsheets as source of data
2003/2007 version
Formatting of cell
Complex relevancy rules in extracts of huge tables (MSEG,
BSEG, JEST)
Custom ABAP transform
Sales/Purchase Orders, Plant Maintenance
Versions of conversion tables
Dedicated Schemas
Persistent Datastore
Communication, communication and communication
Keeping focused on future state (dont duplicate current)

Real Experience. Real Advantage.

[ Examples

ECC to ECC (Material / Customer / Vendor)

Sales History
36 months worth of data (~ 200 M rows)
ETL (72/2/36)
Document Management System (DMS)
ABAP Extract/Load
(SAP Database and Vault) SAP Database

Real Experience. Real Advantage.

[ Resources

Werner Daehn ( SAP )

BusinessObjects Data Services Tips and Tricks:
http://wiki.sdn.sap.com/wiki/display/BOBJ/BusinessObjects+Data+Services+Tips+and+Tricks

Data Services Supplement for SAP ( end user manual )

Real Experience. Real Advantage. 26

[ Information Steward

Sustainability
Proof of Concept
Information Services Platform / Data Services 4 / IS: 64 bit
no memory restrictions

Real Experience. Real Advantage.

[ Key Learnings

Data Conversion is a business process, not just a technical

exercise
Data conversion is a collaborative process that requires
coordination across all teams
The business owns the data
Development and testing of conversion components should be
an iterative process
Data Services is a good tool to help facilitate conversion

Real Experience. Real Advantage. 28

[ Q/A

Real Experience. Real Advantage. 29

[
Thank you for participating.
Please remember to complete and return your
evaluation form following this session.
For ongoing education on this area of focus, visit the
Year-Round Community page at www.asug.com/yrc
]
[ SESSION CODE: 0504

Real Experience. Real Advantage. 30