[原创essay] 市场价格与市场份额的关系

最恨苦瓜

文章第N稿,完成度90%(差conclusion)等,盼指教 第一贴连接http://www.21manager.com/dispbbs.asp?boardID=13&ID=32951&page=2 数据文件连接http://www.21manager.com/dispbbs.asp?boardID=32&ID=32958&page=1

Changes in Prices and Its impact of Market Share: A Case Study on Cornflake Sub-market

Abstract

Keywords:

1. Introduction

Despite the increased role of non-price factors in marketing process, price remains an important element in the marketing mix. Companies keep on changing their price strategies and tactics with a view to increasing sales and market share. The immediate impact of price changes on market shares and the potential for strong reactions from consumers and competitors provoked managers’ great concerns. In this article I intend to interpret the effects of price changes on the market share, based on a case from a particular sub-market of cornflakes.

The case is about a particular sub-market of cornflakes (KC250, KC500) and the superstore’s own brands of cornflakes (SMC250, SMC500). In this context it assumes that the market is consists of these four kinds of products (because it is impossible to analyse the market for the breakfast cereal market as a whole), preliminary graph plots show that the price changes of these four products have significant influences on market share. For instance, the price increases of KC500 in week 12 and week 16 (Fig 1.1) led to its slash in market share from then on, the downwards trend lasted until week 34 (Fig 1.3), part of which was brought about by the price cut of its competitor products i.e. SMC250 and SMC500.

2. Literature View

Pricing was ranked the third in overall importance among 15 marketing issues in a survey by Davidson and Stacey (1997), 78% of the respondents considered it as “extremely important”. Mercer’s study (1991) on FCC biscuit market indicates that market shares are primarily determined by price. He used both of multiplicative model and attraction model, assuming elasticities are constant, and came out with similar results. In his study he also indicated other factors like advertising and non-price promotions are also important and should be included in the models. More recently in 1996, a research by Mercer (1996) shows there are non-linear price effects on share vary from brand to brand, depending on where each is positioned in the market, and more importantly, he demonstrated that it is incorrect to assume price elasticities are constant. My approach of analysis will be based on Mercer’s model (1996).

Some price changes are temporary movements around a fairly stable level (e.g. temporary price decrease of KC500 during week 74 and 79), others are permanent if there is no return to the previous level. One type of permanent price changes is evolving prices, if a company engages in a regular practice of discount policies (e.g. price changes of SMC500), another type of permanent price change is structural change in favour of a new level different from the previous level (e.g. permanent price increase of KC250 in week 16). These three types of price changes and their impacts were discussed in detail in Market share response and competitive interaction: The impact of temporary, evolving and structural changes in prices (Shuba Srinivasan, Peter T.L. Popkowski, Frank M. Bass, 2000). The conclusions of this article are shown in table 2.1.

Temporary price changes

1. Only a temporary effect on shares.

2. Short-run response elasticities are greater than long-run elasticities evolving prices.

3. Temporary price changes could be more effective for lower-selling brands than for higher selling brands.

4. Strong competitive reaction consistent with retailer-dominated reactions.

Evolving price changes

1. Significant permanent effects on share.

2. Response elasticities lower than those for temporary price changes.

3. Immediate competitive reactions consistent with retailer-dominated reactions.

4. Long-run price matching behavior price wars.

5. Could be less profitable than structural changes since the latter allow the firm to get to desired market share level sooner.

Structural price changes

1. Permanent effects on share could lead to shift in share level.

2. No change in consumer response coefficients.

3. Share response could be greater than response to temporary changes.

4. Subsequent competitive reaction does affect the prediction of price response and needs to accounted for by managers.

5. A structural decrease in price could be more profitable since firm gets to desired share level sooner.

6. Major competitors respond with a 12 to 13 week lag to an unexpected price break consistent with retailer-dominated reactions.

Table 2.1

Shuba Srinivasan (2000) used VAR (Vector Auto Regression) model with time series in order to display the long-term and short-term changes of market share, which is reasonable in terms of its context. However, Alan (1992) argued that lags should not be included since it would overestimate the influences of adstock and fail to explain the impacts of sudden price drops. Here I am not going to put lags into my model since no specific data available for further research on lags i.e. the customer segmentation and their behavioural responses to cornflake.

[此贴子已经被作者于2004-3-25 9:30:24编辑过]

最恨苦瓜

3. Methodology

Firstly I will conduct a component analysis on market shares of the four products in order to indicate empirically the relationships among them, and then I will try to build a model to illustrate the relationships between market shares and prices statistically. The process of modeling is as follows:

The attraction of a brand depends on its marketing mix. Let Zit be the attraction of product i in period t, then the market share attraction models can be defined as:

Sit = (E 3.1)

Where Sit is the market share of product i in period t, n is the number of brands on the market, the sum of Sit should be equal to 1. Bell et al (1975) demonstrate that the following four axioms are necessary in attraction model:

1. Each brand in the market has a positive attraction (Zit >0)

2. Zero attraction means zero market share

3. Brands which have equal attractions obviously have equal market shares

4. Equal effects on brands

Equation 2.1 represents the overall structure and the attraction function itself remains to be specified. There are different specifications for it according to different context, in a differential-effects version of the well-known Multiplicative Competitive Interaction model (Cooper and Nakanishi, 1988), the attraction Ait is specified as

Ait = exp(ait) X (E 3.2)

Where ai = constant

Xkit = the k-th attribute for brand I (e.g. distribution, prices, advertising)

K = the number of independent variables (the marketing mix)

= the impact coefficient for the k-th marketing mix variable

Here in this case study each of the two brands, namely SMC and KC, has its cereal products in two sizes and more importantly, according to the scanner data, the sum of market shares of all of the 4 products is equal to 1, hence in the modeling I assume that they make up of the “whole” sub-market and they are competitors of each other.

In this case study I specify the function as follows:

Sit = ai exp( )P D A (E 2.3)

Where a i = constant

= 0 before price change c

= 1 after price change c

= the price of brand r in period t

D = Distribution of product i in period t

A = Advertising of product I in period t

By introducing natural logarithm E2.2 can be changed into

Ln Sit = ai + + + eii

+ terms for distribution and advertising (E2.4)

where = proportional change in share of brand i for price change c

= price cross-elasticity of competitor r, own price elasticity when i = r

ei = a disturbance term

E(2.4) was provided by Alan (1996) in his “Non-linear price effect”, however in this case there is no data available for distribution and advertising, I had to eliminate them in model building at the moment.

最恨苦瓜

4. Empirical Analysis and Discussion of Results

According to the SPSS results, this regression models for market shares of the four products don’t fit well (R square values are around 0.3 only, significant level .05). The reasons for it could be various, i.e. distribution and advertising factors are excluded in the model due to the lack of data, the size of dataset is small, and aggregations of data (The scanner data for prices and shares here are of an aggregation of several stores within the superstore group) also influence the accuracy the model. Although including lags of market share in the model can improve R square values by about 0.3, it is not recommended to do so with reasons discussed in literature review.

From the table, none of the market shares of the four products is elastic to the prices of SMC500 and KC250 since both of their coefficients are insignificant. That indicates the price changes of SMC500 and KC250 have no significant influence on the market shares of the four products (including its own market share). On the other hand, he market shares of SMC250, SMC500 and KC250 are elastic to the prices of both SMC250 and KC500 while the market share of KC500 is only influenced by its own price.

The table also shows that Dummy2F (the 6th price change of SMC500 in week 50) has significant impacts on the market shares of the four products. Similarly, Dummy2A, 2D and 4D led to market share changes of SMC500, KC500 and SMC250, -12.2%, -11.4% and 7.5% respectively.

It is worth noting, again, that the results from the model can only reflect the truth with about 30% confidence.

最恨苦瓜

table怎么也贴不上来。。。小小一个表格大小会超过16xxx字节？！

最恨苦瓜

table怎么也贴不上来。。。小小一个表格大小会超过16xxx字节？！

虎子

苦瓜兄啊

能不能来国文啊

e文真的我不行啊

拜托