What gets measured gets improved

1
(No Transcript)
2
(No Transcript)
3
"What gets measured gets improved"
4
"What gets measured gets improved"
5
Introduction
6
Reducing "out of stocks" satisfies a key consumer
need
Consumer needs
67
Shorter queues
1
52
More promotions
2
30
Fewer items OOS
3
17
More employees
4
. . .
6
Cleaner facilities
7
7
Questions to the audience
What are you doing when you as a shopper are
confronted with an out-of-stock?

How many of you would leave the store without
purchasing anything?
8
9 of purchases are lost due to "out of stocks"
Consumer responses , 15 categories
21
37
9
17
16
Buys a different size
Returns later
Doesn't buy anything
Buys a different brand
Buys brand elsewhere
12-31
11-20
4-10
21-65
8-41
9
Questions to the audience
When you as a shopper enter a supermarket how
many of the usually stocked items are missing on
the shelf?

10
All existing studies show an OOS rate of between
7 and 10
!
OOS rate 7-10
Global study on OOS
16 previous European studies on OOS
11
11 national ECR initiatives and the GLP are
working on OSA
12
What's at stake?
13
Industry view OOS in Europe cost EUR 4 bn a
year in lost turnover
Lost turnover p.a. EUR
4.0 bn
14
Moreover, losing only one loyal family due to
OOS is a serious financial blow
Family value
EUR 150 per week for 20 years
EUR 150,000
15
Questions to the audience
Who is currently working on an OSA improvement
program and what are you doing specifically?

16
European measurement approach
17
Excuses not to measure OOS are quite common and
convenient
Excuses
In reality

• ECR measurement methodology available
• Detailed root cause catalogues exist
• Many pilot experiences documented

1
Measurement is toocomplex
2

• OSA performance depends on whole supply chain
• Operational inefficiencies lead to resource
  requirements

Need for resources
3
Detailed POS data is required

• Manual measurement is possible
• Educational impact is significant

18
Measurement and management attention are
important prerequisites
3
Replenishment system
1
2
Coherent consumer centric business system
4
Merchandising
Manage- ment attention
5
Measure- ment
Inventory accuracy
6
Promotion management
7
Ordering system
19
OSA projects should be based on a common
measurement method
PARTICIPATINGCOUNTRIES
COMMONMEASUREMENT
OBJECTIVES
UK

• Initiates and facilitates measurement process
• Enables benchmarking
• Increases management attention
• Stimulates European knowledge exchange

D
Countries
I
SM
HM
Store formats
DC
Candies
Ice cream
Catego- ries
Pasta
20
Two basic approaches are possible
OSA MEASUREMENT
Direct
Indirect
Physical counting
POS data
Without third party
With third party service provider
Without third party
With third party service provider

• Effective out-of-stocks
• Combination with root causes

• Indicative potential OoS
• Day-to-day alert
• Store performance indicator

FOCUS
21
The common measurement method is based on PoS data
REPRESEN-TATIVE
SIMPLE TOOPERATE

• Reference product basket
• Store format
• Local coverage

• Data collection
• Index calculation

22
Based on ECR France and pilots, a common
measurement approach was developed
Jan. / Feb. 04
A
Feb. 04
B
March 04
C
June / July 04
D
OOS DETECTION METHOD
DATA GATHERING AND PROCESSING
OOS CALCULATION
COMMON OOS MEASUREMENT METHOD
PHASE

• Detecting OOSbased on abnormal sales patterns
• - Sales velocity
• - Sales volatility

• 12m transaction items covering 2 months of POS
  data
• About 37,000 SKUs

• Analysis of sales patterns
• Index calculation for stores and categories

• Documentation of OOS measure-ment approach
• Roll-out re-commendation

CENTRAL TOPIC
23
For measuring OOS with POS data sales velocity
and sales volatility are key
DEFINITION
SUGGESTED CUT-OFF VALUE
KPI

• Number of SKUs sold per day and store

• More than 10 units per day

SALES VELOCITY
gt Sales variation and percentage of days with
zero-sales per SKU

• Variation coefficient (VC) of more than 80 or 100

SALES VOLATILITY
24
POS-based measurement works best for SKUs with
high velocity and low volatility
Results pilot in Sweden of sales in each
quadrant
gt 10 items / day
16
SALES VELOCITY
32
11
lt 10 items / day
VC gt 100
VC lt 100
SALES VOLATILITY
25
The Greek pilot experiences confirm this approach
gt 10 items / day
82 counts (1) 1 sales
SALES VELOCITY
1,124 counts (13) 15 sales
6,874 counts (81) 25 sales
lt 10 items / day
VC gt 80
VC lt 80
SALES VOLATILITY
26
Decisions on category definition and on
measurement are key
CATEGORY AND FORMAT DEFINITION
DATA GATHERING
DATA PROCESSING

• 20-30 categories
• Top 10 items per category
• Participating retailers
• Hypermarkets and supermarkets

• SKU
• Store ID
• Units sold
• Date
• EAN code
• Additional information

• Clear definition of basket definition and
  measurement method
• Initiative and res-ponsibility to kick off
  measurement lies within national ECR initiatives
  and individual companies

27
Summary key learnings ECR Europe
Measurement of OOS is possible at limited costs
and with a jointly developed European methodology
The methodology has been tested and proven in
different countries, store formats and
categories
The basis for companies to improve the OOS
situation is set now it is time to act and
work on all improvement levers
28
ECR Greece Index validation and on-line
monitoring
FORWARD
REWIND
START
Click to
29
Objective

• To verify the European Methodology of O.O.S.
  measurement through a physical survey

30
Comparing ?M to Physical OOS Survey

• Physical OOS Survey
• 2 Retailers Veropoulos and Metro
• 12 stores
• 4 suppliers
• 1 service provider
• 1 week 2-7 February 2004
• 110 products
• 5 clusters
• One cluster with products participating in
  in-store promotions
• One cluster per quadrant

31
The European Methodology (EM)
High
SALES VELOCITY
Low
High
Low
SALES VOLATILITY
32
Calculation of OOS index

• 1. SALES VELOCITY
• Average number of items sold per day (Monday,
  Tuesday, )
• Best performing value 10 items per day
• Other alternatives tested
• 7 items per day
• Lower number of items for smaller stores

33
Calculation of OOS index

• 2. SALES VOLATILITY
• Comparing average to standard deviation
• Original threshold value 100
• Best performing value 80
• Other tested value 70

34
EM Results
Q2 82 counts (1) 6 sales
Q1 490 counts (5) 53 sales
gt10 items / day
SALES VELOCITY
Q4 6.874 counts (81) 25 sales
Q3 1.124 counts (13) 15 sales
lt 10 items / day
Var. coefficient lt 80
Var. coefficient gt 80
SALES VOLATILITY
35
POS data versus physical counting our results
are encouraging
PHYSICALOOS SURVEY
gt 10 items / day
Total OSS 17
SALES VELOCITY
EM
Total OSS 23, Including all 17 plus 6 more
lt 10 items / day
High
Low
SALES VOLATILITY
36
Example chocolate bars Good match between POS
data and counting
Chocolate Bar 30GR items / day
30
Avg. Std.
POS
25
20
15
10
Avg. Std.
5
0
MON
TUE
WED
THUR
FRI
SAT

• The index captures all full OOS situations for
  products in Q1 (high velocity, high volatility)

37
As expected, the POS indication does not always
match with the physical survey
Chocolate Bar T.8
18
Avg. Std.
14
10
6
Avg. Std.
2
POS
-2
MON
TUE
WED
THUR
FRI
SAT

• This is one of the 6 incidents where the index
  detected an OOS but the Physical OOS Survey didn't

38
An Out-of-shelf index for slower-moving items
Q2 82 counts (1) 6 sales
Q1 490 counts (5) 53 sales
1
SALES VELOCITY
92
Q4 6.874 counts (81) 25 sales
5
Q3 1.124 counts (13) 15 sales
O.O.S.
SALES VOLATILITY
39
Identify "forgotten" products

• "Forgotten" products
• Belong to the store's assortment
• Used to sell some time ago
• Haven't sold anything for quite some time unlike
  their past behavior

• Look at past periods with zero salese.g. how
  many days on average did a product not sell in
  the past?
• Compare to last period with zero sales

40
An Out-of-shelf index for slower-moving items

• Use assortment data
• to identify situations where the product belongs
  to the typical store assortment but not in the
  store
• Use sales data
• to identify situations where the product has been
  "forgotten"

41
How the new index works
Avg 4,3 days no sales
Last period 29 days no sales
Last day of sales
42
New index for slow movers our results
Physical OOS survey vs. new index
10
8
4
2
0
Physical Survey
New Index
43
Online access to the out-of-shelf measurement
RETAILER
SUPPLIER
Replenishment
Collaboration Platform
STORE
SALESMAN
INTERNET
INTERNET
Dynamic Statistics
CENTRAL OFFICES
CENTRAL OFFICES
44
Incorporating the OOS measurement in the Order
Proposal
Store Assortment
Order Proposal
Physical shelf check
Final order
45
Whats is in it for retailers?

• Veropoulos
• Both indices are considered useful
• Next step Incorporation in the store
  replenishment tool
• Metro
• The POS-based index is useful there is
  confidence about its validity
• The index for slow movers would be useful, but
  further validation is required
• Next step Validate the new index using longer
  history data

46
What is in it for suppliers?

• Measuring the problem is the first step towards
  solving it
• A common measurement available to both retailer
  and supplier is a prerequisite for further
  collaborative action
• Easy, daily access is what makes it feasible
• Confidence in the EM is high, but it has limited
  scope
• The "New Index" looks more promising as it can
  capture a broader range of SKUs, but needs
  further validation
• The ultimate objective is to solve the OOS
  problem, which requires commitment on both sides

47
Summary key learnings ECR Greece
POS-based measurement is working as the
validation in Greece shows
An index for slow movers has been developed with
promising results. Further validation seems
necessary
An Internet-based collaboration platform is set
up to allow for "real-time" measurement and
corrective actions
48
ECR France From OOS measure-ment to OSA
management
49
A joint continuous improvement process is the
basis of collaboration
ALERT
STAKE EVALUATION
RESULTS
MEASUREMENT
RANKING
IMPLEMENTATION
ROOT CAUSES INVESTIGATION
CORRECTIVE ACTIONS IDENTIFICATION
50
The methodology works and the collabo-ration
gives results
As validated by 3 pilots
RESULTS (2001)
gt AUCHAN-LESIEUR gt CORA-NESTLE gt CORA-L'ORÉAL
OOS
Ø
Shower gel
Deodo- rants
gt AUCHAN-LESIEUR gt CORA-NESTLE gt CORA-L'ORÉAL
6.4
1.1
8.5
BEFORE
2.6
0.0
3.7
AFTER
MeasurementManual store check
51
In 2001 we developed a new measurement tool based
on PoS data and scanning
Daily rate full stock-outs 1 / 6 17 Daily
rate partial stock-outs 2 / 6 33
Zero sales
Lost sales Usual sales () real sales ()
MON
TUE
WED
THUR
FRI
SAT
52
In 2004 more than 2700 stores, 11 retailers and
250 categories are using this tool
CATEGORIES
250
12
8
3
9
8
6
2
11
RETAILERS
53
What's new about measurement?

• We now take into account
• Seasonal effect of the category in the trading
  area of the PoS
• The daily traffic evolution of the studied PoS

Trigger point of partial OOS Standard approach
Trigger point of partial OOS including seasonal
effects and PoS traffic
54
Daily results from the new barometer started in
2004 example ice cream
T. FAMILLE CHOCOLAT AU LAIT
7
8
9
8
6
5
7
6
12
9
8
8
9
DRPS
19
11
12
13
20
10
14
5
6
8
7
9
7
DRFS
M. FAMILLE
1
2
3
3
1
2
2
3
4
6
5
6
5
DRPS
19
17
19
16
29
12
18
9
12
19
21
27
18
DRFS

• 10 to 25 items followed during one year
• From 1700 hypermarkets and supermarkets
• On a daily basis
• Each day (from Monday to Saturday)
• Delivered every 4 weeks, 12 days after the end of
  the period

55
More and more players are involved "Learning
from 11 pilots"

• Validation of the measurement tool
• Amora Maille, Auchan, Bel, Candia, Coca-Cola,
  Cora, Ferrero, Evian, Heineken, Henkel,
  Kellogg's, Lesieur, L'Oréal, Nestlé, SVF
• One to one pilots
• Auchan Nestlé
• Auchan Vania
• Carrefour Evian
• Match - Ricard

• Système U Amora Maille
• Système U Coca Cola
• Système U Vania

56
From OOS measurement to OSA manage-ment What
are the key questions?

• Is my company ready to implement an OSA project?
• How to choose the products and partners I work
  with?
• How to go from awareness to project
  implementation?
• How to set up the project team?
• How to mobilize on the subject?

• How to break the barriers?

57
Is my company ready to implement an OSA project?

• 3 REQUIREMENTS
• Top management involvement
• Organizational ability to collaborate
• Organizational changes anticipated and accepted
• 4 ENABLERS
• People availability
• Value relationships (collaboration level)
• Performance measurements
• Technology

58
How to mobilize the company? Measure the stakes!
Partial OOS
13,1
Full OOS
6,5
2,4
Total lost sales A
Total lost sales B
Total lost sales C
59
How to choose the products and partners to work
with?
OOS STAKES

High potential but with another partner
High potential
ABILITY TO WORK TOGETHER


High potential but on another category
Low potential

60
Which categories to focus on?
Daily rate partial stock-outs
Daily rate full stock-outs
61
Which banners and brands to focus on?

• Internal studies (e.g. sales force statement )
• Studies of panellists
• Consumer inquiries
• Companies relationship
• Marketing and trade internal strategy

62
Find out the POS to begin with the case of
Ricard and Match
Banner average
HIGH POTENTIALS
PRIORITY
10
PARTIALRATE OOS
8
6
4
2
BEST IN CLASS
FULLRATE OOS
0
0
2
4
6
8
10
12
63
Where to begin The Auchan-Nestle Pilot
Potential lost turnover per store
18
PARTIALRATE OOS
16
4,495 Euros
14
1,684 Euros
12
10
8
654 Euros
6
1,110 Euros
4
2
FULLRATE OOS
327 Euros
0
0
2
4
6
8
10
12
14
64
How to set up the project team?
MANUFACTURER
RETAILER
Top management of both companies
SPONSOR
Category manager, Buyer, Supply Chain, sales
Category manager, sales management, Supply Chain
Service rate to consumer
PROJECT TEAM
PROJECT TEAM
IT, POS, Marketing...
IT, sales force, Marketing...
EXPERTS
EXPERTS
65
Summary key learnings ECR France
The developed measurement approach has been
adopted by a vast majority of French retailers
and manufacturers
Refinements of the basic approach have been made
to account for seasonal and regional differences
ECR France is building the bridge from
measurement to management through identifying key
implementation barriers and requirements
66
ECR Italy OSA roadmap and country monitoring
system
67
ECR Italy has developed a roadmap from OOS
Measurement to OSA Management
2003 OOS Understanding
2004 OSA Methodology
Measurement and Store Check

• Common language
• Collaborative approach
• End-to-end supply chain diagnostics

35 CATEGORIES 51 SEGMENTS
BLUE BOOK
10 Retailers
ECR Europe GLP
26 Manufacturers
68
ECR Italia and ECR France approach OSA is a
system issue
FROM OOS MEASUREMENT
TO OSAMANAGEMENT
OOS understanding and testing
Country monitoring
FOCUS
Prototyping
Standard sharing, Common language and algorithms
METHODOLOGY
One-to-One
Many-to-Many
APPROACH
Task force
Company organization
PEOPLE
Short term
System solution
SOLUTION
69
One key pillar of the roadmap is our country
monitoring system
2003
2004
10 Retailers 26 Manu-facturers
Understanding and testing
OSA management

• Standard measurement
• Best practices
• Balanced scorecard

• Common language and definitions
• Market survey
• OOS Measurement methodology

6 Retailers 12 Manufacturers
6 PILOTS AIMED AT STORE CHECKING

• MONITORING SYSTEM
• CONTINUOUS IMPROVEMENT

VARIOUS WORKSHOPS
PHYSICAL COUNTING
ELECTRONIC COUNTING
PERFOR-MANCE
PEOPLE
70
34 Companies have been involved in our market
survey (1)
RETAILERS
Categories
12
8
INTERDIS
CARREFOUR / GS
No. of returned questionnaires
PAM
COOP ITALIA
Detergents
RINASCENTE / SMA
CRAI / CODÉ
Total Grocery
2
5
SELEX / MIDA3
DESPAR / SADAS
Beverages
3
3
2
3
Dairy
2
Personal Care
Food
71
34 Companies have been involved in our market
survey (2)
MANUFACTURERS
Categories
12
26
Segments
51
J J
BARILLA
JOHNSON WAX
BEIERSDORF
Pet Care
LA MOLISANA
CAMEO
Beverages
LAVAZZA
COCA-COLA
2
2
2
2
2
2
Personal Care
4
4
4
4
4
4
LEAF
CARAPELLI
Dairy
4
LEVER FABERGÉ
COLGATE
17
17
17
17
17
17
LORÈAL
FATER
MANETTI R.
FERRERO
10
10
10
10
10
10
Food
MASTERFOODS
GILLETTE
NESTLÉ
GLAXO SMITHKLINE
14
14
14
14
14
14
PROCTER GAMBLE
GUABER
Detergents
RECKITT BENCKISER
HEINEKEN
SANPELLEGRINO
HENKEL
No. of returned questionnaires
72
Key results The best performing catego-ries are
detergents, food and dairy
95 4
94 5
92 7
85 10
82 15
80 17
On shelf availability
Promotional
Manufacturers
Retailers
Average
Iper
Continuous
Dispersion
Super
Deter-gents
Personal Care
Food
Dairy-Frozen
Pet Care
Bever-ages
73
The survey also highlighted the logistic KPIs
connected with OSA
Quality ( Stock-out)
OSA
Food
Detergents
20
Dairy - Frozen
15
Personal Care
10
"LEAN"
"RESPONSIVE"
KPI
5
5
10
KPI
15
20
25
30
"FLEXIBLE"
Time (order-delivery cycle) (supply frequency)
Costs (Coverage days)
KPI
74
In addition, manufacturers and retailers rank
improvement levers differently
Improvement areas and priorities
Improve Program Precision
Motivate Human Resources
Computer Aided Replenishment (CRP)
Increase Stock Visibility (VMI)
Decrease the Time Cycle order / delivery
Increase order frequency
75
Pilot store checks have been conducted with 18
companies in 12 categories
PILOT RETAILERS
CO-PILOT MANUFACTURERS
SEGMENTS
Detergents / Laundry Personal Care Oral Care Hair
Care Pasta Snacks Bakery Coffee Dairy
Foods Frozen / Ice Cream Beverages / Beer Pet Care
76
Country monitoring requires common language and
rules on OSA Measurement
MANU-FACTURER OOS
PoS

• Two criteria for the OOS
• Retailer Focus on PoS (all items)
• Manufacturer Focus on items (all PoS)

1.000

7
6
5
4
3
2
1
1.000

7
6
5
4
3
2
1
2
2
3
3
4
4
Items
5
5
6
6
One item in many PoS OOS of the item (ex.item
4)
7
7


100
100

Same PoS for all items OOS of the PoS (ex. PoS
5)
RETAILER OOS
1
2
3
100 BRAND 1-2-3
Brand ABC

Different criteria for fast moving items selection
Category ABC
100 SHELF / CATEGORY
77
Physical counting allows to investigate the
entire supply chain
Check points from the shelf to the plant
Plant
Back Store
DC
CWH
PWH
4/5 times / day
Retailer(All products)
Retailer(Pilot products)
Manufacturers (co-Pilot)
Only Promotional Merchandising
Direct Gathering
Interface Visibility
Presence
Degree of Participation
78
The checks highlighted the main improve-ment
areas and levers
DC
Retailer DC
Backstore
Shelf
SINGLE PLAYER LEVERS
Forecast accuracy
Space allocation
Computer aided re-ordering
Shelf replenishment
Manufacturer availability
Manufacturer availability
COLLA-BORATIVE LEVERS
SKU phase in / phase out
Promotion rules alignment
Packaging size and consumer units per box
Data base alignment
79
Two months later most categories have performed
significantly better
OOS
ORAL CARE
HAIRS
CAT CARE
WASHING
DISH WASHING
SNACKS
DES-SERT
YOG-URT
GASTRO-NOMY
FIRST CHECK
SECOND CHECK
80
The General Learning Program on OSA needs to
involve a wide population
OOS Best lt 2-3 / Worst gt7-8 / Peak gt 30-35
81
A comprehensive in-store inventory control is
necessary
DAILY INBOUND QUANTITY
OOS VALUE
OOS DURATION
Stock
OVERSTOCK
INVENTORY LEVEL
DAILY OUTBOUND QUANTITY
STOCKOUTLINES
(52 SKU)
82
Our country monitoring system and self-assessment
provide useful benchmarks
SELF ASSESSMENT QUESTIONNAIRES
MONITORING DATA
Quality
Country OOS average by category
Time
Costs
COMPANY DATA BENCHMARKING
COMPANY DATA BENCHMARKING
83
The OSA Balanced Scorecard links the main KPIs in
a systematic way

• FINANCIAL RESULTS
• Sales
• Margin

• CONSUMER VALUES
• Service
• Reliability

OSA BALANCEDSCORECARD

• KEY PERFORMANCES
• Effectiveness KPI
• Efficiency KPI

• KBP AND HUMAN RESOURCES
• Growth
• Evolution

84
Summary key learnings ECR Italy
ECR Italy is also following the basic measurement
approaches of ECR Europe and ECR France
The current focus of ECR Italy is on the OSA
roadmap by investigating the link between OSA and
the performance of the whole supply chain
Key pillars of this roadmap are a country-wide
monitoring system and an OSA Balanced Scorecard
85
Conclusion
86
How to get started with OSA
87
Outlook RFID will come, but it will take time
88
Optimal Shelf Availability
gt
89
(No Transcript)