단계적 매출액모형을 이용한 전단지 효과분석 및 최적 전단지 게재주기 도출 (Modeling the Impact of Feature Advertising on Sales to Derive the Optimal Advertising Interval)

서지정보

· 발행기관 : 한국마케팅학회
· 수록지 정보 : 마케팅연구 / 22권 / 1호 / 59 ~ 80페이지
· 저자명 : 김동훈, 이승연

초록

다양한 업계에서 전단광고에 지출하는 마케팅비용이 증가하면서, 전단광고의 효과에 대한 관심이 고조되고 있다. 이
런 현실에 발맞추어 전단광고의 효과를 설명하는 매출액모형을 형성하여, 효율적인 광고비용 지출을 위한 지침을 제
시하고자 한다. 본 연구는 전단광고가 브랜드매출에 긍정적인 영향을 미친다는 점과 전단광고 게재주기에 의해 그
영향력이 조절된다는 점을 고려하여, 매출을 극대화할 수 있는 최적전단게재주기를 도출하는 것이 목적이다. 구체적
으로, 매출액에 미치는 전단지 게재효과를 판매확률 증대효과와 조건부 판매액 증대효과로 분해하여, 2단계로 구성
된 매출액 모형을 도출하였다. 국내의 한 유명 백화점의 두 지점에 입점해 있는 여성의류 브랜드들을 대상으로 모형
을 분석한 결과, 각 지점의 전체 여성의류 브랜드 중에서 각각 76.1%와 70.7%에 해당하는 브랜드들이 판매확률과
조건부 판매액 두 측면 모두에서 긍정적인 효과를 누리고 있는 것으로 나타났다. 한편, 모형에서 도출된 최적전단게
재주기 전략을 따르게 되면, 현재의 주기로 시행한 경우에 비해 브랜드의 연간매출액이 두 지점에서 평균적으로 각
각 16.0%와 17.6% 증가하는 것으로 나타났다. 본 연구에서 제시한 매출액 모형은 두 가지 측면에서 전단광고 효과
를 설명할 뿐 아니라, 브랜드 전단광고를 계획하는 마케팅관리자에게 효율적인 비용지출을 위한 전단게재주기의 결
정 기준을 동시에 제시하고 있다.

영어초록

While advertising in general leads to long term franchise-building benefits and sales promotion leads to a more
short term and immediate sales response, feature advertising(sales flyers) is a marketing communication medium
that integrates the benefits of the two (Burton, 1999). As such, feature advertising has been used extensively in
Korea by a wide variety of retail businesses such as department stores, discount stores, and supermarkets. It has
been reported that over 50% of Korean department store industry’s total communication budget has been
allocated to feature advertising and the total feature advertising spending in Korea amounted to 250 billion won
(approximately US$ 300 million) in 2004.
The effects of feature advertising on sales have been studied widely in marketing. Feature advertising has been
found to increase sales through store switching(Kumar and Leone, 1988), brand switching(Guadagni and Little,
1983; Gupta, 1988; Lemon and Nowlis, 2002), purchase quantity increases(Neslin et al, 1985; Ailawadi and
Neslin, 1998; Gupta, 1988), and shortening of inter-purchase times(Neslin et al, 1985 Gupta, 1988). On the other
hand, research have also shown that the effects of feature advertising diminish with repeated exposure(Tellis,
1988; Nordheilm, 2002; Campbell et al, 2003).
Based on these findings, this paper presents a brand level model that describes and measures the impact of
feature advertising on sales and then uses the model to derive the optimal average interval of feature advertising
that maximizes sales. The impact of feature advertising is modeled as a two stage process where the feature
advertising first increases the probability of purchase and then increases the sales amount given purchase. The
logic of the model can be presented in the following diagram. In essence, the unconditional sales model that
measures the impact of feature advertising was specified as the product of two equations where one represents the
probability of sales incidence and the other, sales amount given sales incidence. Three variables related to feature
60 마케팅 연구 M a r c h . 2 0 0 7
Brand appears in feature advertising
Total sales impact of feature advertising
=Sales probability increase Sales amount increase
Increase in sales likelihood
Sales Probability Model
Increase in sales amount
(Given sales)
Conditional Sales Amount Model

advertising were included in the model: (1) Insertjt: whether brand j appeared in the feature advertising at time t
(0-1 variable), (2) Sincejt: number of valid days since brand j appeared in the feature advertising at time t (the offer
contained in the feature advertising is valid for three days and hence the Since variable takes on the values 1, 2, or
3 if Insert=1 and 0 otherwise), and (3) INTjt: number of days since brand j last appeared on a feature advertising.
In the first stage of the analysis, this combined model was calibrated using women’s clothing sales data from
two stores of a major Korean department store chain. The data contained one year’s worth of daily purchases by
2,000 selected customers for 227 brands of women’s clothing. Out of the 227 brands, 142 were chosen for
analysis on the basis that they were featured at least 5 times during the one year period. The model parameters
were estimated separately for each of the brands. The results showed that 76.1% and 70.7% of the brands in the
two stores respectively experience a statistically significant increase in both sales probability as well as sales
amount due to feature advertising.
As the second stage in the analysis, the combined sales model and the estimated brand-level parameters were
used to derive an equation representing expected total yearly brand sales as a function of the INT (feature
advertising interval) variable. Specifically, the yearly brand sales is specified as the sum of the expected sales
during the period in which the brand is featured and the expected sales during the period in which the brand is not
featured.
The expected sales during featured period is specified as follows. First, the average number of times a brand(j)
is featured in a year is 365/INTj. Since each time a brand is featured, the content is valid for 3 days, the
expected(predicted) sales during the 3 valid days is summed and in turn multiplied by 365/INTj. This yields the
total brand sales over the featured period. The expected sales during the time the brand is not featured is the
expected(predicted) sales for non-featured days multiplied by the number of days the brand is not featured.
The yearly expected brand sales equation was then maximized with respect to INT to yield the optimum feature
interval for each of the brands. The optimization results indicate that when all brands follow the optimal feature
advertising schedule, the brand sales will increase on average by 16.0% and 17.6% respectively in the two stores.
The results also showed that the optimal interval was longer than the current average interval for 24% of the
brands. This suggests that for these brands, reducing the frequency of feature would lead to higher sales. The
optimal interval for the remaining brands(76%) was either same or shorter than the actual. The paper concludes by
discussing the managerial implications for managers making feature advertising decisions.

참고자료

· 없음