날씨불쾌감과 쇼핑선호도가 소매점 선택에 미치는 영향 (The Effect of Weather Discomfort and Shopping Preference on Retail Store Visits)

서지정보

· 발행기관 : 한국유통학회
· 수록지 정보 : 유통연구 / 25권 / 1호 / 1 ~ 25페이지
· 저자명 : 이지연, 윤여홍, 최정혜, 정예림

초록

본 연구는 당일 날씨가 소매점 방문 선택에 미치는 영향을 살펴본다. 구체적으로 날씨불쾌감에 초점을 맞추어, 이에 따라 소매점의 대표적 매장 유형인 대형마트와 슈퍼마켓에서 방문 확률이 변화하는 양상을 확인하였다. 또한 계절에 대한 기후 변동이 뚜렷한 한국의 특성을 반영하여 날씨불쾌감의 효과를 계절별로 살펴보았다. 나아가 소비자 개인의 특성인 쇼핑선호도에 따라 날씨 효과가 어떻게 변화하는지 고찰함으로써, 유통 연구와 실무 업계에 유의미한 시사점을 제시하고 있다. 국내 대형 유통업체로부터 확보한 쇼핑 데이터와 날씨 데이터를 통합하여 실증 분석한 결과, 저자들은 세 가지 주요한 결론을 도출하였다. 첫째, 봄·가을철 발생하는 날씨불쾌감은 모든 소매점 방문에 긍정적인 영향을 미친다. 둘째, 여름·겨울철 발생하는 날씨불쾌감은 모든 소매점 방문에 부정적인 영향을 미치지만, 여름철 대형마트에서는 영향력이 유의하지 않다. 셋째, 소비자가 기존에 쇼핑을 선호하는 성향이 높을수록 소매 업종의 종류에 따라 행동 양상이 다르게 나타난다. 즉, 날씨불쾌감이 발생하여도 쇼핑선호도가 높은 사람들은 모든 계절에 대하여 대형마트 방문에 일관되게 긍정적으로 반응하나, 슈퍼마켓 방문에는 변화가 없거나 부정적이다. 이와 같은 연구 결과를 바탕으로 저자들은 학문적·실무적 시사점을 논의하였다.

영어초록

Companies have a keen interest in taking advantage of weather to improve business performance. Previous studies have examined outcomes in a single store or one retail brand (Arunraj and Ahrens 2016; Parsons 2001) or focused on one dimension of weather among a possible multitude including temperature, sunlight and rainfall (Moon et al. 2018; Zwebner et al. 2014). However, little is known about complex weather effects, especially in a setting of multiple retail brands. This is where we aim to contribute. This research investigates the effect of weather in the context of grocery shopping. We focus on weather discomfort or unpleasant weather conditions from the perspective of grocery shoppers and examine how the weather discomfort affects their store visits when shopping for grocery. To be specific, we attempt to obtain answers to the following questions. First, does weather discomfort affect grocery store visits (megastore vs. supermarket)? Second, how does this weather effect vary by season? Third, to what extent does individuals’ shopping preference moderate the weather effect? To this end, we study related literature, propose the hypotheses and then empirically test them.
Figure 1 shows the proposed model. As mentioned above, we focus on the two types of grocery stores, megastore and supermarket. To understand store visits by grocer shoppers, we first focus on weather discomfort. This weather discomfort is designed to reflect weather conditions outside comfort zones or beyond bearable boundaries. The season-specific variables of weather discomfort are constructed using the multiplex measure of the NET index (Hentschel 1986). Spring and fall are grouped together while summer is combined with winter due to their pairwise similarity of climate mildness or severity. Next, we examine the moderation effect of shopping preference. Shopping preference is measured by counting the lagged number of store visits within the 2.5 week data window. As such, we can examine in what way season-specific weather effects are moderated by individuals’ shopping preferences when explaining store visits to megastores and supermarkets.
H1: In spring and fall, weather discomfort positively affects visits to megastores and supermarkets.
H2: In summer and winter, weather discomfort negatively affects visits to megastores and supermarkets.
H3: In spring and fall, shopping preference moderates the positive effect of weather discomfort.
H3-(1): The positive effect of weather on megastore visits becomes stronger as shopping preference gets greater.
H3-(2): The positive effect of weather on supermarket visits becomes weaker as shopping preference gets greater.
H4: In summer and winter, shopping preference moderates the negative effect of weather discomfort.
H4-(1): The negative effect of weather on megastore visits becomes weaker as shopping preference gets greater.
H4-(2): The negative effect of weather on supermarket visits becomes stronger as shopping preference gets greater.
We obtain store visit data of 2,400 shoppers residing in Seoul in 2015 from one of the top three retail stores. To reiterate, store visits consist of megastore and supermarket while non-shopping serves as a baseline. We merge this store visit data with the following additional datasets. We obtain daily weather data from the open portal of Korea Meteorological Administration and construct the variable measuring weather discomfort (WD). The season-specific variables of weather discomfort are constructed using values deviated from the median of the NET index or the weather stress index. We also collect public data available from the Korean government to control for local environments and rule out alternative explanations. In particular, we obtain the total numbers of megastores and supermarkets in each region (at the sigungu level) along with area size. As such, a variety of control variables includes local retail characteristics (overall densities of megastores and supermarkets), demographics (age and gender), and daily weather conditions (rain and snow).
Multinomial logistic regression is employed to estimate the probability of consumer i visiting store j (=1 for megastore; =2 for supermarket) on day d in season s while taking ‘no visit’ as a baseline. Model 1 focuses on the weather effect by season whereas Model 2 extends Model 1 by adding the moderation effect of shopping preference (SP). Note that both empirical models include time fixed effects by month and day of week.
(1) Model 1: (2) Model 2: (3) Table 1 shows the estimation results of Model 1 and Model 2. The empirical findings are as follows. First, weather discomfort in spring and fall helps consumers go grocery shopping whereas that in summer and winter hampers shopping trips. These weather effects stay the same for both megastores and supermarkets, thus supporting H1 and H2. Second, when visiting megastores, shopping preference moderates the weather effect by increasing (lowering) its influence in spring and fall (summer and winter), suggesting the varying sensitivity to the weather effect across individuals when visiting megastores. That is, as shopping preference gets greater, the positive effect of weather in spring and fall becomes stronger and the negative effect in summer and winter becomes weaker. These thus support H3-(1) and H4-(1). Third, as per supermarket visits, shopping preference moderates the weather effect by intensifying the negative effect of weather only in winter, moderately supporting H4-(2). Taken all together, our empirical findings in general support the proposed hypotheses.
This study empirically demonstrates that weather plays a key role in understanding retail store visits and this weather effect varies by season. Furthermore, individual-level shopping preference can both help and hamper this main effect of weather depending on seasons and store types. Thus, our empirical findings contribute to the field of offline retailing by examining the role of weather in driving store visits as well as the additional effect by individuals’ shopping preference. In today’s era, the retail environment continues to evolve rapidly with new channels and unprecedented formats. As such, this sort of disaggregate level analyses helps both academia and practitioners deepen related understanding and employ strategies accordingly above and beyond studies using aggregate sales data. Further this study offers managerial insights into weather-based promotions to drive store traffic and improve sales performance. The positive effect of weather needs to be taken advantage of and the negative effect should be mitigated by targeting consumers of varied shopping preferences.

참고자료

· 없음