dominick’s retail analysis

14
Retail Demographics Analysis Eeshan Srivastava Balaji Vanjinathan Jia Xie

Upload: eeshan-srivastava

Post on 14-Apr-2017

149 views

Category:

Marketing


0 download

TRANSCRIPT

Page 1: Dominick’s retail analysis

Retail Demographics AnalysisEeshan SrivastavaBalaji VanjinathanJia Xie

Page 2: Dominick’s retail analysis

2

Agenda• Background• Problem Statement• Methodology• Results and Analysis• Recommendations• Questions

Page 3: Dominick’s retail analysis

3

Background• Dominick's was a Chicago-area grocery store chain and subsidiary of

Safeway Inc.• Closed operations in Dec, 2013 after a series of failed strategic

initiatives and high competition

• Data from 1989 – 1994 when Dominick’s was a regional leader• 9 years of store-level data on the sales of more than 3,500 UPCs• Collected by The Kilts Center for Marketing at Chicago Booth and

Nielsen

Page 4: Dominick’s retail analysis

4

Problem Statement• To identify the demographic makeup of the market in terms of

store clusters, discover sales patterns and recommend a targeted positioning strategy

• What’s in it for Dominick’s? What are the characteristics of customers visiting each store? Which stores attract higher sales in which categories? Where are the store clusters located geographically?

Page 5: Dominick’s retail analysis

Methodology

5

Identify relevant data• Store-level sales and

traffic data ~ 300K transactions

• Store demographics data ~ 100+ stores

Cleanup data• Remove

junk• Remove

stores with very few transactions

Preliminary testing of assumption• Discover variance in

sales across stores for at least one product

K-Means Segmentation• Identify segmenting

variables• Perform clustering and

discover correct # of clusters

Generate Results• Merge cluster data with

sales data• Collect sales data per

cluster

Page 6: Dominick’s retail analysis

6

Preliminary Test• Assumption: there is a variance in sales across different stores in at

least one product• Test Product: Beer

2 6 12 19 32 44 48 52 56 62 68 72 76 81 88 92 97 102

106

110

114

118

123

129

133

137

142

301

305

309

313

318

$(500.00)

$-

$500.00

$1,000.00

$1,500.00

$2,000.00

$2,500.00

$3,000.00

Avg Beer Sales by Store Number

Page 7: Dominick’s retail analysis

7

K-Means Segmentation

Ethnicity – % of Hispanics/Blacks

Household size – average number of members in

the family

Household Value – average

house value in the area

Income – average income

of the neighborhood

Identified the following variables which could sufficiently differentiate clusters from one another

Education – % of college graduates

Page 8: Dominick’s retail analysis

8

K-Means Segmentation

1 2 3 4 5 6 7 80

0.05

0.1

0.15

0.2

0.25

0.3

K-Means Elbow Chart

Cluster # Income Education Ethnicit

yAvg Family

SizeMean House

ValueAvg Beer

SalesAvg Cust

Count# of

stores

1 $ 38,991 20.2% 12.6% 2.66 $ 140,090.08 $

615.27 2535 32

2 $ 58,657 44.4% 7.5% 2.52 $ 246,393.53 $

626.14 2762 6

3 $ 47,731 29.6% 7.6% 2.62 $ 181,861.19 $

613.88 2621 25

4 $ 33,438 11.9% 30.1% 2.77 $ 91,575.16 $

715.09 3048 22

We decided to pick 4 as our number of clusters to avoid having clusters with too few number of stores

Page 9: Dominick’s retail analysis

9

Analysis

1 2 3 4 $550

$600

$650

$700

$750

Avg $ Sales of Beer/day

1 2 3 4 $-

$200 $400 $600 $800

$1,000 $1,200

Avg $ Sales of Fish/day

1 2 3 4 $-

$1,000 $2,000 $3,000 $4,000 $5,000 $6,000

Avg $ Sales of Dairy/day

1 2 3 4 $-

$1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000

Avg $ Sales of Meat/day

Page 10: Dominick’s retail analysis

10

Store segments on the map

Page 11: Dominick’s retail analysis

11

Recommendation

Identification of market segment

Targeting right segments for maximized

profits

Product Positioning/Place

ment

Page 12: Dominick’s retail analysis

12

Limitations• Need to avoid the stereotyping pitfall

• Needs to be coupled with other types of segmentation – Psychographics and Purchase Behavior

• Hypothesis testing could be done for each segmenting variable before performing segmentation

• Costs and profitability data should also be analyzed across store segments

Page 13: Dominick’s retail analysis

Questions?

Page 14: Dominick’s retail analysis

Thank You!