Finding & Testing SIMPLE Solutions for COMPLEX Problems. CASE STUDY 1 Helping achieve “ One-Simple-ABB ” with “ TOC in SAP” within Complex Manufacturing Environment. CASE STUDY 2 Finding & Testing a solution to SHORTAGES & SURPLUSES within Book Publishing Supply Chain. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Case Study 1Case Study 1Helping Achieve “Helping Achieve “One-Simple-One-Simple-
ABBABB” with ” with TOC in SAPTOC in SAPPresenter: Dr Alan Barnard, CEO Goldratt Research LabsContributors: Dr Katja Rajaniemi, Improvement Manager ABB BU
Fredrik Nordstrom, Regional Manager ABB Ops Development Group Lukasz Krupa, Leader, ABB Manufacturing IS Solutions Alex D’ ’Anci, Regional Manager, ABB Ops Development Group Eli Schragenheim, Goldratt Schools
− Facts about ABB− ABB’s Continuous Improvement Evolution at ROI− The “One-Simple-ABB” Challenge for Operations Excellence
− Conflict in leveraging and standardizing on best practices across ABB− Conflict in leveraging and standardizing on a single IT Platform and ERP
system
− Finding a Solution to the “One-simple-ABB” Challenge− Defining a simple yet robust TOC solution for 300+ different factories− Defining “TOC in SAP” using Goldratt’s “Strategy and Tactic” (S&T)
process − Testing “TOC in SAP” solution through series of Pilots
− Results achieved to date & next steps• Research Conclusions
RESEARCH PROBLEM− Most companies today, have left the final choice of which
which Planning, Execution and Continuous improvement (PECI) methods and which IT systems to support these to Business Units because BUs are:− Responsible for continuously improving on their
result.− In the best position to decide which methods &
systems best meet their specific requirements. − Normally responsible for pay for these.
− No wonder BU’s have resisted most attempts to standardize or centralize IT and PECI’s …
RESEARCH QUESTIONSQ1: Is it possible to find a "one-solution-fits-all" both in rules
and ERP technology when it comes to planning, execution and improvement of operations especially considering the many specific local considerations resulting in "we are different”)?
Q2. Even if it was possible to theoretically find such a solution that could meet most business requirements, would it be possible to get buy-in from regional and plant managers under pressure to improve performance & reduce costs to test such a solution?
RESEARCH METHOD1.Literature Review to identify which companies
have tried and succeeded and which not and Why?2.Consultation with Experts to determine if a “One-
Solution-fits-all” solution design was possible and then whether such a solution can be implemented within a standard ERP System (the hypothesis to be tested)
3.Followed Action Research method using “Plan, Do, Review & Act” cycles to validate and continuously improve hypothesis
Case Study: Facts about ABBCase Study: Facts about ABB
ABB is the world’s leading provider of power and automation technologies with strong market positions in core businesses
ABB’s goal is to create value for their stakeholders by helping customers Use Electrical Power more Efficiently, Increase Industrial Productivity, Lower Environmental impact in a sustainable way
Revenues in 2007: $29.2 billion and Orders for $34.3 b (large backlog) Headquarters: Zurich, Switzerland About 107,000 employees in 100 countries with Market-leading positions in most key products Robust global value chain to serve established and emerging markets
What is the benefit of using TOC as a focusing tool What is the benefit of using TOC as a focusing tool for LEAN & Six Sigma vs. using each in isolation?for LEAN & Six Sigma vs. using each in isolation?
Sanmina-SCI Case Study•Leading global electronics manufacturing services (EMS) company employing over 48,000 people and revenue of $11.7 Billion
•Set-up experiment over 2.5 years to test financial impact of LEAN, Six Sigma and TLS (TOC focusing LEAN & 6S)
•21 Participating plants with:11 Plants on Six Sigma 4 Plants on Lean6Plants on TLS
•TLS, Lean, and Six Sigma all offered benefits
•TLS showed 3.9 times greater financial benefit delivering 89% of total benefits achieved from only 6 out of 21 plants
Per Project Financial Return
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Lean Six Sigma TLS
Methodology
Units Return
Contribution % to realized savings by method applied
Lean4%
Six Sigma7%
TLS89%
Source: Apics Magazine, May 2006 and TOCICO 2007 presentation by Dr Russ Pirasteh, Sanmina-SCI,
But how do you achieve Step-Change AND But how do you achieve Step-Change AND Continuous Improvement within any organization?Continuous Improvement within any organization?
Source: ‘Kaizen: The Key to Japan’s Competitive Success’, by Masaaki Imai
Despite the many differences between the plants where TOC+LEAN were implemented, it seemed to deliver similar results without any significant differences in the “what” and “how to”…
Need
Roll-outbest practices
effectivelyObjective
Ensure leverage of Best-Practices
in all units NeedEnsure out-
standing resultin each unit
Direction of solution:Use TOC to focus & synchronize Planning, Execution & Cont. Improvement solution (TOC + LEAN) and an participative engagement approach for getting contribution and consensus
ABB Gate Model for Process Improvement ABB Gate Model for Process Improvement projects of Best Practices using a holistic projects of Best Practices using a holistic TOC approachTOC approach
Project scope
definition
ProjectExecution
Plan
Final solution
agreement
ProjectHand -over
Pilot results
Close project
Validate resultsProject
StartAgreement
2 3 4 51 6 7Development Piloting Implementation0
Analysis Planning Execution Sustain
1. Agreement on the problem2. Agreement on the direction of the solution
4. Agreement that no disastrous side effects will result5. Agreement on the implementation requirements and the plan itself
3. Agreement that the solution will yield the desired results
…. addressing resistance to change as an organizational constraint:
6. Agreement by all collaborators including management that we can move forward with confidence
But how do a standardize “Best Practices” without standardizing the IT Platform?Source: ABB developed (based on Cooper´s Stage-gate model).
ABB’s ERP System reality after years of ABB’s ERP System reality after years of acquisitions and empowered BU and acquisitions and empowered BU and Regional IT decision making..Regional IT decision making..• As a result of numerous acquisitions, ABB at one time had more
than 500 ERP systems and 70 different ERP brands in operation across its businesses.
• Most of the factories and distribution centres were also using in-house developed Spreadsheets and Databases for supporting their preferred Planning, Execution and Continuous Improvement Methods
• Integration of systems and data (and support) had become a nightmare and major risk to the business.
• In late 2006, a decision was made at the ABB board, to embark on a (challenging) journey of standardization on a common IT / ERP Platform (or at least range of platforms) and simplification based on “best-practices”.
• The initiative was titled “One-simple-ABB” or “OsA”.
ABB’s “One-simple-ABB” decision ABB’s “One-simple-ABB” decision and consequences...and consequences...
Press Release in March 2007
“Today, ABB signed a strategic agreement with the German-based ERP provider, SAP to help deploy common SAP ERP software through its global operations to help unify and simplify some of ABB’s most important business processes. The OsA initial target is one ERP per country, and ultimately one ERP platform per region for all of ABB that would enable a high degree of standardization in the human resources, finance and administration and corporate governance functions as well as in the operational excellence “best-practices”
Yes, BUT....Would it really be possible to find and get agreement that a “One-simple-ABB” solution was possible for managing the
The Operations dilemma in supporting ABB’s The Operations dilemma in supporting ABB’s global “One-Simple-ABB” objective?global “One-Simple-ABB” objective?
NeedEfficient and fully integrated Finance, HR & OPS business processes across
ABBObjective
Create long term competitive edge in
our industry NeedContinuously
Improve customer service and sales
through efficient & effective operations
ActionDeploy a common
ERP platform across ABB and limit
selection of SCM & MES systems
ActionAllow each site to develop own / use 3rd party ERP, SCM
& MES systems
Each BU/Factory is unique and using a common ERP platform or limiting choices in SCM / MES methods & systems will jeopardize results and or ownership
Efficient and fully integrated Finance, HR & OPS business processes across
ABB
Create long term competitive edge in
our industry
Continuously Improve customer service and sales
through efficient & effective operations
Direction of win-win solution:•Get agreement that TOC can provide “one-solution-fits-all”•Find a way to Modify SAP R3 to support TOC (TOC Planning, Execution & POOGI rules).•ABB to launch a “TOC in SAP” initiative to validate above hypothesis
The “One-Simple-ABB” Solution for Operations
Yes, BUT…Q1: Would it really be possible to define a TOC “One-solution-fits-all”?Q2: Would it really be possible to modify SAP to support this solution?
The Operations dilemma in supporting ABB’s The Operations dilemma in supporting ABB’s global “One-Simple-ABB” objective?global “One-Simple-ABB” objective?
Finding a simple & robust SCM solution for Finding a simple & robust SCM solution for ABB that is generic enough to be used by all ABB that is generic enough to be used by all factoriesfactories
“The best solutions start with the right questions…”
2.1Reliability/Availability Competitive
Edge
2.2Rapid Response
Competitive Edge
1ABB Goal:
Profitable Growth
3.1.1Remarkable
Due Date Perf & Availability
3.1.2Selling
Reliability/Availability Selling as Comp. Edge
3.1.3Coping
with Sales Growth
3.2.1Remarkable
Rapid Response
3.2.2Protective
Capacity for RR orders
3.2.3Selling Rapid Response as Comp. Edge
3.2.4Expanding RR Client
Base
Using Goldratt’s Strategy & Tactic tree method to find a simple & robust Planning, Execution & Cont. Improvement solution…
Releasing Too much / Too earlyMaterials / Purchased Parts not available
Unsynchronized Priorities in OperationsOther “Hidden” Disruptions to Flow
What really causes low Tput, poor DDP / Availability & Long LT?
Finding a simple & robust SCM solution for Finding a simple & robust SCM solution for ABB that is generic enough to be used by all ABB that is generic enough to be used by all factoriesfactories
How do we control release of ETO, MTO & MTS WO’s for Low WIP/high
CCR utilization?
Execution
How do we maintain right
priority of ALL WO’s on shop floor?
Ongoing Improvement
How do we continuously Improve flow by
identifying & removing local optima & other
“hidden” disruptions/ CCRs to improve flow?
Finding a simple & robust SCM solution for Finding a simple & robust SCM solution for ABB that is generic enough to be used by all ABB that is generic enough to be used by all factoriesfactories
Finding a simple & robust SCM solution for Finding a simple & robust SCM solution for ABB that is generic enough to be used by all ABB that is generic enough to be used by all factoriesfactories
Defining, communicating & validating the Defining, communicating & validating the proposed ERP changes using S&T structure proposed ERP changes using S&T structure
4.11.1 (Sales)
TOC’s PLANNED
LOAD (PL) for Quoting Safe
Due Dates
Why is the change needed?
Necessary Assumption
What is the objective of the change?Strategy
Why will the change achieve the objective?
Parallel Assumptions
How will the change be implemented?
Tactic
Quoting Safe Due Dates using CCR(s) Planned Load4.11.1• When Sales/Planning accept an order due date which the
factory cannot achieve, it jeopardize ABB’s Reliability CE.• Even if the constraint is in the market, fluctuations in
demand or supply can cause specific resources (CCRs) to be overloaded. Under such circumstances, quoting fixed lead times is very likely to result in missed due dates resulting in poor DDP.
Due-dates given by the sales force/planning are (almost) always met even during periods of capacity overloads (Target: 99% DDP)
1. Safe Due-date are given according to first available slot on CCR(s) + ½ Production Time Buffer
2. The ERP system is modified to provide Safe Due Date based on CCR Planned Load + ½ PTB within minutes and Sales/Planning is trained to use it.
• It is relatively easy to meet all due-dates when the commitments are given based on actual planned loads on the CCR(s) and S-DBR and BM are in place to control release & align priorities.
• The ERP systems can be modified to provide (within mins) a safe due date based on CCR(s) planned load + ½ Production Time Buffer (if CCR is about in middle of flow)
Quoting Safe Orders Due Dates using CCR(s) Planned Load4.11.1Tactic2. The ERP system is modified to provide Safe Due Date based on CCR Planned Load + ½ PTB within
Maintaining Optimum RM & PP Inventory Levels4.11.2• Having too much Raw Material (RM) and or Purchased Parts
(PP) in the stores or ordering too early can easily drain the company’s cash /WH space.
• Having too little RM, PP can and most frequently do cause delays that can cause lost production and potentially missed due dates (jeopardizing building of a “Reliability” Comp. Edge)
• .The target levels of RM & PP inventories held are continuously monitored and when needed are suitably modified (not too much/too little)
1. Target Levels are sized based on Max Demand within Reliable Replenishment time (Rt=OLT + SLT)
2. Inventory levels are maintained by replenishing on actual consumption & resizing on buffer penetration
3. The ERP system is modified to enable DBM & TOCR functionality
•TOC;s Dynamic Buffer Management & Replenishment of RM & PP Inventory, is a robust mechanism that enables setting, replenishing & adjustment of inventory targets, according to the actual level of demand & supply ensuring low levels of inventory& high availability.
•Most ERP systems do not provide a mechanism for auto resizing inventory target levels based on actual changes in demand/supply
TOC REPLENISMENT & DBM DETAILS
4.11.2 (Buying)
TOC’s DYNAMIC BUFFER MGT (DBM)
& TOC Replenishment
(TOCR) for Inv Mgt
Defining, communicating & validating the Defining, communicating & validating the proposed ERP change using S&T structure proposed ERP change using S&T structure
Auto Up-size Buffer based on level of Red-zone penetration
Auto Down-size Buffer based on no Yellow Zone penetration
TOC Planning Rule: Size Buffer Target Level based on “Peak Demand within Reliable Replenishment Time”TOC Execution Rule: Order daily & Replenish up to Target Level frequently (e.g. daily) on actual demandTOC Feedback Rule: Re-size buffers based on red-zone penetration
TOC Replenishment & Dynamic Buffer Management RulesWhat are the New TOC Planning, Execution & Feedback Rules…
Tactic1. Target Levels are sized based on Max Demand within Reliable Replenishment time (Rt=OLT + SLT)2. Inventory levels maintained by replenishing on actual consumption & resizing on buffer penetration
TOC Release Control to prevent Over-production4.11.3• Having too many orders on the shop floor masks priorities,
promotes local optima behavior and therefore prolongs the lead-time and significantly disrupts due-date-performance (DDP).
• Having too little orders on the shop floor will cause starvation of CCRs and cause lost production and potentially missed due dates (jeopardizing building of a “Reliability” Comp. Edge)
• .The shop floor is populated ONLY with orders that have to be filled within a predefined horizon.
1. For each group of products currently having similar lead times, a buffer time is set to be equal to 50% of the current avg lead-time. Orders are released to the floor only a buffer time before their committed due-date, and if all materials are available. (WIP frozen until its time arrives).
2. The ERP system usage is modified to support S-DBR release control and automatic material availability check.
• In traditionally run plants touch time is a very small fraction (<10%) of the lead time.• Vast experience shows that, in traditionally run plants, restricting the release of materials, to be just half the current lead time before the corresponding due date, leads only to good results and to no negative ramifications* (lead time shrinks to less than half, DDP improves considerably, throughput goes up and excess capacity is revealed). These results are achieved irrespective of whether or not a bottleneck exists.
• * Except for environments which are dominated by heavily dependent set-up matrixes. Those environments have to be dealt in a different way.
TOC RELEASE CONTOL (S-DBR) DETAILS
4.11.3 (Planning)
TOC’s RELEASE CONTROL (S-
DBR) to control WIP and improve
T, DDP & LT
Defining, communicating & validating the Defining, communicating & validating the proposed ERP change using S&T structure proposed ERP change using S&T structure
Manuf Parts Release Date= Due Date – ATO Shipping Buffer
Shipping Buffer for ATO
RM1
RM2
PP1
RM3
PP2
MP1
1. Raw Materials for Stock Manuf Parts Release based on “Replenish Frequently on Actual Consumption
MPS = DRUM = ORDERSPart # Qty Due Date Buffer FG001 5 13/3 66% FG002 10 16/3 23%
MANUF PARTS RELEASEPart # Qty OD DD Buffer TypeM001 5 05/2 04/3 66% ATOM002 10 07/2 07/3 20% Cust
Part sales to customers
GREEN ZONE RED ZONEYELLOW ZONE
Shipping Buffer for MTOGREEN ZONE RED ZONEYELLOW ZONE
2. Raw Material for Non-Stock Manuf Parts Release = DD – MTO Shipping Buffer
Stock Manuf (Common) Parts
Non-Stock (Unique) Manuf Parts
RAW MATRL RELEASE (ROPE)Part # Qty OD DD Buffer TypeP001 5 05/2 04/3 66% TimeP002 10 07/2 07/3 30% Stock
DefinitionsBuffer Status = (Buffer Time – Remaining Duration) / Buffer TimeRed Zone = Time req’d for expediting a medium-sized order
4.11.34.11.3 Controlling the Release in Manufacturing (MTO/ATO)
“Customer” Orders
GREEN ZONE RED ZONEYELLOW ZONE
S-DBR Planning (Release Control) Solution
Quote LT’s on Planned CCR Load
Tactic1. Target Levels are sized based on Max Demand within Reliable Replenishment time (Rt=OLT + SLT)2. Inventory levels maintained by replenishing on actual consumption & resizing on buffer penetration
Single Priority System to Synchronize Execution4.11.4• Hectic priorities (hot, red-hot and do-it-NOW) cause
chaos on the floor• Even when material release is properly choked, not
having a priority system can cause some orders to still be late..
The shop floor is governed by a simple, yet robust, priority system
1. TOC Buffer Management is the ONLY priority system used on the shop floor.
2. The ERP system (and MES/RFID systems) are modified to ensure the Purchase Order List, “Work-to-List” and Distribution Shipment List are prioritized based on buffer status (1st Black, 2nd Red, 3rd Yellow, 4th Green)
• Vast experience has shown that when work is released according to set time buffers, excellent results are obtained by using a crude priority system that is based solely on the time lapsed since the release.
• Buffer Management (BM) is setting priorities only according to the degree the buffer-time is consumed (four color code system - green: less than one third of the buffer time passed is lowest priority and black: more than the time buffer passed is the highest).
TOC SINGLE PRIORITY SYSTEM DETAILS
4.11.4 (Execution)
TOC’s Buffer Mgt (BM) as
synchronized SINGLE
PRIORITY SYSTEM
Defining, communicating & validating the Defining, communicating & validating the proposed ERP change using S&T structure proposed ERP change using S&T structure
Stock Replenishment Orders do not have “Due Dates”
Customer Production Orders do have “Due Dates”
Sample
Scr
een f
or sh
owing
Buffer
statu
s (Prio
rity)
of ea
ch or
der a
nd en
ablin
g
capt
uring
of re
ason
code
s for
“blac
k” an
d “re
d” or
ders
4.11.4 Single Priority System to Maintain Execution SynchronizationTactic1. Target Levels are sized based on Max Demand within Reliable Replenishment time (Rt=OLT + SLT)2. Inventory levels maintained by replenishing on actual consumption & resizing on buffer penetration
Focusing Mechanism for Process Improvement/Elevation4.11.5• Most local improvement initiatives in manufacturing, which use good tools (Root Cause analysis, Lean and Six Sigma techniques) do improve the local performance but, many times, those local improvements do not translate into global improvements
•Not having sufficient protective capacity results in long LT & poor DDP
All local improvements initiatives in manufacturing do contribute meaningfully to the global performance and there is enough protective capacity for high DDP
1. For all Red & Black Orders and Stock Buffers, Production Supervisors/workers/purchasers record “what was black/red order waiting for?”
2. CCRs are identified and effectively removed through focused process improvement / elevation to the extent that most loaded resource has at least 20% protective capacity.
3. The ERP (and MES) systems are modified for POOGI.
If a CCR exists, work-in-process piles up in front of it. When materials release is restricted, the only work centers that have work-in-process piling up in front of them are the real CCRs.
CCRs can be identified also by recording “what black & red orders were waiting for?” In most of the cases additional capacity can be exposed by better EXPLOITATION like: - Ensuring that CCRs do not take lunch or shift change breaks, - Offloading work from the CCRs to less “effective” work centers that have ample excess
capacity, - Using LEAN /Six Sigma techniques to shrink the set-up time/reduce variation on the CCRs, - ELEVATING capacity with overtime or capex approval for the CCRs, etc.
TOC BUFFER ANALYSIS (POOGI) DETAILS
4.11.5 (Improve)TOC’s BUFFER ANALYSIS to
Focus Improvements
& Capacity Elevation
Defining, communicating & validating the Defining, communicating & validating the proposed ERP change using S&T structure proposed ERP change using S&T structure
4.11.5 Focusing Mechanism for Process Improvement/ElevationTactic1. For all Red & Black Orders an Stock Buffers, Production Supervisors record “what was black/red order
waiting for?” 2. CCRs are identified and effectively removed through focused process improvement / elevation to the
extent that most loaded resource has at least 20% protective capacity.
“After analyzing SAP recommendations and data from nearly 2 months, we are confident that the the TOC rules were implemented correctly and that we are now confident that we can switch of our Off-line & 3rd Party systems.
The new Dynamic Buffer Management Functionality enables us to react quickly to changes in inventory values, which was not possible earlier, as we didn’t have time to monitor values manually.
We are very happy, are continuing to use the functionality and will be expanding it to other product lines”
Operations Managers
Factories in Brno and Vaasa
TOC in SAP at ABBTOC in SAP at ABBResults from Brno (Czech Republic) and Vaasa Results from Brno (Czech Republic) and Vaasa (Finland) (Finland)
TOC in SAP at ABBTOC in SAP at ABBNext StepsNext Steps
Implemented "TOC and SAP” now in 8 factories:•in Brno, Czech Republic•in Chonan, Korea, 2 factories•in Vaasa, Finland, 2 factories•in Baroda, India•in Przasnysz, Poland, 2 factories
And planning the implementations (in the work queue)•In Vaasa, Finland, 3rd factory•in Dalmine, Italy•in Ratingen, Germany. Has “SDBR in MES”•in Xiamen and Beijing China•in Skien, Norway.•…until all 300+ factories are reached…
Research ConclusionResearch Conclusion• The “TOC in SAP” initiative, showed that the ambitious goal of “One-
Simple-ABB” can be achieved even in the management of operations (Planning & Execution and Continuous Improvement) in complex and different operational conditions.
• The partnership between ABB and Goldratt Research Labs made it possible to:
− Define a ROBUST enough TOC Planning, Execution and Continuous Improvement solution that could cope with all the variations and complexities of 300+ factories
− Define a SIMPLE enough TOC Solution which could be supported with relatively small modifications within a standard ERP system such as SAP.
− Break new ground by using the Strategy & Tactic Tree to define, communicate and get buy-in for the new TOC solution in a way that focused and accelerated the achievement of consensus and development of the required SAP functionality…which is also accelerating all new SAP implementations
− Deliver this project and its associated results in a record time for SAP solution design, development and testing in less than 9 months….
Acknowledgments to Project Acknowledgments to Project Team ContributionsTeam Contributions• Stefan Forsmark, Operations Manager for ABB’s Power Products Division/BU Transformers.• Thomas W. Schmidt, IS Manager for ABB’s Power Products Division• Goethe Wallin, Operations Manager for ABB’s Power Products Division/BU Medium Voltage• Karol Kaczmarek, Global Project Manager Operational Excellence, ABB BU Transformers• Erich Beeler, IS Manager, ABB BU Transformers, Engineering IS• Dr Katja Rajaniemi, Global Process Improvement Manager, ABB BU Medium Voltage• Bill Vick, Global Operations Improvement Manager, ABB BU Medium Voltage • Miroslaw Bistron, Leader of Manufacturing IS program, ABB’s Power Products Division• Lukasz Krupa, Leader of Manufacturing IS program, ABB’s Power Products Division• Fredrik Nordstrom, ABB GF Q&O / ODG• Vesa Enestam, Eliaps Oy, SAP SCM Consulting• May-Jing Li, Project Manager, ABB Group Function IS, 'One Simple ABB' and SAP R/3• Martin Korthaus, SCM II Consulting, SAP• John Trip, Goldratt-TOC Ltd• Dr Alan Barnard, CEO, Goldratt Research Labs• Eli Schragenheim, Principal, Goldratt Schools• Dr Eli Goldratt, Chairman, Goldratt Group
Presented By: Presented By: Dr. Alan Barnard (CEO Goldratt Research Labs, TOCICO Chairman 2003-2006)Dr. Alan Barnard (CEO Goldratt Research Labs, TOCICO Chairman 2003-2006)Date:Date: 2727thth August 2009 August 2009
1. Generic Research Problem1. Extent and Consequences of Shortages & Surpluses2. Cause(s) & Direction of Solution
2. Case Study: Random House Publishing1. Publishing Industry Background 2. Random House Company Overview3. Phase 1: Quantifying the Extent, Consequences & Causes4. Phase 2: Finding a Solution5. Phase 3: Testing the Solution
Research ProblemResearch ProblemExtent & Consequences of Extent & Consequences of SHORTAGESSHORTAGES
1. How many of YOU frequently don’t find what you want at a SHOP? *Either because the retailer was “Out-of-Stock” (OOS) or it decided not to carry the item you wanted?
2. What is YOUR (consumer) response to this SHORTAGE?− Come back later to same retailer to buy same product?− Buy a similar product from same retailer and same supplier? − Buy a similar product from same retailer but different supplier?− Go to other retailer to buy same product?− Do not buy anything?
Results from International GMA Research Study (2002)Title: “Retail Out-of-Stocks: A Worldwide Examination of Extent, Causes and Consumer Responses”Scope: 661 Retail Outlets, 32 Product Categories, 71,000 Consumers Interviewed, 29 Countries
Retailers Lost Sales8 to 9% OOS x (9+31=40%) Consumer Response = 3 to 4%
Supplier Lost Sales8 to 9% OOS x (9+26=35%) Consumer Response = 3 to 4%
Impact on Lost Sales & Profitability
INDUSTRY RESEARCH: Extent, INDUSTRY RESEARCH: Extent, Consequences & Consumer Response to Consequences & Consumer Response to SHORTAGESSHORTAGES
But at 1:10 Leverage, 3 – 4% Lost Sales = Lost Profitability of 30 to 40%
1. Where is the real Constraint in the Consumer Goods Supply Chain (e.g. for printed books)? − Number of Customers willing to buy (books)……and what limits customers to buy more of the available books (OR what prevent
retailers to sell more books)?− Shortages (Out-of-Stocks) OR− Shelf Space available in Book Shops to display all available books that would sell if
they were available to buy...
2. What is the consequences of SURPLUSES on “Exploiting” the Retailer’s shelf space?
Surpluses is like having BRICKS on the Shelf……and BRICKS = LOST SALES of non-stock items selling elsewhere
Research ProblemResearch ProblemAnother important consequence of Another important consequence of SURPLUSESSURPLUSES
• More Sophisticated (and expensive) Forecasting Systems to predict consumer Demand…and more Collaboration efforts to share forecasts and supply problems
• Implementing Central & Regional Distribution Centers (CDC / RDC) to improve availability and responsiveness
• Bar Coding, High Tech Materials Handling & EDI
What was the result?Significant improvements in availability at CDC and RDC’s (95 – 98%)
and in responsiveness (daily deliveries)…
But…did it really improve Shortages and Surpluses at Retail level and if not, WHY?
Research ProblemResearch ProblemPrevious attempts to solve the problemPrevious attempts to solve the problem
Case StudyCase StudyPublishing Industry BackgroundPublishing Industry Background
• The publishing industry, due to slow or even declining sales, has been under SEVERE pressure to find ways to improve for many years…
• The book publishing industry think shortages is low but has experienced very high returns (surpluses) for many years (no-sale-return policy)* A Typical Book Publisher have 30% - 50% returns of all books printed
Pile’em high…. Watch’em fly…”
• Many Book Publishers & Retailers assumed that high returns was simply the “price you pay” to not lost any sales…but most still considered it as an improvement opportunity* A Typical Book Publisher’s Print & Distrib. Costs is only 30% of Selling Price
Case StudyCase StudyRandom House PublishingRandom House Publishing• In February 2008, Goldratt Research Labs was approached
by Random House Publishing.• They are considered one of the best in managing returns,
and still have 28% of every book printed returned to be shredded.
• There were no reliable data about the level of surpluses (and or shortages within retailers), and it appeared as if the consequences of these were not fully understood.
Could Theory of Constraints help to quantify the extent, consequences and causes of the problem… and help to find a solution?
• Market Position: Random House is the world’s largest general-interest book publisher, with ore than 11,000 new books issued a year and 500 million books sold annually
• Employees: 5,779 (as of December 31, 2008)• Revenues: € 1.7 billion (Fiscal Year 2008)• Shareholders: Bertelsmann AG (100%)• Published Authors include: F. Scott Fitzgerald, Ernest Hemingway,
John Gresha, Danielle Steel, Dan Brown and Barack Obama…
Case StudyCase StudyRandom House Key FactsRandom House Key Facts
Applying TOC’s 5 Focusing Steps to the Book Publishing Supply Chain…
Step 1: IDENTIFY the System’s Constraint
Step 2: Decide how to EXPLOIT the System’s Constraint
Step 3: SUBORDINATE everything to the above decision
“Until the consumer has bought, nobody has really sold…”The System Constraint therefore is the No. of Consumers willing to buy
“Exploiting the System Constraint” means “Having the Right Product, at the right place and right time (when consumer is ready to buy).
What conditions block better exploitation?1)Shortages of Products already stocked2)Unavailability of Products that sell elsewhere but which is not stocked due to shelf-space constraints (occupied by Surpluses).
Actually Surpluses in a retail shop is like having bricks on the shelf…
Change any Policy, Measurement and or Behaviour that contribute to current high level of both Shortages and Surpluses
YES, BUT, how to we find these and will it really be a win:win to change these…
Book Publishing Research Challenge #1What makes quantifying the Extent, Consequences and
Causes of Surpluses & Shortages so difficult?• Despite excellent “Point-of-Sale” data available from most retailers,
Publishers do not have visibility of actual shortages and surpluses at retails.
• Retailers can buy directly from Publishers (to get best price) or from Wholesalers (to get best availability and response).
• Therefore, knowing the availability (or unavailability) at one Publishers and Wholesalers do not give a true picture of unavailability at retail level.
Is there a simple way to quantify the likely level of shortages and surpluses at retail level?
Overcoming the challenges to reduce Overcoming the challenges to reduce shortages & surpluses in Book Publishingshortages & surpluses in Book Publishing
Book Publishing Research Challenge #2What makes reducing Surpluses & Shortages so difficult?
• Uncertainty of consumer demand• Weak incentives for retailers & distributors to improve returns…due
to an industry policy of “No Sales-Return Any time” creates an incentive to rather order too much than too little..
• Distributors & retailers create buffers for the same “end-consumer demand.
• Printing & Binding economies of scale that drive large batches.• A strong held belief that “Title visibility in the marketplace drives
demand (but there is little research to quantify this relationship)• Supply chain constraints like warehouse proximity and limited rush
reprint capacity at printers
The goal is to print & distribute closer to what sells to the end consumer
Auto Upsize Buffer based on level of Red-zone penetration
Keeping correct inventory levels4:25
Auto Downsize Buffer based on lack of Yellow zone penetration
Sizing Stock Buffers with Sizing Stock Buffers with Uncertainty and variability in Uncertainty and variability in demand and supplydemand and supply
Planning Rule: “Be Paranoid but not Hysterical”Target Stock Level for each Title at each Storage Location = Maximum Demand within Reliable Replenishment Time
Actual Demand
Execution Rules: Replenish up to Target level in sequence of buffer status more frequently (e.g. daily)
Too much
Too little
Good Enough
Goa
l Uni
ts
Time / Stock BufferFeedback Rules: Adjust Target Level (Buffer Size) based on Level of Buffer Penetration
• In general, Goldratt Research Labs have found that if a company reports formally a 1-3% “Lost Sales” due to unavailability of product, improving availability to close on 100% (to end consumer) typically will result in increase in sales of 10 to 30%
• The Table below shows the results from actual implementation tests:
Industry Official Lost Sales
Availability before TOC
Availability after TOC
Avg Increase in Sales within 1st year of TOC rollout
Fruit Supplier 2-5% 85 -90% >98% 30% (worst 5%, best 200%)
Bread & Flour Supplier
2-5% 80-95% >99% 20% (worst 5%, best 300%)
Cosmetic Supplier
2-5% 90-98% >99% 10% (worst 2.5%, best 50%)
Sportswear Supplier
10-20% 80 – 85% >97%(Total Inv down
60%)
25%(worst 10%, best 100%)
Note: In reality, most “lost sales” are not recorded. The dramatic increase in sales achieved with improved availability verifies this.
Results from Past TOC’s Replenish-to-Results from Past TOC’s Replenish-to-Consumption (RTC) ImplementationsConsumption (RTC) Implementations
Testing the Solution with 12 Testing the Solution with 12 Test vs. Control ShopsTest vs. Control ShopsResearch Questions:1. Acceptance Rate? • % of Shops accepting the offer for the Publisher to “Replenish Daily on Actual
Consumption
2. Difficulty to setup up RTC system?• How difficult would it be to set up a system for a shop to share daily sales with the
publisher and to receive daily deliveries?
3. Really Win:win (Value = Benefits – Costs)?• What will be the impact “Less more Frequently” on Shortages (Out-of-Stock) and
Results Achieved and Lessons Results Achieved and Lessons Learned so far?Learned so far?
• Importance of knowing how to answer simple question of “By how much has Sales gone up?” due to this change...
• Very high level of acceptance rate of offer to replenish daily based on actual sales (>80%)
• “Replenishing Daily on Actual Consumption” can help significantly reduce Shortages & Surpluse:
− Sales Increase by preventing Shortages around 2.5-10%− Sales increase by replacing Surpluses (if done on wide enough scale) can be 5-20%.
• Implementing Dynamic Buffer Management (with so many SKUs) and frequent bad decisions is expected to increase Sales even more
• Replacing a BRICKS with any Item that sells will increase Sales (BRICKS = More than Needed / Really Slow Runner)
• Major leverage opportunities to reduce SURPLUSES & SHORTAGES are the really HIGH Runners and the LOW Runners (1’s, 2’s & 3’s that should be 0’s,1’s & 2’s)…