Assessing Grocery Store Attraction via Cross-Shopping Linkages – Methodology

Assessing Grocery Store Attraction via Cross-Shopping Linkages - MethodologyData were collected from a random sample of grocery shoppers in Charlotte, NC. A mail survey was sent to the respondents along with an offer to participate in an opportunity to win fifty dollars. For the present study one thousand surveys were mailed and 169 (16.9 %) usable surveys were returned.
Demographics were gathered and include gender, age, number of people living at home, approximate income, education level, and race (see Table 1). Respondents tended to be female (65%). Average age is 45 years old. On average respondents reported two or three people living at home, including themselves. Approximately 75% of respondents earned incomes less than $60,000, 44% had some college education, and the majority of respondents classified themselves as Caucasian (86%). Approximately 64% of respondents identified a second store that they shop at least once per month. These respondents formed the basis for the cross-shopping analysis. The questionnaire consisted of twenty-four items related to service quality and eight items related to consumers’ perceptions of private label brand quality and value. A review of the service quality and private label literature provided the basis for items used in the study. Respondents were asked to think about “the grocery store they shop at most often” when completing the survey. A seven-point likert scale (i.e., strongly disagree/strongly agree) accompanied each statement. At the end of the survey, respondents were asked to identify their primary and secondary grocery stores and their frequency of shopping.
Cronbach’s alpha was used to measure the construct reliabilities. Results of the analysis indicate that all of the constructs were reliable. The coefficient alpha’s ranged from.83 for service reliability to.91 for tangibility and private label quality. See Appendix A for examples of survey questions, number of items per construct, and reliability measures.
The survey data will provide a number of useful insights related to cross-shopping patterns within the grocery store market. The degree to which the stores are linked to each other store will be determined by calculating the proportion of a store’s shoppers who visited at least one other store. For example, if store A has 100 shoppers and 80 of its shoppers visited another store (e.g., Stores B, C, or D) then we could say that 80% of store A shoppers are cross-shoppers. In order to assess intra-type competition (i.e., cross-shopping linkages), the proportion of Store A shoppers who also shopped at each of the other stores will be computed. For example, 20% of Store A shoppers who identified Store A as their primary store may have shopped at Store B, 30% of Store A shoppers may have shopped at Store C, and 50% of Store A shoppers may have shopped at Store D. The latter information when compiled for all stores will produce a cross-shopping matrix which displays the strength of the cross-shopping linkages between all possible pairs of stores in the center.
Cross-shopping patterns are also examined to determine if there are clusters of stores with strong cross-shopping linkages. Each store’s status in the cross-shopping linkages can be further revealed by computing both its attraction (i.e., receiving, pulling) and sharing (i.e., sending, pushing) percentages and the resulting attraction/sharing ratio. A store’s attraction percentage is the mean percentage of primary shoppers that come from other stores. For example, if 50% of shoppers whose primary store was B shopped at Store A, 30% of shoppers whose primary store was C shopped at Store A, and 20% of shoppers whose primary store was D shopped at Store A, then Store A’s attraction percentage would be 33.3% (i.e., (50+30+20)/3). This number is a measure of the importance of the store in generating shopping linkages. A store’s sharing percentage is the mean percentage of shoppers that shop at the other grocery stores and is an indicator of dependency between stores. For example, if 25% of shoppers whose primary store was A shopped at Store B, 20% of shoppers whose primary store was A shopped at Store C, and 15% of shoppers whose primary store was A shopped at Store D, then Store A’s sharing ratio would be 20% (i.e., (25+20+15/3). By creating an attraction/sharing ratio it is possible to examine the attraction level of a store relative to other stores in the market. Finally, the effects of service quality and the perceived quality and value of private label brands on the strength of a store’s cross-shopping attribute will be addressed using multivariate analysis of variance.

Table 1. Respondents’ Profile

N %
Gender
Female 110 65.1
Male 59 34.9
169 100.0
Household size
1 39 23.2
2 76 45.2
3 16 9.5
4 24 14.3
5+ 13 7.7
168 100.0
Income
$20,000 or less 21 13.2
$20,001 to $40,000 47 29.6
$40,001 to $60,000 52 32.7
$60,001 to $80,000 24 15.1
$80,001 or more 15 9.4
159 100.0
Education
Less than high school 2 1.2
Completed high school 18 10.7
Some college 53 31.5
Completed college 60 35.7
More than college 35 20.8
168 100.0
Race
African American 19 12.4
White/Caucasian 134 87.6
153 100.0

Tags: , ,