SlideShare a Scribd company logo

Measure Data Quality

Z
ZavalaJV

The document discusses establishing metrics to measure data quality initiatives at Delphi by building an Information Quality Index to track attributes like missing, late, or inaccurate data across 5.96 million material master records and fields, extracting relevant data using SAP tools like QuickViewer to analyze data quality. It also covers monitoring business processes and information flows to simplify data and improve quality.

Related topics:
Data Quality Insights
1 of 60
Downloaded 11 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Measure Data Quality
Case Study: Utilize Quantitative Standards and Metrics to Measure Data Quality Initiatives: A Real- World Case Study from Delphi Jose Zavala Delphi © 2008 Wellesley Information Services. All rights reserved.
In This Session ... • Reveal the magnitude and complexity of data required to satisfy Generation C (consumers) demand Build a map with these interacting business elements Quantify effects of these elements to the auto industry • Find the value in creating a strong, yet simple data strategy (simplify) • Review a model of a dynamic query creator (rule of thumb) to easily create/expand logical tests applied to your static dataset (Material Master Records) Build from this model into other dynamic data elements (effectiveness metrics) • Extend this approach with closed-loop cycles monitoring the bottom line (cash flow/inventories) 2
What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 3
Generation C (Consumers): Aggressive Demand Creative Communicated Connected Conversation Challenged Community Charming Customized e-Commerce Content Cash Control Channel Consumer 2.0 4
Consumer Touch Points and the Information Flows FACT: Order-To-Cash Research Blogs Music Peoples’ needs and Movies Social Price? desires are easily captured with current technology Gaming Shopping Products are Quality engineered, Speed manufactured, and delivered fast! Mass-Customization 5
Example: Buying a New Car We want this ... DATA s BUSINESS r ate e ne ELEMENTS a tg Th But OEMs offer all these choices Speed Color Technology Size Class DVD Products are Multiple CD Service Finance Headlights engineered, Brand Class Price Seats Value manufactured, and Transmission Economy Sound delivered faster! Safe Sunroof Engine Reliable Sharp … When supported Power Doors Resistant with good data Shape 6
Magnitude and Complexity of Data Required Currencies and Markets Modular Products Delphi Divisions Electronics & Safety Packard Electrical/Electronic Architecture Powertrain Systems Steering Thermal Systems Product & Service Solutions Laws and Regulations US SEC EU Canada Lead Free 7
Magnitude and Complexity of Data Required: Material Masters EU Material Master Recs: 237,000+ 1.99 M Delphi Packard 84 fields NA Material AP Material Master Recs: Master Recs: 165,000+ 237,000+ 116 fields 87 fields 1.91 M 2.06 M Combined Material Masters/Fields = 5.96 million 8
The Power to Simplify: Material Masters Quality Aspect • Each intersection (data element) could be: Missing or late Delphi Packard Inaccurate ATTRIBUTES (fields) MATERIALS Sales Engineering Purchasing Finance Prod Control Logistics 5.96 Million Material Masters/Fields = Error Opportunities 9
What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 10
SAP Data Architecture • So, how do you check for a data element that is: Missing or late? Inaccurate? • If data is created/maintained by Sales, Engineering, Purchasing, Finance, Product Control, or Logistics … … Then how do you measure its quality? 11
SAP Data Architecture (cont.) • SAP has a well structured set of inter-related tables to minimize size of storage as well as to improve response time • Realizing that we are building a data quality index and because size of data files are not a restriction, Sales VBEH VBLB VBSS we can proceed to “fill in the blanks” and create a VBEP VBKD VBFA VBRK data matrix with key data elements VBKE VBPA VBAP VBUK VBBE VBRP T16OQ TLGR T001L Unfold! SO11 T161T MSEG System Purch T069 T437L EKET EORD T160R T024D SO12 MVER MARA EKPO MARA MARC T161F T134T EKNN MKPF T160W T604 T157H EINE EKKO SO31 T006 T005 ADRP STKO PLKO MAKT CRCO CRTX KAPE BVOR BSIP BSAS System Eng Plan FICO T100 T024 PLPO AFPO CRHS KAKT PAYR BSAK T247 ADR2 STPO MVER CRHD KAZY BSAD KNC1 MARA MAST CRCA MLAN TAPLT AOQD AFKO MARM CRID KAPA BKPF BSIS T777A T023 ADCP MAPL PLSO T001W CRHH KAKO T024C BSID BSIK LFC1 12
SAP Data Architecture and the QuickViewer SQVI Tool • Off-the-shelf SAP contains over 100 data fields as part of master data records in multiple views • First, identify data fields as part of the master data record of interest • Then, define ownership for data creation and maintenance 13
Four Steps to Extract Data for the Information Quality Index 1 3 4 2 1: Join Definition 2: Field Selection 4: Validate Results 3: Save/Test 14
Step 1: Join Definition • Using QuickViewer — SQVI Keep table joins simple, as this will drive your processing time Field selection should consider current and future functionality 15
Step 2: Field Selection • Select data fields of interest — SQVI • Data will be generated in the same order • Selection fields are part of interface screen created, if run with transaction START_REPORT 16
Step 2: Field Selection (cont.) • Create All Queries According to Areas of Interest — SQVI Use consistent query names according to the nature of the project Do not bring unnecessary data fields to the model 17
Step 3: Save/Test • Generate programs and get report names — SQVI Test queries using transaction START_REPORT These queries can also be used to validate data 18
Step 4: Validate Results • Schedule a download job (t-code SM37) Add all queries to the download job as steps Consider execution times to avoid system overloads 19
Step 4: Validate Results (cont.) • Schedule a download job (t-code SM37) Daily analysis seems to be a good choice Data needs to be fixed, but most important is to enhance the business processes as well 20
Step 4: Validate Results (cont.) • Get to your spool list and export items as text (t-code SP02) Queries over empty data tables result in no spool output Download to user SAPGUI folder for conversion and upload to SQL Once files are downloaded to your local drive, user should get an SAP notification 21
Step 4: Validate Results (cont.) • Find your items in the SAPWorkDir folder using Windows Explorer Make sure the file size is manageable Downloaded jobs can be directed to other users when scheduled 22
Step 4: Validate Results (cont.) • Complete the validation process (text editor) This is a standard output when the spool item is “Export as Text” Use the tool of your choice to upload to SQL 23
Step 4: Validate Results (cont.) • Upload the files to the SQL Server (MS-SQL) Data fields should be uploaded in the same sequence There should be one table for each query created 24
A Self-Sufficient Data Analysis System Algorithm Truncate existing data Build New SQL Microsoft START Statement/Query Internet MM tables + Results Explorer Reload MM tables Select Table called Browsing from SAPGUI in Audit Rules Exceptions Tag each record Report with Client, Region (AuditResults) Business Unit & Plant Open Apply Audit Rules Audit Rules Table Scope (filter records) SP02 SAP SM37 SM36 Initiate variables Row = 0 Apply SQL Command under Rule to Field SQVI Row = Row + 1 Segregate non-complaint Read AuditRule Row # data to AuditResults END Yes End of file No No End of Target Audit Rules? Table found? Yes 25
Daily Refresh of Data Loaded to SQL Engine • Preparing SAP download jobs Once target tables and data fields are identified, jobs are scheduled to run at 1:00 AM EST Monday through Friday A Master Data Engineer (MDE) gets them into their SAP account • Retrieving data from SAP to bring to a local system Files are then downloaded as text to a local PC Information is not structured at this time • Uploading data to the SQL Server from a local system Files are uploaded directly to the SQL Server from the production environment 26
Rules of Thumb (ROT) Examples 27
Additional ROT System Tables (Part of the SQL Model) • MMPlantPBU table Helps classify each record by Plant and Business Unit Plant (key), Business Unit, Plant Description, Master Data Manager (coordinator) • MMAuditSummaryHeader Keeps daily audit results Client, Region, Business Unit, Audit Fields, Total Records, RunDate & Plant • MMAuditSummaryItem Provides count and links for non-conforming records by rule RuleNumber, ErrorMessageExplanation, ErrorLevel, Owner, Client, Region, Business Unit, Errors, RunDate, Plant 28
What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 29
Rules of Thumb: Browsing the AuditResults Records Report Name Selecting a Target Dataset MMs records audited Total fields audited Navigation MM Recs X Fields Zoom Total exceptions found Search PPM calculation % deviations Date/time % compliance to ROT stamp Info Quality Sigma Level Name of Errors per BPO area Business Process Owners (BPOs) and Total Rules created by them Exporting Formats Continuous Improvement Model 30
Rules of Thumb: Browsing the AuditResults Records (cont.) • Non-conforming data to ROT are presented by business unit (PBU) or plant level, following a standard set of information 31
Data Views Available — Following the Rules of Thumb • When users are browsing non-conforming records, they can target a given client, region, and business unit data set • Specific Rules of Thumb (logical conditions) are established or approved by the business process owners part of a business unit Then they are put together in SQL Server language syntax by Master Data Engineers • The list of non-conforming records looks like this: 32
Multiple Formats Available When Exporting Records • The SQL Server Reporting Service contains a set of standard formats End users can manipulate data after non-conforming records are identified 33
Benefits of the Reporting Services • Quick access to information Find any existing value By rule By plant Etc. • Exporting formats available Most common formats are available Helps processing errors Facilitate Error Analysis such as: Counts, average, etc. 34
Benefits of the Reporting Services (cont.) • Data is available for massive updates (t-code MM17) Target Material Masters are copied to the clipboard and provided to MM17 MM17 can change up to 800+ records at one time Tables and fields are identified Update is done in just a few steps Processing time is minimized 35
Describing the Global ROT System • Benchmarking is made possible by networking among Master Data managers Within each region Within each business unit At different levels of deployment phase (QN4 environment) Comparing: Error Level Logical statement Scope Applicability Customized values RefPackingMaterial <> “REFPACK” PN1 AP DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active PN1 AP DEEDS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active PN1 NA DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active QN4 NA DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active QN4 NA DEEDS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active 36
Elements of the Global ROT • 5-digit rule number 3-digit data field number A sequential, numerical ID that goes along with a given SAP data field It’s standard for all PBUs in every region Currently have 148 fields available for creation of rules, and only 103 are partially covered Can grow as data becomes available within SAP, in a solid table structure 37
Elements of the Global ROT (cont.) • 5-digit rule number (cont.) 2-digit sequential rule number It’s also a global standard This means we can create up to 100+ rules for every data field (by using alphabet) The creation of a rule will help other regions to evaluate the applicability of the rule 38
Elements of the Global ROT (cont.) • Error message explanation A short text message that describes the condition being tested in the database (human version of the rule) Will be used in reports to drive action on data maintenance Error Level Description I Info Only W Warning is a condition that could be improved, but is completely functional E Error is a condition that will not have an immediate impact, but will create data accuracy deteriorated in the mid term C Critical indicates this condition has the ability to stop a shipment $ Potential impact to cash flow 39
Elements of the Global ROT (cont.) • Owner Specify the corresponding Business Process Owner of the data These are the only groups with authority to create or approve a given rule System architecture allows the creation of additional groups • Table It’s a technical element passed to the query generator during runtime Narrow the focus of ROT applied to the specified content of that SQL table This is only used by the MDG Engineers. examples of those SQL tables are: 40
Elements of the Global ROT (cont.) • Fields Same as the table element, indicating to the engine which data field the query will be applied to This is only for Master Data Managers • Rules Correspond to the technical SQL Server restricted language statement that will be applied by the engine to the data set Require technical knowledge of the expected syntax needed by the engine 41
Elements of the Global ROT (cont.) • Scope The technical statement that isolates the records to which the rule will be applied It’s also used strictly by the MDG • Status This is a flag that shows if a rule has been deactivated for any reason, avoiding the need for the deletion of rules 42
Elements of the Global ROT (cont.) • Client Identifies the specific SAP environment from which data was obtained (examples: QN4-040, QN4-050, PN1-025) SQVI queries are recreated on those SAP environments used for data cleansing and conversion activities, and can be uploaded to the ROT engine • Region Regions might have slightly different needs for a given topic or variable within SAP This field allows the engine to keep separate rules for every region/client • PBU This pull-down menu allows the user to display non-conforming records for a given business unit 43
What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 44
Enhancing the Business Processes 45
Enhancing the Business Processes (cont.) 3 ROI $$$$$ 2 SAP Effectiveness Metrics 1 ROT (Rule of Thumb) 46
Enhancing the Business Processes with a Purpose • Sigma level as a 1 measure of speed and accuracy • Supporting optimal business performance ROLE: SHIPPING MASTER INVENTORY SUPPLIER ORDER MANAGEMENT Error in Backflush Error in Backflush 3 Outbound Interplant 2 Unexecuted Build (caused by (caused by missing Inventory with Inventory with External Supplier REPORT NAME: Shipments not Supplier Past Plans missing Material Purchasing Master negative quantity negative dollars Past Dues completed Dues Master data) data) Inventory is Inventory is overstated, Missing Inventory is overstated on Raw Customer ASN, and overstated on Raw Increased Raw Material and Understating Cannot execute Cannot execute Missing Customer Material and Understating Material Inventory, understated on inventory, increased Build Plan, Potential Build Plan, Potential Invoices OR understated on inventory, increased Potential for Finished raw material Inbound and Inbound and IMPACT: Customer Finished raw material inventory, Overtime and Goods/Higher inventory, Potential Outbound Premium Outbound Premium Requirements are Goods/Higher Potential for Financial Outbound Premium Assemblies, Labor is for Financial Cash Freight, Potential Freight, Potential understated and Assemblies, Labor is Cash Flow Issue Freight understated, Flow Issue Overtime Overtime Inventory Potential for understated, Production Outbound Premium Production Downtime Downtime Shipments OWNER: GOAL: Shipping Role No open deliveries over 1 Scheduling Role Inv Analyst / PC&L Inv Analyst / PC&L Team Team Number of No orders over 2 No errors over 1 component part Inv Analyst / PC&L Inv Analyst / PC&L Team Number of parts - No parts with Team Dollar value of Supplier Order Management No orders older Supplier Order Management No orders older Optimization weeks old week old numbers per negative inventory than 2 weeks than 2 weeks week old negative inventory assembly Plant FW61 Plant Zacatecas SAP T-CODE: Target: Plant Manager R. Nunez VL06O 0 > 1 WEEK 0 Y_DN3_4700017 2 0 E-Parts > 2 WEEKS 2 ZCOGIA 0 > 1 WEEK 0 MF47 0 > 1 WEEK 358 MB52 0 REAL-TIME 48 MB52 0 REAL-TIME ($646) Y_DN3_4700037 Y_DN3_4700037 8 0 > 2 WEEKS 24 8 0 > 2 WEEKS 10 $ XXX.5 M FW62 Fresnillo 1 A. Lozano 5 33 0 1432 176 ($835,130) 44 37 By Dec 08 FW63 Fresnillo 2 J. Moreno 1 4 0 820 80 ($42,603) 63 83 FW80 Laredo Carlos Leyva / Gene Lindgren 13 0 0 0 6 ($663,931) 726 11 FW81 Neuvo Laerdo R. Vega 0 0 0 515 423 ($454,177) 41 41 FW84 Guadalupe 2 F. Olivas 0 17 0 152 25 ($7,774) 20 7 FW86 Linares R. Mendoza 0 0 0 124 13 ($1,046) 23 2 FW91 Victoria 1 R. Gutierrez 0 0 0 1121 82 ($49,836) 7 0 FW92 Victoria 2 R. Gutierrez 0 0 0 3344 112 ($115,766) 91 5 FW96 Guadalupe 3 J. Navarro 0 8 0 347 186 ($117,771) 0 0 47
Monitoring Business Performance • Access the tool • Use the specific intranet site where the reporting service is located • Select the dataset of interest • Check for the information quality level • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 48
Monitoring Business Performance (cont.) • Browse the detailed results • Act on the exceptions • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 5a. Clean the data 5b. Enhance/fix the business process (see next slide) 49
Monitoring Business Performance (cont.) 5a. Clean the data 5b. Enhance/fix the business process • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 50
Enhancing the Business Processes with a Purpose Delphi Communications with Supplier Customer Communications with Delphi DATA DATA Expertise Roles Shipping Receiving Supplier Order Master Planning and Customer Order Management Scheduling Management DATA DATA Supplier Communication with Delphi Delphi Communication with Supplier DATA DATA 51
Enhancing the Business Processes with a Purpose (cont.) P4 Repetitive Pulls Transformation to Be Customer-Centric PD ERP Driven Pulls 52
Enhancing the Business Processes with a Purpose (cont.) DOH Index for FW62 17.3% Party Numbers with Excessive DOH INV: 72.0% 5/28/2008 Part numbers with potential Premium 10.7% 2,500 INVENTORY (by SLOC) Pieces Dollars 1,932 1,600 Blanks: In Transit 3,749,388 324,373 2,000 0001: Receiving 80,334,594 3,813,069 0002: WIP 57,566,936 2,158,153 1,500 0003: to LADC 20,134 298,597 1,000 0004: at LADC 65,797 1,700,415 1,400 1,355 0007: Others 2,014,332 88,664 360 500 287 0009: Finished 24,528 185,885 104 Total 140,026,321 $8,244,783 0 Red Yellow Green EXCESS INV PNs w/DaysOnHand ($$$) 1,200 16.8% 17.7% 19.6% INVENTORY ANALYSIS by Status Flag COMPONENTS CABL HARN TOTAL % LwrLimit UpperLim 1 Red 198 66 23 287 10.70% -999 0 2 Yellow 224 105 31 360 13.42% 0.1 5 1,000 3 Green 59 30 15 104 3.88% 5.1 7 4 EXCESS INV ($$$) 1,206 560 166 1,932 72.01% 7.1 999 Total Part numbers 1,687 761 235 2,683 5 PNs with -999 DOH 8 1 2 11 0.41% 800 6 PNs with over 100 neg(DOH) 0 0 0 0 0.00% 7 PNs with over 30 neg(DOH) 3 0 3 6 0.22% 8 PNs with less than 0 DOH 180 59 16 255 9.50% 9 PNs with 0 DOH 7 6 2 15 0.56% 10 PNs with DOH less than 5 222 104 30 356 13.27% 600 11 PNs with DOH less than 10 114 63 22 199 7.42% 12 PNs with DOH less than 30 190 118 21 329 12.26% 13 PNs with DOH less than 999 101 43 13 157 5.85% 14 PNs with 999 DOH 862 367 126 1,355 50.50% 400 356 15 Avg neg(INV_DOH) excl -999 DOH -4.0 -2.8 -12.8 329 16 Avg INV_DOH excluding 999 DOH 15.8 15.1 15.8 17 Generic MRP Controllers (no owner) 101 20 206 327 12.19% 255 18 MRP Type = PD 1,686 761 1 2,448 91.24% 199 19 MRP Type = P4 1 0 234 235 8.76% 200 157 20 MRP Type = ND 0 0 0 0 0.00% Exceptions Groups COMPONENTS CABL HARN TOTAL % 11 0 6 15 1 Late in moving to a proposal 0 0 0 0 0.00% 0 2 Late in moving to a commitment 0 0 0 0 0.00% PNs with over 100 neg(DOH) PNs with over 30 neg(DOH) PNs with -999 DOH PNs with less than 0 DOH PNs with 0 DOH PNs with 999 DOH PNs with DOH less than 5 PNs with DOH less than 10 PNs with DOH less than 30 PNs with DOH less than 999 3 Stock should have been there 252 119 1 372 13.87% 4 A new requirement 0 0 0 0 0.00% 5 BOM related issues 0 0 0 0 0.00% 6 Too much or too little stock 8 1 2 11 0.41% -200 7 Dates when needed/available differs 437 89 120 646 24.08% 8 Marked for Deletion? 0 0 0 0 0.00% Total Part numbers 697 209 123 1,029 38.35% 53
Enhancing the Business Processes with a Purpose (cont.) ACTIVE PARTS ONLY DOH Index for FW62 78.3% Party Numbers with Excessive DOH INV: 18.6% Date: 09/23/2008 Part numbers with potential Premium 3.2% 6,000 5,381 INVENTORY for ACTIVE PARTS Pieces Dollars 5,000 6,000 COMPONENTS 46,759,315 3,857,620 4,000 CABLE 11,334,460 703,710 3,000 HARNESS 51,951 1,277,758 5,381 2,000 1,372 1,000 236 405 0 5,000 Total 58,145,726 $5,839,088 Red (pot Yellow (risk for Green (Opt EXCESS INV shortage) shortage) Days Supply) ($$$) TOTAL INV IN EXCESS VALUE 3,925,414 INVENTORY IN EXCESS VALUE 3,012,811 404,025 508,579 PNs w/DaysOnHand % Optimal PN by groups -> 81.8% 52.6% 83.8% DaysSupply Analisys by Commodity) COMPONENTS CABL HARN TOTAL % LwrLimit UpperLim 1 Red (pot shortage) 129 71 36 236 3.19% -999 0 4,000 2 Yellow (risk for shortage) 229 93 83 405 5.48% 0.1 5 3 Green (Opt Days Supply) 3,560 441 1,380 5,381 72.78% 5.1 7 4 EXCESS INV ($$$) 716 410 246 1,372 18.56% 7.1 999 Total Part numbers 4,634 1,015 1,745 7,394 5 PNs with -999 DOH 0 0 0 0 0.00% 6 PNs with over 100 neg(DOH) 1 0 1 2 0.03% 3,000 7 PNs with over 30 neg(DOH) 12 9 3 24 0.32% 8 PNs with less than 0 DOH 92 58 25 175 2.37% 9 PNs with 0 DOH 229 93 83 405 5.48% 10 PNs with DOH less than 5 3,560 441 1,380 5,381 72.78% 11 PNs with DOH less than 10 171 75 40 286 3.87% 12 PNs with DOH less than 30 144 65 14 223 3.02% 2,000 13 PNs with DOH less than 999 94 66 20 180 2.43% 14 PNs with 999 DOH 292 200 159 651 8.80% 15 Avg neg(INV_DOH) excl -999 DOH -27.9 -12.7 -157.4 16 Avg INV_DOH excluding 999 DOH 865.5 575.3 936.9 17 Generic MRP Controllers (no owner) 2,919 253 812 3,984 53.88% 18 MRP Type = PD (SAP generated) 4,457 1,015 1,744 7,216 97.59% 1,000 19 MRP Type = P4 (user sched some) 115 0 1 116 1.57% 20 MRP Type = ND (NO SAP MRP) 25 0 0 25 0.34% 651 21 Rounding Values undefined 2,686 137 1,584 4,407 59.60% 405 286 175 223 180 Exceptions Groups COMPONENTS CABL HARN TOTAL % 0 2 24 1 Late in moving to a proposal 0 0 0 0 0.00% 0 PNs with over 100 neg(DOH) PNs with over 30 neg(DOH) PNs with -999 DOH PNs with 0 DOH PNs with 999 DOH PNs with less than 0 DOH PNs with DOH less than 5 PNs with DOH less than 10 PNs with DOH less than 30 PNs with DOH less than 999 2 Late in moving to a commitment 0 0 2 2 0.03% 3 Stock should have been there 389 107 57 553 7.48% 4 A new requirement 0 0 0 0 0.00% 5 BOM related issues 0 0 0 0 0.00% 6 Too much or too little stock 0 0 4 4 0.05% 7 Dates when needed/available differs 466 130 37 633 8.56% -1,000 8 Marked for Deletion? 0 0 0 0 0.00% Total Part numbers with Exceptions 855 237 100 1,192 16.12% 54
What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 55
Resources • www.sap-img.com SAP Tables Help File and ABAP Programming • www.dmreview.com/channels/data_quality.html White paper library • www.findwhitepapers.com/index.php Technology Research For Business Professionals • www.ittoolbox.com/ Professional IT Community 56
7 Key Points to Take Home • Focus on data fields of interest (remember the SAP built-in validation for data during the creation process) • Keep current and future SAP functionality in mind during development • Identify proper SAP tables required for the Information Quality Model • Create simple queries (minimize more than two joins per query) • Use structured names for SQL tables and programs • Share system ownership with functional areas by co-authoring rules and resolving their issues • Maintain high coloring standards for Information Quality and Business Performance Assessment (Red/Yellow/Green) 57
Your Turn! How to contact me: Jose V Zavala jose.v.zavala@delphi.com 58
Disclaimer SAP, R/3, mySAP, mySAP.com, xApps, xApp, SAP NetWeaver®, Duet™, PartnerEdge, 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. Wellesley Information Services is neither owned nor controlled by SAP. 59

More Related Content

PPT
Supply chain presentation 11 2006
Girard Brewer
 
PPTX
TeleManagement Forum OSSera Case Study - AIS Thailand Service Manager Present...
Mingxia Zhang, Ph.D.
 
PDF
Emakina Academy 6 - Boost your intranet - Web Content Management for SAP
Emakina
 
PDF
Emakina Academy 6 - Boost your intranet - STIB
Emakina
 
PDF
VW EMS case March 2010
guest49c269
 
PDF
Cloud Computing -- Organizational Shift
Raman Kannan
 
PDF
Enterprise Architecture
Raman Kannan
 
PPTX
What is @hand??
cleehand
 
Supply chain presentation 11 2006
Girard Brewer
 
TeleManagement Forum OSSera Case Study - AIS Thailand Service Manager Present...
Mingxia Zhang, Ph.D.
 
Emakina Academy 6 - Boost your intranet - Web Content Management for SAP
Emakina
 
Emakina Academy 6 - Boost your intranet - STIB
Emakina
 
VW EMS case March 2010
guest49c269
 
Cloud Computing -- Organizational Shift
Raman Kannan
 
Enterprise Architecture
Raman Kannan
 
What is @hand??
cleehand
 

What's hot (18)

PPTX
Make Your Business More Flexible with Scalable Business Process Management So...
Perficient, Inc.
 
PPTX
Internship Experience
Amit Chaudhari
 
PDF
Dell 20805[1]
mevada
 
KEY
Designing the User Experience
Jason Wehmhoener
 
PPTX
Værdikæder i netværk og plug'n play supply chains af John Johansen, CIP på AAU
InfinIT - Innovationsnetværket for it
 
PDF
SQL Server Data Mining - Taking your Application Design to the Next Level
Mark Ginnebaugh
 
PPT
Henrik thomsen
InfinIT - Innovationsnetværket for it
 
PDF
Microsoft Data Mining 2012
Mark Ginnebaugh
 
PDF
Implementing a request fulfillment process
IESS
 
PDF
WIKIOCEAN
mahesh.panchal
 
PDF
Neecom 2010 (Inovis-Dell case study)
Doug Kern
 
PDF
Corporate Overview
Bobby Brea
 
PPTX
Newgen Solutions for Telecom
newgenpartners
 
PDF
Ccs onesheet datasource_en
CNET Content Solutions
 
PDF
Managed Solutions Professional Services Presentation
shawnrobertmartin
 
PPT
All Roads Lead to SaaS
Ruann Tiller
 
PDF
Scalability problems and E-business
We Learn - A Continuous Learning Forum from Welingkar's Distance Learning Program.
 
PDF
B2B eCommerce
We Learn - A Continuous Learning Forum from Welingkar's Distance Learning Program.
 
Make Your Business More Flexible with Scalable Business Process Management So...
Perficient, Inc.
 
Internship Experience
Amit Chaudhari
 
Dell 20805[1]
mevada
 
Designing the User Experience
Jason Wehmhoener
 
Værdikæder i netværk og plug'n play supply chains af John Johansen, CIP på AAU
InfinIT - Innovationsnetværket for it
 
SQL Server Data Mining - Taking your Application Design to the Next Level
Mark Ginnebaugh
 
Henrik thomsen
InfinIT - Innovationsnetværket for it
 
Microsoft Data Mining 2012
Mark Ginnebaugh
 
Implementing a request fulfillment process
IESS
 
WIKIOCEAN
mahesh.panchal
 
Neecom 2010 (Inovis-Dell case study)
Doug Kern
 
Corporate Overview
Bobby Brea
 
Newgen Solutions for Telecom
newgenpartners
 
Ccs onesheet datasource_en
CNET Content Solutions
 
Managed Solutions Professional Services Presentation
shawnrobertmartin
 
All Roads Lead to SaaS
Ruann Tiller
 
Scalability problems and E-business
We Learn - A Continuous Learning Forum from Welingkar's Distance Learning Program.
 
B2B eCommerce
We Learn - A Continuous Learning Forum from Welingkar's Distance Learning Program.
 
Ad

Viewers also liked (6)

PDF
Creating Innovation in Schools
Rafael Parente
 
PPTX
Innovation, Change and Technology
Stephen Murgatroyd, PhD FBPsS FRSA
 
PPTX
Leading for innovation in schools workshop 2 singapore
Stephen Murgatroyd, PhD FBPsS FRSA
 
PPT
Successful Remediation of the Unsatisfactory Teacher
Richard Voltz
 
PPT
Understand Innovation in 5 Minutes
Gordon Graham
 
PPTX
Slideshare ppt
Mandy Suzanne
 
Creating Innovation in Schools
Rafael Parente
 
Innovation, Change and Technology
Stephen Murgatroyd, PhD FBPsS FRSA
 
Leading for innovation in schools workshop 2 singapore
Stephen Murgatroyd, PhD FBPsS FRSA
 
Successful Remediation of the Unsatisfactory Teacher
Richard Voltz
 
Understand Innovation in 5 Minutes
Gordon Graham
 
Slideshare ppt
Mandy Suzanne
 
Ad

Similar to Measure Data Quality (20)

PDF
SAP EIM
Sybase Türkiye
 
PDF
SAP HANA and Apache Hadoop for Big Data Management (SF Scalable Systems Meetup)
Will Gardella
 
PDF
Automotive products selection software - Right Information
Right Information
 
PDF
Greenplum hadoop
Chiou-Nan Chen
 
PDF
Greenplum hadoop
Chiou-Nan Chen
 
PDF
Saleseffectivity and business intelligence
marekdan
 
PDF
High-precision engineering products selection software- - Right Information
Right Information
 
PPTX
Open Source@etailing v1.2 (video)
Iksula
 
PPTX
Presentatie Duncan Rogers NMD2010 17 juni 2010
OGZ
 
PDF
Sapphire Online 2009 Or1005
Shereen Zubair
 
PDF
Unraveling Data Sharing Challenges to Improve Collaboration atGlobal Ford Pow...
safisam
 
PDF
Tooling systems selection software - Right Information
Right Information
 
PDF
Intel Cloud summit: Big Data by Nick Knupffer
IntelAPAC
 
PDF
Electronic equipment selection software - Right Information
Right Information
 
PDF
Power tools selection software - Right Information
Right Information
 
PPSX
Linked In_Thoughts
TimFeltham
 
PDF
Scaling MySQL: Benefits of Automatic Data Distribution
ScaleBase
 
PDF
Product Customization, Personalization and Customer Centricity: Market Opport...
Assomac
 
PDF
Big Data at #WADAY11
Cosimo Accoto
 
PPTX
UBM Electronics... By the Numbers
bmcclean
 
SAP EIM
Sybase Türkiye
 
SAP HANA and Apache Hadoop for Big Data Management (SF Scalable Systems Meetup)
Will Gardella
 
Automotive products selection software - Right Information
Right Information
 
Greenplum hadoop
Chiou-Nan Chen
 
Greenplum hadoop
Chiou-Nan Chen
 
Saleseffectivity and business intelligence
marekdan
 
High-precision engineering products selection software- - Right Information
Right Information
 
Open Source@etailing v1.2 (video)
Iksula
 
Presentatie Duncan Rogers NMD2010 17 juni 2010
OGZ
 
Sapphire Online 2009 Or1005
Shereen Zubair
 
Unraveling Data Sharing Challenges to Improve Collaboration atGlobal Ford Pow...
safisam
 
Tooling systems selection software - Right Information
Right Information
 
Intel Cloud summit: Big Data by Nick Knupffer
IntelAPAC
 
Electronic equipment selection software - Right Information
Right Information
 
Power tools selection software - Right Information
Right Information
 
Linked In_Thoughts
TimFeltham
 
Scaling MySQL: Benefits of Automatic Data Distribution
ScaleBase
 
Product Customization, Personalization and Customer Centricity: Market Opport...
Assomac
 
Big Data at #WADAY11
Cosimo Accoto
 
UBM Electronics... By the Numbers
bmcclean
 

Measure Data Quality

  • 2. Case Study: Utilize Quantitative Standards and Metrics to Measure Data Quality Initiatives: A Real- World Case Study from Delphi Jose Zavala Delphi © 2008 Wellesley Information Services. All rights reserved.
  • 3. In This Session ... • Reveal the magnitude and complexity of data required to satisfy Generation C (consumers) demand Build a map with these interacting business elements Quantify effects of these elements to the auto industry • Find the value in creating a strong, yet simple data strategy (simplify) • Review a model of a dynamic query creator (rule of thumb) to easily create/expand logical tests applied to your static dataset (Material Master Records) Build from this model into other dynamic data elements (effectiveness metrics) • Extend this approach with closed-loop cycles monitoring the bottom line (cash flow/inventories) 2
  • 4. What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 3
  • 5. Generation C (Consumers): Aggressive Demand Creative Communicated Connected Conversation Challenged Community Charming Customized e-Commerce Content Cash Control Channel Consumer 2.0 4
  • 6. Consumer Touch Points and the Information Flows FACT: Order-To-Cash Research Blogs Music Peoples’ needs and Movies Social Price? desires are easily captured with current technology Gaming Shopping Products are Quality engineered, Speed manufactured, and delivered fast! Mass-Customization 5
  • 7. Example: Buying a New Car We want this ... DATA s BUSINESS r ate e ne ELEMENTS a tg Th But OEMs offer all these choices Speed Color Technology Size Class DVD Products are Multiple CD Service Finance Headlights engineered, Brand Class Price Seats Value manufactured, and Transmission Economy Sound delivered faster! Safe Sunroof Engine Reliable Sharp … When supported Power Doors Resistant with good data Shape 6
  • 8. Magnitude and Complexity of Data Required Currencies and Markets Modular Products Delphi Divisions Electronics & Safety Packard Electrical/Electronic Architecture Powertrain Systems Steering Thermal Systems Product & Service Solutions Laws and Regulations US SEC EU Canada Lead Free 7
  • 9. Magnitude and Complexity of Data Required: Material Masters EU Material Master Recs: 237,000+ 1.99 M Delphi Packard 84 fields NA Material AP Material Master Recs: Master Recs: 165,000+ 237,000+ 116 fields 87 fields 1.91 M 2.06 M Combined Material Masters/Fields = 5.96 million 8
  • 10. The Power to Simplify: Material Masters Quality Aspect • Each intersection (data element) could be: Missing or late Delphi Packard Inaccurate ATTRIBUTES (fields) MATERIALS Sales Engineering Purchasing Finance Prod Control Logistics 5.96 Million Material Masters/Fields = Error Opportunities 9
  • 11. What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 10
  • 12. SAP Data Architecture • So, how do you check for a data element that is: Missing or late? Inaccurate? • If data is created/maintained by Sales, Engineering, Purchasing, Finance, Product Control, or Logistics … … Then how do you measure its quality? 11
  • 13. SAP Data Architecture (cont.) • SAP has a well structured set of inter-related tables to minimize size of storage as well as to improve response time • Realizing that we are building a data quality index and because size of data files are not a restriction, Sales VBEH VBLB VBSS we can proceed to “fill in the blanks” and create a VBEP VBKD VBFA VBRK data matrix with key data elements VBKE VBPA VBAP VBUK VBBE VBRP T16OQ TLGR T001L Unfold! SO11 T161T MSEG System Purch T069 T437L EKET EORD T160R T024D SO12 MVER MARA EKPO MARA MARC T161F T134T EKNN MKPF T160W T604 T157H EINE EKKO SO31 T006 T005 ADRP STKO PLKO MAKT CRCO CRTX KAPE BVOR BSIP BSAS System Eng Plan FICO T100 T024 PLPO AFPO CRHS KAKT PAYR BSAK T247 ADR2 STPO MVER CRHD KAZY BSAD KNC1 MARA MAST CRCA MLAN TAPLT AOQD AFKO MARM CRID KAPA BKPF BSIS T777A T023 ADCP MAPL PLSO T001W CRHH KAKO T024C BSID BSIK LFC1 12
  • 14. SAP Data Architecture and the QuickViewer SQVI Tool • Off-the-shelf SAP contains over 100 data fields as part of master data records in multiple views • First, identify data fields as part of the master data record of interest • Then, define ownership for data creation and maintenance 13
  • 15. Four Steps to Extract Data for the Information Quality Index 1 3 4 2 1: Join Definition 2: Field Selection 4: Validate Results 3: Save/Test 14
  • 16. Step 1: Join Definition • Using QuickViewer — SQVI Keep table joins simple, as this will drive your processing time Field selection should consider current and future functionality 15
  • 17. Step 2: Field Selection • Select data fields of interest — SQVI • Data will be generated in the same order • Selection fields are part of interface screen created, if run with transaction START_REPORT 16
  • 18. Step 2: Field Selection (cont.) • Create All Queries According to Areas of Interest — SQVI Use consistent query names according to the nature of the project Do not bring unnecessary data fields to the model 17
  • 19. Step 3: Save/Test • Generate programs and get report names — SQVI Test queries using transaction START_REPORT These queries can also be used to validate data 18
  • 20. Step 4: Validate Results • Schedule a download job (t-code SM37) Add all queries to the download job as steps Consider execution times to avoid system overloads 19
  • 21. Step 4: Validate Results (cont.) • Schedule a download job (t-code SM37) Daily analysis seems to be a good choice Data needs to be fixed, but most important is to enhance the business processes as well 20
  • 22. Step 4: Validate Results (cont.) • Get to your spool list and export items as text (t-code SP02) Queries over empty data tables result in no spool output Download to user SAPGUI folder for conversion and upload to SQL Once files are downloaded to your local drive, user should get an SAP notification 21
  • 23. Step 4: Validate Results (cont.) • Find your items in the SAPWorkDir folder using Windows Explorer Make sure the file size is manageable Downloaded jobs can be directed to other users when scheduled 22
  • 24. Step 4: Validate Results (cont.) • Complete the validation process (text editor) This is a standard output when the spool item is “Export as Text” Use the tool of your choice to upload to SQL 23
  • 25. Step 4: Validate Results (cont.) • Upload the files to the SQL Server (MS-SQL) Data fields should be uploaded in the same sequence There should be one table for each query created 24
  • 26. A Self-Sufficient Data Analysis System Algorithm Truncate existing data Build New SQL Microsoft START Statement/Query Internet MM tables + Results Explorer Reload MM tables Select Table called Browsing from SAPGUI in Audit Rules Exceptions Tag each record Report with Client, Region (AuditResults) Business Unit & Plant Open Apply Audit Rules Audit Rules Table Scope (filter records) SP02 SAP SM37 SM36 Initiate variables Row = 0 Apply SQL Command under Rule to Field SQVI Row = Row + 1 Segregate non-complaint Read AuditRule Row # data to AuditResults END Yes End of file No No End of Target Audit Rules? Table found? Yes 25
  • 27. Daily Refresh of Data Loaded to SQL Engine • Preparing SAP download jobs Once target tables and data fields are identified, jobs are scheduled to run at 1:00 AM EST Monday through Friday A Master Data Engineer (MDE) gets them into their SAP account • Retrieving data from SAP to bring to a local system Files are then downloaded as text to a local PC Information is not structured at this time • Uploading data to the SQL Server from a local system Files are uploaded directly to the SQL Server from the production environment 26
  • 28. Rules of Thumb (ROT) Examples 27
  • 29. Additional ROT System Tables (Part of the SQL Model) • MMPlantPBU table Helps classify each record by Plant and Business Unit Plant (key), Business Unit, Plant Description, Master Data Manager (coordinator) • MMAuditSummaryHeader Keeps daily audit results Client, Region, Business Unit, Audit Fields, Total Records, RunDate & Plant • MMAuditSummaryItem Provides count and links for non-conforming records by rule RuleNumber, ErrorMessageExplanation, ErrorLevel, Owner, Client, Region, Business Unit, Errors, RunDate, Plant 28
  • 30. What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 29
  • 31. Rules of Thumb: Browsing the AuditResults Records Report Name Selecting a Target Dataset MMs records audited Total fields audited Navigation MM Recs X Fields Zoom Total exceptions found Search PPM calculation % deviations Date/time % compliance to ROT stamp Info Quality Sigma Level Name of Errors per BPO area Business Process Owners (BPOs) and Total Rules created by them Exporting Formats Continuous Improvement Model 30
  • 32. Rules of Thumb: Browsing the AuditResults Records (cont.) • Non-conforming data to ROT are presented by business unit (PBU) or plant level, following a standard set of information 31
  • 33. Data Views Available — Following the Rules of Thumb • When users are browsing non-conforming records, they can target a given client, region, and business unit data set • Specific Rules of Thumb (logical conditions) are established or approved by the business process owners part of a business unit Then they are put together in SQL Server language syntax by Master Data Engineers • The list of non-conforming records looks like this: 32
  • 34. Multiple Formats Available When Exporting Records • The SQL Server Reporting Service contains a set of standard formats End users can manipulate data after non-conforming records are identified 33
  • 35. Benefits of the Reporting Services • Quick access to information Find any existing value By rule By plant Etc. • Exporting formats available Most common formats are available Helps processing errors Facilitate Error Analysis such as: Counts, average, etc. 34
  • 36. Benefits of the Reporting Services (cont.) • Data is available for massive updates (t-code MM17) Target Material Masters are copied to the clipboard and provided to MM17 MM17 can change up to 800+ records at one time Tables and fields are identified Update is done in just a few steps Processing time is minimized 35
  • 37. Describing the Global ROT System • Benchmarking is made possible by networking among Master Data managers Within each region Within each business unit At different levels of deployment phase (QN4 environment) Comparing: Error Level Logical statement Scope Applicability Customized values RefPackingMaterial <> “REFPACK” PN1 AP DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active PN1 AP DEEDS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active PN1 NA DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active QN4 NA DCS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active QN4 NA DEEDS 005-00 E [RefPackingMaterial] != 'REFPACK' [MaterialType] = 'HALB' AND [ProcType] ='E' AND [BUOM]='PC' Active 36
  • 38. Elements of the Global ROT • 5-digit rule number 3-digit data field number A sequential, numerical ID that goes along with a given SAP data field It’s standard for all PBUs in every region Currently have 148 fields available for creation of rules, and only 103 are partially covered Can grow as data becomes available within SAP, in a solid table structure 37
  • 39. Elements of the Global ROT (cont.) • 5-digit rule number (cont.) 2-digit sequential rule number It’s also a global standard This means we can create up to 100+ rules for every data field (by using alphabet) The creation of a rule will help other regions to evaluate the applicability of the rule 38
  • 40. Elements of the Global ROT (cont.) • Error message explanation A short text message that describes the condition being tested in the database (human version of the rule) Will be used in reports to drive action on data maintenance Error Level Description I Info Only W Warning is a condition that could be improved, but is completely functional E Error is a condition that will not have an immediate impact, but will create data accuracy deteriorated in the mid term C Critical indicates this condition has the ability to stop a shipment $ Potential impact to cash flow 39
  • 41. Elements of the Global ROT (cont.) • Owner Specify the corresponding Business Process Owner of the data These are the only groups with authority to create or approve a given rule System architecture allows the creation of additional groups • Table It’s a technical element passed to the query generator during runtime Narrow the focus of ROT applied to the specified content of that SQL table This is only used by the MDG Engineers. examples of those SQL tables are: 40
  • 42. Elements of the Global ROT (cont.) • Fields Same as the table element, indicating to the engine which data field the query will be applied to This is only for Master Data Managers • Rules Correspond to the technical SQL Server restricted language statement that will be applied by the engine to the data set Require technical knowledge of the expected syntax needed by the engine 41
  • 43. Elements of the Global ROT (cont.) • Scope The technical statement that isolates the records to which the rule will be applied It’s also used strictly by the MDG • Status This is a flag that shows if a rule has been deactivated for any reason, avoiding the need for the deletion of rules 42
  • 44. Elements of the Global ROT (cont.) • Client Identifies the specific SAP environment from which data was obtained (examples: QN4-040, QN4-050, PN1-025) SQVI queries are recreated on those SAP environments used for data cleansing and conversion activities, and can be uploaded to the ROT engine • Region Regions might have slightly different needs for a given topic or variable within SAP This field allows the engine to keep separate rules for every region/client • PBU This pull-down menu allows the user to display non-conforming records for a given business unit 43
  • 45. What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 44
  • 46. Enhancing the Business Processes 45
  • 47. Enhancing the Business Processes (cont.) 3 ROI $$$$$ 2 SAP Effectiveness Metrics 1 ROT (Rule of Thumb) 46
  • 48. Enhancing the Business Processes with a Purpose • Sigma level as a 1 measure of speed and accuracy • Supporting optimal business performance ROLE: SHIPPING MASTER INVENTORY SUPPLIER ORDER MANAGEMENT Error in Backflush Error in Backflush 3 Outbound Interplant 2 Unexecuted Build (caused by (caused by missing Inventory with Inventory with External Supplier REPORT NAME: Shipments not Supplier Past Plans missing Material Purchasing Master negative quantity negative dollars Past Dues completed Dues Master data) data) Inventory is Inventory is overstated, Missing Inventory is overstated on Raw Customer ASN, and overstated on Raw Increased Raw Material and Understating Cannot execute Cannot execute Missing Customer Material and Understating Material Inventory, understated on inventory, increased Build Plan, Potential Build Plan, Potential Invoices OR understated on inventory, increased Potential for Finished raw material Inbound and Inbound and IMPACT: Customer Finished raw material inventory, Overtime and Goods/Higher inventory, Potential Outbound Premium Outbound Premium Requirements are Goods/Higher Potential for Financial Outbound Premium Assemblies, Labor is for Financial Cash Freight, Potential Freight, Potential understated and Assemblies, Labor is Cash Flow Issue Freight understated, Flow Issue Overtime Overtime Inventory Potential for understated, Production Outbound Premium Production Downtime Downtime Shipments OWNER: GOAL: Shipping Role No open deliveries over 1 Scheduling Role Inv Analyst / PC&L Inv Analyst / PC&L Team Team Number of No orders over 2 No errors over 1 component part Inv Analyst / PC&L Inv Analyst / PC&L Team Number of parts - No parts with Team Dollar value of Supplier Order Management No orders older Supplier Order Management No orders older Optimization weeks old week old numbers per negative inventory than 2 weeks than 2 weeks week old negative inventory assembly Plant FW61 Plant Zacatecas SAP T-CODE: Target: Plant Manager R. Nunez VL06O 0 > 1 WEEK 0 Y_DN3_4700017 2 0 E-Parts > 2 WEEKS 2 ZCOGIA 0 > 1 WEEK 0 MF47 0 > 1 WEEK 358 MB52 0 REAL-TIME 48 MB52 0 REAL-TIME ($646) Y_DN3_4700037 Y_DN3_4700037 8 0 > 2 WEEKS 24 8 0 > 2 WEEKS 10 $ XXX.5 M FW62 Fresnillo 1 A. Lozano 5 33 0 1432 176 ($835,130) 44 37 By Dec 08 FW63 Fresnillo 2 J. Moreno 1 4 0 820 80 ($42,603) 63 83 FW80 Laredo Carlos Leyva / Gene Lindgren 13 0 0 0 6 ($663,931) 726 11 FW81 Neuvo Laerdo R. Vega 0 0 0 515 423 ($454,177) 41 41 FW84 Guadalupe 2 F. Olivas 0 17 0 152 25 ($7,774) 20 7 FW86 Linares R. Mendoza 0 0 0 124 13 ($1,046) 23 2 FW91 Victoria 1 R. Gutierrez 0 0 0 1121 82 ($49,836) 7 0 FW92 Victoria 2 R. Gutierrez 0 0 0 3344 112 ($115,766) 91 5 FW96 Guadalupe 3 J. Navarro 0 8 0 347 186 ($117,771) 0 0 47
  • 49. Monitoring Business Performance • Access the tool • Use the specific intranet site where the reporting service is located • Select the dataset of interest • Check for the information quality level • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 48
  • 50. Monitoring Business Performance (cont.) • Browse the detailed results • Act on the exceptions • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 5a. Clean the data 5b. Enhance/fix the business process (see next slide) 49
  • 51. Monitoring Business Performance (cont.) 5a. Clean the data 5b. Enhance/fix the business process • Sigma level as a measure of speed and accuracy • Supporting optimal business performance 50
  • 52. Enhancing the Business Processes with a Purpose Delphi Communications with Supplier Customer Communications with Delphi DATA DATA Expertise Roles Shipping Receiving Supplier Order Master Planning and Customer Order Management Scheduling Management DATA DATA Supplier Communication with Delphi Delphi Communication with Supplier DATA DATA 51
  • 53. Enhancing the Business Processes with a Purpose (cont.) P4 Repetitive Pulls Transformation to Be Customer-Centric PD ERP Driven Pulls 52
  • 54. Enhancing the Business Processes with a Purpose (cont.) DOH Index for FW62 17.3% Party Numbers with Excessive DOH INV: 72.0% 5/28/2008 Part numbers with potential Premium 10.7% 2,500 INVENTORY (by SLOC) Pieces Dollars 1,932 1,600 Blanks: In Transit 3,749,388 324,373 2,000 0001: Receiving 80,334,594 3,813,069 0002: WIP 57,566,936 2,158,153 1,500 0003: to LADC 20,134 298,597 1,000 0004: at LADC 65,797 1,700,415 1,400 1,355 0007: Others 2,014,332 88,664 360 500 287 0009: Finished 24,528 185,885 104 Total 140,026,321 $8,244,783 0 Red Yellow Green EXCESS INV PNs w/DaysOnHand ($$$) 1,200 16.8% 17.7% 19.6% INVENTORY ANALYSIS by Status Flag COMPONENTS CABL HARN TOTAL % LwrLimit UpperLim 1 Red 198 66 23 287 10.70% -999 0 2 Yellow 224 105 31 360 13.42% 0.1 5 1,000 3 Green 59 30 15 104 3.88% 5.1 7 4 EXCESS INV ($$$) 1,206 560 166 1,932 72.01% 7.1 999 Total Part numbers 1,687 761 235 2,683 5 PNs with -999 DOH 8 1 2 11 0.41% 800 6 PNs with over 100 neg(DOH) 0 0 0 0 0.00% 7 PNs with over 30 neg(DOH) 3 0 3 6 0.22% 8 PNs with less than 0 DOH 180 59 16 255 9.50% 9 PNs with 0 DOH 7 6 2 15 0.56% 10 PNs with DOH less than 5 222 104 30 356 13.27% 600 11 PNs with DOH less than 10 114 63 22 199 7.42% 12 PNs with DOH less than 30 190 118 21 329 12.26% 13 PNs with DOH less than 999 101 43 13 157 5.85% 14 PNs with 999 DOH 862 367 126 1,355 50.50% 400 356 15 Avg neg(INV_DOH) excl -999 DOH -4.0 -2.8 -12.8 329 16 Avg INV_DOH excluding 999 DOH 15.8 15.1 15.8 17 Generic MRP Controllers (no owner) 101 20 206 327 12.19% 255 18 MRP Type = PD 1,686 761 1 2,448 91.24% 199 19 MRP Type = P4 1 0 234 235 8.76% 200 157 20 MRP Type = ND 0 0 0 0 0.00% Exceptions Groups COMPONENTS CABL HARN TOTAL % 11 0 6 15 1 Late in moving to a proposal 0 0 0 0 0.00% 0 2 Late in moving to a commitment 0 0 0 0 0.00% PNs with over 100 neg(DOH) PNs with over 30 neg(DOH) PNs with -999 DOH PNs with less than 0 DOH PNs with 0 DOH PNs with 999 DOH PNs with DOH less than 5 PNs with DOH less than 10 PNs with DOH less than 30 PNs with DOH less than 999 3 Stock should have been there 252 119 1 372 13.87% 4 A new requirement 0 0 0 0 0.00% 5 BOM related issues 0 0 0 0 0.00% 6 Too much or too little stock 8 1 2 11 0.41% -200 7 Dates when needed/available differs 437 89 120 646 24.08% 8 Marked for Deletion? 0 0 0 0 0.00% Total Part numbers 697 209 123 1,029 38.35% 53
  • 55. Enhancing the Business Processes with a Purpose (cont.) ACTIVE PARTS ONLY DOH Index for FW62 78.3% Party Numbers with Excessive DOH INV: 18.6% Date: 09/23/2008 Part numbers with potential Premium 3.2% 6,000 5,381 INVENTORY for ACTIVE PARTS Pieces Dollars 5,000 6,000 COMPONENTS 46,759,315 3,857,620 4,000 CABLE 11,334,460 703,710 3,000 HARNESS 51,951 1,277,758 5,381 2,000 1,372 1,000 236 405 0 5,000 Total 58,145,726 $5,839,088 Red (pot Yellow (risk for Green (Opt EXCESS INV shortage) shortage) Days Supply) ($$$) TOTAL INV IN EXCESS VALUE 3,925,414 INVENTORY IN EXCESS VALUE 3,012,811 404,025 508,579 PNs w/DaysOnHand % Optimal PN by groups -> 81.8% 52.6% 83.8% DaysSupply Analisys by Commodity) COMPONENTS CABL HARN TOTAL % LwrLimit UpperLim 1 Red (pot shortage) 129 71 36 236 3.19% -999 0 4,000 2 Yellow (risk for shortage) 229 93 83 405 5.48% 0.1 5 3 Green (Opt Days Supply) 3,560 441 1,380 5,381 72.78% 5.1 7 4 EXCESS INV ($$$) 716 410 246 1,372 18.56% 7.1 999 Total Part numbers 4,634 1,015 1,745 7,394 5 PNs with -999 DOH 0 0 0 0 0.00% 6 PNs with over 100 neg(DOH) 1 0 1 2 0.03% 3,000 7 PNs with over 30 neg(DOH) 12 9 3 24 0.32% 8 PNs with less than 0 DOH 92 58 25 175 2.37% 9 PNs with 0 DOH 229 93 83 405 5.48% 10 PNs with DOH less than 5 3,560 441 1,380 5,381 72.78% 11 PNs with DOH less than 10 171 75 40 286 3.87% 12 PNs with DOH less than 30 144 65 14 223 3.02% 2,000 13 PNs with DOH less than 999 94 66 20 180 2.43% 14 PNs with 999 DOH 292 200 159 651 8.80% 15 Avg neg(INV_DOH) excl -999 DOH -27.9 -12.7 -157.4 16 Avg INV_DOH excluding 999 DOH 865.5 575.3 936.9 17 Generic MRP Controllers (no owner) 2,919 253 812 3,984 53.88% 18 MRP Type = PD (SAP generated) 4,457 1,015 1,744 7,216 97.59% 1,000 19 MRP Type = P4 (user sched some) 115 0 1 116 1.57% 20 MRP Type = ND (NO SAP MRP) 25 0 0 25 0.34% 651 21 Rounding Values undefined 2,686 137 1,584 4,407 59.60% 405 286 175 223 180 Exceptions Groups COMPONENTS CABL HARN TOTAL % 0 2 24 1 Late in moving to a proposal 0 0 0 0 0.00% 0 PNs with over 100 neg(DOH) PNs with over 30 neg(DOH) PNs with -999 DOH PNs with 0 DOH PNs with 999 DOH PNs with less than 0 DOH PNs with DOH less than 5 PNs with DOH less than 10 PNs with DOH less than 30 PNs with DOH less than 999 2 Late in moving to a commitment 0 0 2 2 0.03% 3 Stock should have been there 389 107 57 553 7.48% 4 A new requirement 0 0 0 0 0.00% 5 BOM related issues 0 0 0 0 0.00% 6 Too much or too little stock 0 0 4 4 0.05% 7 Dates when needed/available differs 466 130 37 633 8.56% -1,000 8 Marked for Deletion? 0 0 0 0 0.00% Total Part numbers with Exceptions 855 237 100 1,192 16.12% 54
  • 56. What We’ll Cover … • Establishing and tracking metrics for data quality initiatives • Understanding how to build your own Information Quality Index • Browsing the Sigma levels of information quality • Monitoring and enhancing the business processes • Wrap-up 55
  • 57. Resources • www.sap-img.com SAP Tables Help File and ABAP Programming • www.dmreview.com/channels/data_quality.html White paper library • www.findwhitepapers.com/index.php Technology Research For Business Professionals • www.ittoolbox.com/ Professional IT Community 56
  • 58. 7 Key Points to Take Home • Focus on data fields of interest (remember the SAP built-in validation for data during the creation process) • Keep current and future SAP functionality in mind during development • Identify proper SAP tables required for the Information Quality Model • Create simple queries (minimize more than two joins per query) • Use structured names for SQL tables and programs • Share system ownership with functional areas by co-authoring rules and resolving their issues • Maintain high coloring standards for Information Quality and Business Performance Assessment (Red/Yellow/Green) 57
  • 59. Your Turn! How to contact me: Jose V Zavala jose.v.zavala@delphi.com 58
  • 60. Disclaimer SAP, R/3, mySAP, mySAP.com, xApps, xApp, SAP NetWeaver®, Duet™, PartnerEdge, 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. Wellesley Information Services is neither owned nor controlled by SAP. 59