Consulting
The Benefits of Store Clustering
Simon Smallwood Director Email –
[email protected] Tel - +44 7786 387793
7 Garrick St Covent Garden London WC2E 9AR T - +44 (0)203 051 1375 www.riverheadconsulting.com
Page: 2
Not so long ago.......
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Where everyone knew your name......
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 3
But times they were a changing.....
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 4
And the only constant is change.....
Page: 5
Pick n Pay V & A Wharf Cape Town SA
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Mass Merchandise, Mass Market, Mass Range, Mass Inventory...
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 6
Page: 7
So what’s in it for the.... Retailer: • Broadest possible range attracts broadest number of customers • Easy to manage – ‘One size fits all’
Customer:
• Buying & promotion efficiencies
• Vast range of choice
• Out range the competition
• All tastes catered for
• Logistics & Distribution efficiencies
• Secondary & Tertiary options
• Streamlined back office systems
• Competitive environment keeping prices down
Manufacturer:
• One stop shop
• Maximum distribution
• Bulk buying
• Optimum market penetration • Promotional Critical Mass • Minimum number of SKU’s
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 8
What is the real cost to retailers and do customers really benefit? 100
Sales Value
80
20
Inventory Value GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
100
Page: 9
Studies have shown that the annual additional cost of holding excess inventory can be 25% to 32%.
The Diamond of Doom Excess Inventory Leads to
Poor Cash Flow: Pressure from suppliers
Leads to
Excessive Obsolescence Pilferage, maintenance, insurance etc
Leads to
Leads to
Excessive Debt servicing
Lower Gross Margin Leads to
Leads to
High Advertising & Selling Expenses (To eliminate the excess)
High Interest Expense Leads to
Leads to
Lower Operating Profits GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 10
Traditional Retail Models define both ends of the spectrum...
Range & Value
Sales Volumes
High
Low
Local Convenience Store: • Destination Store • 1:1 Service • Knowledgeable Staff • Awareness of Needs
Mass Market Grocers: • Destination Store • Low Cost Provider • Range Breadth & Depth • Broad Appeal Customer Engagement Operating Costs GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
High
New retail models combine service & value to achieve high loyalty & profits
Range & Value
High
SupaValu USA – La Jolla CA
Low
Customer Engagement
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
High
Page: 11
Combining a strong commitment to service and value...
Page: 12
Mission Statement To provide the finest assortment and highest quality fresh and specialty foods from around the world - in a warm, friendly, and uniquely designed atmosphere with service and value that exceeds the expectations of our customers. Service: Knowledgeable, Helpful Staff Each Bristol Farms store maintains a large staff who are always available to offer assistance to customers. Atmosphere: Bristol Farms' stores have been carefully designed and decorated to create a singular shopping experience that evokes the local area.
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Store Clustering - Why do it?
Page: 13
• Introduce a ‘common language’ describing stores across the business • Improve store planning, assortment and merchandising • Tailor store space to match customer demand within each cluster • Provides the potential to offer differential cluster specific promotions
• At category and sub-category level determine optimum assortment • • • •
Enable informed predictions on demand levels for core range and new titles Optimise stockholding v demand Minimise overstocking Eliminate/reduce expensive returns of redundant stock
• Identify the external attributes that drive cluster performance to achieve a closer match to the needs of the customer profile store by store • Results in a higher rate of sale from a lower stock holding – improved ROCE
• Identify the internal factors driving optimum performance and enable the sharing of ‘best practice’ within the group
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 14
The Dynamics of Store Clustering
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 15
The dynamics of store clustering
Stores within a group do not perform in the same way despite how similar the product and price offers Both internal and external factors impact the performance of every store more or less In an ideal world we would treat every store as unique and range and merchandise to suit the customers who walk through each store In the real world we must seek to cluster stores by common attributes and performance patterns
Critical success factor – Simplicity. The entire company should be able to understand the clusters and describe the people and the stores that each cluster most strongly represents GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
The right store clustering programme results in increased customer satisfaction, compliance and improved supply chain efficiency
Page: 16
External variables significantly determine store performance Percentage contribution to store performance variability.
6%
78%
7% 10% 25%
E D
C
30%
B
A
F
Examples of ‘External Variables’ are:
External Variables
A – Local population and Competition (Population, competition, grocery spend within 5,10,15 minutes) B – Store size variables (Revenue, payroll, sq m, opening hours, profit contribution etc) C – Wider demographics (10-15 minute drive time) D – Local demographics (5 minute drive time) E – Store productivity (Productivity index, wastage, shrinkage, FT/PT ratio etc) F – Variability explained (22% not measurable or identifiable i.e. internal variables such as how good store manager is) GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 17
There are several approaches to store clustering used by retailers...
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 18
Size, Format, Spend - Matrix Main+ Average Main Average
First for Food
Proposition
Q F
Range / Choice
P S+ S
E
P = Premium Brands S = Standard Brands S+ = Standard +Brands E = Economy Brands
Promotions Policy
Q v
Q
Q
v
v
Q = Quality
Service levels
J
Format
£ £ Basic Standard No Frills
Mixed Meals
Making Life Taste Better For Less
First for Foodies
First for Fresh
F
B
E
P S+ S
Q
Q
v
v
Q E
F
P S+ S
Q E
P S+ S
Mixed Grab & Go Fast, Fresh, New & Exciting
F
Q
F
E
P S+ S E
Q
Q
Q
v
v
v
V = Value
J
Extended
Business Benchmark
Environment
P S+ S
Main High Main Average
Q
Q F
B
Own Label Levels
Superior Food + GM For Family & Home First for Foodies & Typical Families
Main+ Low Main Low
£
J
J
J
J £
£
J £
J £
J
J £
Flagship
Average Size & Avg Spend
Avg Size & Low Spend
Avg Size & High Spend
Smaller Local Store; Mixed Shoppers
Smaller Local Store: Young Single Shoppers
Q = Quality (TTD, BGTY, Premium Brands, F = Families (Standard +, Standard, Some Economy), B = Budget (Extended Economy, Tertiary Brands) GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Asda Wal*Mart Spectra Advantage System
Asda WalMart describe all stores by one of four spending bands, Core, Core Plus, Core Plus Plus and Core Constrained, then refines at category level. Spectra system takes panel data (ACNielsen /TNS /GFK) and broadcasts national purchasing patterns through demographic profiles on to store trade areas to describe potential demand by each store
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 19
Asda Wal*Mart Spectra Advantage System Store Clusters defined by opportunity – higher priced wines
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 20
Page: 21
Strategic Customer Segmentation
Can’t stay away 3 monthly high spenders
Convenience
Shopping staff
Healthy Living GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 22
Strategic Customer Segmentation Tesco loyalty card analysis Lifestyles in Tesco
(8 Main Segments) Making Pennies Work
Staple Family Meals
Better off Families
Convenience Cooks
Quick Meals
Shoppers on a budget
Cheap and Easy Meals
High Spending Superstore Families
Cosmopolitan Cooks
Ready Meals Fans
Aspiring Foodies
Standard Superstore Families
Cooking from Jars
Calorie Counters
Stylish Foodies
Kids Choice
Eating for Health
Quiche Meals
16.4%
9.7%
13.0%
0.9%
Substantial Family Fodder
Basic Family Meals
Cost Conscious Cooks
Sausage and spuds families
3.4%
3.3%
4.8%
4.0%
11.8%
11.2%
4.2%
3.0%
4.6%
3.6%
5.2%
2.4%
Biscuits and quick meals
8.7%
3.4%
1.6%
2.0%
Well off Pizza Families
3.3%
1.7%
(Percentage of total number of Clubcard holders) GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Good Cuisine
9.4%
4.5%
2.6%
Good Taste is Green
2.3%
Conservative Quality
Traditional Living
Upmarket & Traditional
Traditional Elderly
First Rate Meals
Old Fashioned Brands
15.9%
4.0%
5.4%
Middle Market Conventionalists
3.8%
13.6%
7.4%
2.5%
Northern Band Loyalists
3.7%
Comfortable but Cautious
2.7%
(27 Sub-segments)
Page: 23
Case Study
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 24
Case Study
• Russian book retailer – Ranges include stationery, toys, music & video • Strong & sustained organic growth • 500 Stores throughout Russia and continuing to grow • Diverse locations • Large range of store sizes • Several ‘Banners’ • Introducing ‘Category Management’ • Implementing major new systems platform
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 25
Concept & benefits of ‘Clustering’ recognised... Different approaches had been tried, but without success
Store brand?
Store size? Store fascia?
Store geography? Store location?
Best practice is to develop a customer profile / shopping occasion based model
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 26
Diverse people, lifestyles & culture how do you profile & group them?
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 27
Shopper based clustering challenges... • Russian market evolving rapidly • Demographic data is difficult to obtain and not granular enough to be useful • Consumer data is patchy and non-existent in book retail channel • Customer profiles are too broad to be applied in this channel • Shopper behaviour understanding in this environment does not exist The only reliable data available was..... Store & Item Level POS Data:
Store Attributes:
Item type Item sales value, volume, history
Location, size, type of locality, adjacencies
Supplemented by observational data... Customer types: Age, single or family, children’s age, affluence
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 28
Analysis of similar stores indicated clear differences in sales profiles Media
Stationery Science & Technology Medicine, Economics, Law Culture & Society Languages & Dictionaries School, Education For child Fiction Home, Leisure, Life
-8
-6
-4
-2
0
Store 1
2
4
Store 2
Total sales values Store 1 = 6.5 million R, Store 2 = 5.8 million R GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
6
8
Analysis of similar stores indicate clear differences in sales profiles
Page: 29
• Same size stores do not deliver the same mix of business • Clear evidence of a bias in store profiles.
Core Range Education bias store cluster
Family bias store cluster
Store 02 has 35% sales in education and sciences Store 01 has 77% sales in Home, fiction, children and stationery
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 30
A detailed analysis of the entire estate identified 6 ‘obvious’ clusters Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster 6
Overall
Actual Sales index.
15
30
55
45
31
89
38
Projected Sales index using cluster 4 as a factor of 1
20
27
40
38
72
72
30
Actual
12
22
41
31
24
52
27
Projected
14
19
29
27
51
51
21
Business Economics Law
Culture And Society
Fiction Actual
39
69
132
98
86
124
82
Projected
43
59
87
82
156
155
65
Home Lifestyle, Leisure Actual
35
60
104
76
63
97
65
Projected
34
47
70
65
125
124
52
Actual
3
6
13
9
7
17
8
Projected
4
6
9
8
16
16
7
Actual
37
63
75
80
64
81
66
Projected
35
48
71
66
126
125
53
Actual
39
75
107
105
81
161
87
Projected
46
63
93
87
166
165
69
Actual
4
8
15
11
9
20
10
Projected
5
7
11
10
19
19
8
Actual
25
49
27
58
56
51
45
Projected
24
33
48
45
86
86
36
Linguistics
Literature for Children
Schools, education and Pedagogics
Science, Technology and Medicine
Toys
Significantly Low Sales
Reduce Space Allocation
Significantly High sales
Increase Space Allocation
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
A detailed analysis of the entire estate identified 6 customer-centric store clusters
1. “Counting the Roubles” Catering to less well off customers buying across all categories on a limited budget in smaller stores outside of major population centres
2. “Children First” Serving and middle income customers mainly buying children’s books and toys in mid-sized town centre and suburban stores
3. “Well Read” Attracting high traffic of high spending customers mainly buying books in larger town centre and suburban locations
6. “Young, better off & Well read”
4. “Middle of the Road” The average store attracting middle-income customers buying across all categories in all types of location
Page: 31
5. “Stationery Stars” Providing an offer for a heavy flow of customers with a strong bias to buying a high number of low value stationery items in town centres and GS1 Baltics Retail Forum 5th November 2008 suburbs © Riverhead Consulting Ltd– 2008
Attracting the highest income, highest spending customers - mainly under 30 years of age, in large numbers, buying across all categories in town centre stores
Page: 32
Cluster comparisons
Descriptor
Sales Profile
Customer Profile Store Profile
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster 6
Counting the Roubles
Children First
Well Read
Middle of the Road
Stationery Stars
Young Better Off & Well Read
• Item sales value is rising • Toy sales lower v cluster 2 • Item sales value higher • Children’s books relatively • Lowest number of item • Value per item rising high sales • Book sales up on cluster 1, • High sales of business, • Value of each item is lowest toys, stationery & culture, fiction, linguistics, of all clusters children's books much science, home & life higher • Stationery sales flat v overall sales • Income profile is lowest of • all groups • Rising income profile • Age profile highest • • Age range & presence of • More households with children similar to cluster 1 children •
• • • •
• High performing cluster Average value of items sold • Highest total item sales of all stores • Stationery sales high but is reverse of cluster 3 lower than cluster 5 • Not the highest value Focus on lower value items • High book sales in every • Category sales of Sales of media, toys & category stationery & toys stationery high outperform all other • Overall value per item Book sales lower than clusters sold is higher than all cluster 3 other clusters • Books are in line with cluster 4
Income profile higher than • Income levels are higher • Income profile similar to • Highest income profile of cluster 1 & 2 than clusters 1 – 4 cluster 3 all categories • Age profile slightly Age range broadly same as • Age range & presence of • More shoppers under 30 younger 1&2 children similar to cluster 3 and fewer with children Less households with • More households with families older children
• Majority of stores are • Size slightly larger than • Sizes similar to cluster 1 smallest cluster 2 • Higher number of visitors • Traffic estimates are lowest • Traffic sharply higher than • Located in centres & of all stores cluster 1 & 2 suburbs, few in rural & • More stores in industrial & • No stores in rural or industrial rural areas industrial areas
• Sizes similar to cluster 3 • Traffic noticeably lower than cluster 3 • Located throughout most areas
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
• Store traffic is rising • Stores located mainly in centres & suburbs
• Highest traffic numbers of all clusters • All stores are in centres
Page: 33
Cluster development... • Clusters were not developed... • ...based on store size • ...using only sales value or volume sales • Clusters were developed... • ...based on item sales mix of categories • ...using customer profile (customers who shopped in the store) • ...store attributes that determine the customer profile
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Customer centric Store Clustering drives benefits across the entire business.. Better understanding of the Value Chain Dynamics
Better understanding of the Market Dynamics
Better understanding of the Customer Dynamics
Factors influencing stores’ performance
Category Strategy
Inventory Management
Stock cover & replenishment planned and managed by cluster
Assortment
Core & discretionary category ranges planned and managed by cluster
Category Plans Space Allocation
Micro & macro category space allocation planned and managed by cluster GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 34
Revenue Management
Supplier Management
Promotional events tailored to cluster-specific requirements
Transparent communication of the implications of the store cluster model
Store assortment by category can be precisely targeted to customer profile For each cluster we can now define…..
Core Range • • • • •
Titles / SKUs Share of category space Position in store Stock levels / target availability Replenishment frequency
Discretionary Range • Based on cluster attributes – Store size – Category participation – Catchment preferences
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Promotions • Participation in promotion • Use of display materials • Position in store
Page: 35
Page: 36
The results can be significant... • Sales uplift in underperforming test stores: +87% • Overall sales uplift: +22% • Availability: +18% • Overall reduction in inventory levels: -17% • Promotional response: +35% • Average spend per visit: + 12%
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Impact on Retailers business model... • Store Clustering enabled the retailer to improve efficiencies across a wide range of measures. • Retailer is now able to discuss ‘Ranging Solutions’ with suppliers on a ‘Cluster’ basis. • Macro & micro space allocation reflects customer demand – optimising stock holding and improving availability • The business has become more ‘Customer Centric’ in its approach and thinking. • Promotions are targeted to drive volume and profit in the stores where impact will be greatest. • Performance measures at store level are focused on ‘customer service’ • Stores are benchmarked ‘like for like’.
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 37
Page: 38
‘Store Cluster’ models should be developed using the best data available to a retailer... ... their own!
Effective ‘Store Cluster’ modelling should not be a ‘black box’ solution... ... it is a combination of high level analytics and retailing expertise.
‘Store Cluster’ modelling is a collaborative process within the retailer and with suppliers... ...the benefits can only be realised by working together .
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Effort, this is. But worth it, effort is. Interesting this may become.
GS1 Baltics Retail Forum 5th November 2008 © Riverhead Consulting Ltd– 2008
Page: 39
Consulting
The Benefits of Store Clustering
Simon Smallwood Director Email –
[email protected] Tel - +44 7786 387793
7 Garrick St Covent Garden London WC2E 9AR T - +44 (0)203 051 1375 www.riverheadconsulting.com