Demand Estimation and Optimization Approaches

Transcription

1 Retail Assortment Planning: Demand Estimation and Optimization Approaches Gürhan Kök The Fuqua School of Business Duke University CERET 2011

2 Outline Industry Approach Best Buy Borders Albert Heijn (Ahold) Tanishq, India (Kok and Fisher and Vaidyanathan, 2008, Book chapter) Academic Approach Kok and Fisher, 2007, Operations Research Fisher and Vaidyanathan, 2007, Working paper Kok and Xu, 2011, Management Science Bernstein, Kok, and Xie, 2010, Working paper Gurhan Kok, Retail Assortment Planning 2

3 Definition of an Assortment Category examples Men s dress shirts, Sunglasses, DVD players, Cereals Assortment at category level Breadth: Number of categories Depth: size of assortment in each category Assortment at product level A selection of product variants within a category Products are differentiated by some attributes Products are potential substitutes Number of SKUs or facings is limited by category shelf space Gurhan Kok, Retail Assortment Planning 3

4 Assortment Planning at Best Buy Hierarchical planning Focus on buying at chain level Deployment to stores (based on store clusters, climate zone etc.) Dynamic environment Identifying growth opportunities (digital vs. traditional camera) New products Short life cycle (Newsvendor kind of addresses that) Markdown pricing Not every category is the same Identifying critical resources (Promo, labor, impulse, price, selection) ect for different e categorieses Gurhan Kok, Retail Assortment Planning 4

5 Critical to Align Assortment with Company Strategy Primary Attribute Price Service Product Experience Access Price Attribu ute Sec condary Service Product Experience Access Gurhan Kok, Retail Assortment Planning 5

6 and Allocate Resources based on Marginal Returns Category Promo Labor Impulse Price Selection Computer High High Low High Medium Refrigerator Medium High Low Medium High Accessories Low Low High Low Low Movies High Med High High High Gurhan Kok, Retail Assortment Planning 6

7 Identify Growth Opportunities Gurhan Kok, Retail Assortment Planning 7

8 Assortment Localization Gurhan Kok, Retail Assortment Planning 8

9 Assortment Planning at Borders Basic Premise Except for best sellers, customer is interested NOT in title BUT category 300, titles grouped into 300 categories Category popularity assessed by computing Relative Sales Per Title Category Sales RST No.of Titles Shelf space periodically reassigned from low RST to high RST Following the principle of Darwin s Natural Selection and Survival of the Fittest, categories fight for shelf space Store managers allowed to pick titles to be stocked within each category, thereby decentralizing a part of the decision process Gurhan Kok, Retail Assortment Planning 9

10 Shelf Space Allocation at Albert Heijn Frank Jensen s method Category 1 Category sales curves are estimated by using data from multiple stores Gross Pro ofit Category 2 Category 3 Metric shelf space allocated to category j For each store: Allocate meters of space with Greedy method until the store runs out of space Gurhan Kok, Retail Assortment Planning 10

11 Assortment Planning at Tanishq No.1 Jewelry retailer in India 30,000 SKUs, 15 categories, 52 retail outlets Diverse demand patterns across stores Assortment planning decisions are Hierarchical at higher attribute levels (category, theme, design) Decentralized for lower attribute levels (models, size), to provide for local customization of stock mix For example, each store to carry 30% of national best sellers, 20% of regional best sellers, 10% of store best sellers, and rest decided at store level Gurhan Kok, Retail Assortment Planning 11

12 12

13 Localized Assortment Research at Albert Heijn Localized Assortment Research at Albert Heijn Kok and Fisher (2007)

14 Consumer Response to Unavailability of a Product Gurhan Kok, Retail Assortment Planning 14

15 Substitution Sales response to more choices depends on substitution % 100% substitution Sales 50% substitution 0% substitution Number of items available to consumer on given day. Assume items have equal demand Gurhan Kok, Retail Assortment Planning 15

16 Local Assortment Research with Albert Heijn n j = number of facings assigned to SKU j. w j = width of a facing for SKU j Max j GP j (n j ) s.t. j w j n j Available Shelf Space Gross profit of product j with facing assignment n j n j 0 and integral Deployment challenges Estimating demand for SKUs not currently carried in the store Substitution Impact of stock outs on product gross profit Maximizing j GP j (n) given interconnectedness stock out of product A increases demand for B Gurhan Kok, Retail Assortment Planning 16

17 Local Assortment Optimization SKU 1 GP increase with more facings due to increased inventory (lower lost sales Gross Prof fit SKU 2 SKU 3 Number of facings allocated to each SKU (inventory level) KEY TRADEOFF: Breadth (number of items) vs Service Levels (max InventoryLevels) Gurhan Kok, Retail Assortment Planning 17

18 Substitution Matrix Examples Gurhan Kok, Retail Assortment Planning 18

19 Estimating Assortment-Based Substitution: Carried in store Not carried Observed demand D 1 D 2 D D k Estimate using store specific models Substitution True demand d d 1 d 2 d d k d k d J Estimate using national models on full assortment stores Substitution probability from product i to j: ij d j d l l N {i} Gurhan Kok, Retail Assortment Planning 19

20 Observed Demand vs. Full Assortment Demand observed demand full assortment demand demand d rate When a store carries less than full assortment, observed demands are higher than full assortment demand due to substitution store 1161 (segment 3 - deepfry) Gurhan Kok, Retail Assortment Planning 20

21 Estimated Substitution Rates Week = 91 segments SR Error reduction % % 3 1 1% % % % % % % % % % SR Week = 38 Error reduction 0 0% 0.3 0% 1 1% % % % 1 19% 0 0% 0.8 0% 1 30% 1 7% 1 35% Gurhan Kok, Retail Assortment Planning 21

22 Results: Impact on Gross Profit Gurhan Kok, Retail Assortment Planning 22

23 Impact of Substitution Gurhan Kok, Retail Assortment Planning 23

24 Sources of Improvement: Example Gurhan Kok, Retail Assortment Planning 24

25 More Improvement When Shelf Space is Tight 25% 20% group 3 group 55 Linear (group 55) Linear (group 3) nt % improveme 15% 10% 5% R 2 = R 2 = % Shelf space (more space corresponds to larger stores) Gurhan Kok, Retail Assortment Planning 25

26 Attribute-Based Substitution Estimation Attribute-Based Substitution Estimation Vaidyanathan and Fisher (2007)

27 Identify Relevant Product Attributes sub-category thread count fabric price color weave size Attribute-based view is useful for Understanding substitution behavior Choosing localization objectives Estimating demand for new products (Fader and Hardie 1996) Gurhan Kok, Retail Assortment Planning 27

28 Snack Cakes: Estimate Sales Potential of all SKUs and Substitution Rates Gurhan Kok, Retail Assortment Planning 28

29 Snack Cakes: Estimated Substitution Rates Gurhan Kok, Retail Assortment Planning 29

30 Assortment Changes for one Store Gurhan Kok, Retail Assortment Planning 30

31 Tires: Sales Data from one Store Gurhan Kok, Retail Assortment Planning 31

32 Tires: Substitution Estimates Gurhan Kok, Retail Assortment Planning 32

33 Impact of Localization 6% 32% Current assortment Gurhan Kok, Retail Assortment Planning 33

34 Hierarchical Choice Models Hierarchical Choice Models Kok and Xu (2011)

35 Subgroups Based on Product Type Gurhan Kok, Retail Assortment Planning 35

36 Subgroups Based on Brand Gurhan Kok, Retail Assortment Planning 36

37 Assortment Planning and Pricing Regimes Pepsi Brand Manager Coke Brand Manager Category Manager Gurhan Kok, Retail Assortment Planning 37

38 Brand-Primary Model vs. Type-Primary Model Product types are not Product types are functionally functionally differentiated or differentiated brand loyalty is strong Regular vs. diet drinks, Dress shirts with different colors Ice creams with different flavors Detergents with different scents Regular vs. decaf coffee, Sedan vs. mini-van, Point-and-shoot vs. SLR camera (Kannan and Wright 1991, Urban et al. 1984, Grover and Dillon 1985, Allenby 1989) Gurhan Kok, Retail Assortment Planning 38

39 Type-Primary Model: Centralized Management Product types Brand X X X X Brand Y Y Y Y Y Can be optimal Brand X X X X X Brand Y Y Y Y Y Y Cannot be optimal Gurhan Kok, Retail Assortment Planning 39

40 Dynamic Assortment Customization in Online Retailing Bernstein, Kok and Xie (2011)

41 Example: Online Retailer Gurhan Kok, Retail Assortment Planning 41

42 Example: Online Retailer Gurhan Kok, Retail Assortment Planning 42

43 Example: Online Retailer Gurhan Kok, Retail Assortment Planning 43

44 Characterizing the Policy y 2 10 Product 1 9 not 8 offered to segment y 1 *(y 2 ) y*1(y2) y*2(y1) y 2 *(y 1 ) Full assortment offered to all segments 3 Product 2 not offered to 2 segment y 1 Gurhan Kok, Retail Assortment Planning 44

45 Broader Questions on Customization Customer behavior Impact of sort order on purchasing behavior Can we refine estimates based on clicks during a search? Revenue maximization and demand management Customization ti based on customer characteristics ti Customization based on margins Customization based on inventory levels or supply ppy outlook Gurhan Kok, Retail Assortment Planning 45

46 Thank you! Papers are available at my web site. Please send comments and stories to Gurhan Kok, Retail Assortment Planning 46