The main objective of this paper is to analyze the heterogeneity of distribution intensity on 3C products and find out its major determinants in the context of emerging China. We collect four 3C benchmark brands in China, Nokia, Haier, Lenovo, and HP, and try to describe their channel marketing strategies, especially in concentration and penetration of their channels distribution. Using a variety of model evaluation strategies, we conclude that NBD regression does indeed provide a better fit of distribution intensity data than the standard Poisson or NBD model. Particularly, our results suggest the estimated parameters from inference in models as convenient descriptors of the nature of distribution intensity that are easy to operate in a meaningful way. As a scientific approach to marketing involves gathering observations about marketing phenomena and then attempting to develop generalized explanatory statements (Bass, 1993), this study takes a scientific approach to the investigation of the distribution intensity for four categories of 3C products in different product life cycle.
In addition, in the current marketing knowledge, CDI/BDI matrix is considered as a marketing investigation tool for investors to evaluate market potential and to help make marketing resources allocation decision. Current marketing knowledge is concerned about the place with high CDI and high BDI, which is more likely the first-tier city to enter into and to be allocated most resources, like resellers. Followed by the second-tier market, which is the one with high CDI and low BDI, and then the market with low CDI and high BDI is the third tier market. Finally, when the market’s CDI and BDI are both low, this market will be the worst place to enter into and investors should put the least resource in this market (Belch and Belch, 2004).
However, CDI/BDI matrix will be quite different when talking about emerging market where the economic status is not stable, especially in emerging China.
Because 3C distribution intensity in China is quite different from the common CDI/BDI matrix in developed country markets (Belch and Belch, 2004), BDI>1 and CDI<1 will be the first candidate to implement market entry strategy. In our results, BDI is positively correlated with distribution intensity in BDI>1 and CDI<1, because investors there will pay more attention to distribution capabilities. We also find that two of these brands, Nokia and Haier, choose to allocate most of their marketing channel resource in the market with high BDI and low CDI (city group C), and the others choose to enter into the market with both high BDI and CDI (city group A).
There are some reasons that cause this difference. We insist that the major reason should be based on product life cycle. Mobile phone and TV are already in the saturation or mature stage, in the cases of Nokia and Haier, entering into the markets where categories are still not well developed, but brands can be popular. However, PC and printer are still in the growth or introduction stage, so Lenovo and HP still put most of distribution intensity in the market with high CDI and BDI because this kind of markets is not saturated. The product life cycle might not be proper since the Chinese emergent market is still growing, and our research data do not have time frame. Another explanation might replace PLC with market penetration levels from low to high.
In reality, these four products look like following a product life cycle from the statistics of these product owners in China of the report of International Telecommunication Union (2003). Besides, the linkage of relative concept of PLC in a generalized manner will help set priority to these products in the channel markets. In brief, the success of channel establishment in emerging markets has always depended not just on the market size according to significant gross domestic product or specific
industrial development, but on the reach of brand’s market penetration, and on distributors’ capability to extend opportunity to gain more profits.
Limitations and Future Research Directions
In this study, we collect four leading brands’ distribution intensity data, which is not easy to obtain for marketing research. However, because we only get data for 2002, we cannot see the dynamic activities of these brands. If the longitudinal data could be obtained, it could be used to observe the patterns of channel market entry of these brands, and it would be useful to understand the spatial diffusion in China’s 3C products market. A cross sectional data set also has its limitations, as does the collection period of one year, which does not allow for observed/theoretical comparisons over different periods of observation. This paper has taken a single time period of observation and used the model to extrapolate to longer periods. There is no empirical verification of this extrapolation. Further research could include observed measures for other time periods of observation.
Theoretical Contributions
But in this study, we suggest that the two parameters, α and β, of fitting the category’s NBD, the coefficient of variation, Gini index, and Pareto Shares are the concentration statistics based on theoretical distribution of unobservable or latent true level market concentration structure. Schmittlein, Cooper, and Morrison (1993) demonstrated that the NBD could be extrapolated to estimate the market behavior for periods beyond the observed period. In addition, the increase of Pareto Share based on NBD over longer time periods of observation is predictable due to only shape parameter α, and theoretically well substantiated. Managers need to note this in their assessment of their channel market concentrations. From the result, it seems that the
“Pareto Share” is not a clear 80/20 relationship but that is actually found among all
different 3C categories to a somewhat difference. Shape parameter α can be regarded as a simple concentration measure of how dependent a brand (in the case of 3C categories) is to their distribution of channel markets in the long run.
We organize our findings of 3C market structure in China around the notion that firms can improve brand perceptions through channel investments and can result in concentration patterns, as in Sutton’s endogenous sunk cost theory. In the data, observed distribution intensity escalates in larger markets across product categories.
Furthermore, a new product case, such as printer for HP, is limited to entry some markets; its distribution intensity concentration levels will be bounded away from zero. In contrast, concentration levels in distribution-intensive brand, such as TV for Haier, are bounded away from zero as well. In general, our findings highlight several persistent patterns in the 3C market structures in China.
From the dataset, the scale of each reseller cannot be revealed, and one distributor might also have its own resellers. We only can observe the quantity of number of channel members, but cannot understand the quality. Therefore, by devising the BDI divided by the number of intermediaries, such measure will be a proxy variable for representing distribution capabilities. The ratio indicates their management capability on that particular city. The reasons why the ratio is high might be due to all kinds of possibilities, including the city’s buying power, good product, capable retail partnership, good price, and less competition, and etc.
Conclusion
We have demonstrated penetration, concentration, market growth, and distribution capabilities in a meaningful way in China 3C market, and considered the likely patterns of these channels as model parameters vary. As such this paper constitutes more on a replication of Scmittlein, Cooper, and Morrison’s (1993) work
in a different context. We have introduced the terminology of “Pareto Share” and defined it as the percentage of distribution intensity establishment made to the top 20% of channel markets, and demonstrated how it may serve as a guide to a brand’s dependence to its channel market. The results of different Pareto Shares across 3C products show that Pareto Share can vary from category to category. A deeper understanding of these parameters is likely to give the brand insight as to how to grow the brand, and maybe give brand managers less cause to panic when they look at their sales in channels panel data.
References
[1] Anschuetz, N., "Profiting from the 80-20 rule of thumb", Journal of Advertising Research, 37 (6): 51, 1997.
[2] Atkinson, Kendall E., An Introduction to Numerical Analysis (2nd edition ed.), John Wiley & Sons, 1989.
[3] Bass, Frank M., “A New Product Growth Model for Consumer Durables”, Management Science, 15, 215-217, 1969.
[4] Bass, Frank M., "The Future of Research in Marketing: Marketing Science", Journal of Marketing Research, 30(1), 1-6, 1993.
[5] Belch, G. E., Belch, M. A., Advertising and Promotion: An Integrated Marketing Communications Perspective (6 ed.). McGraw-Hill/Irwin, 2004.
[6] Bonoma, Thomas V. and Kosnik, Thomas J., Marketing Management: Text and Cases, Homewood, IL: Richard D. Irwin., 1990.
[7] Bronnenberg, B.J., Dhar, S.K., and Dube, J.P.H., "Market Structure and the Geographic Distribution of Brand Shares in Consumer Package Goods", University of Chicago, Working paper, 2005.
[8] Bucklin, L., Venkatram R., and Sumit M., “Analyzing Channel Structures of Business Markets via the Structure-Output Program, International Journal of Research in Marketing, 13, 73 - 87, 1996.
[9] Czinkota, M.R. and Ronkainen, I.A., International Marketing, 6th edition, Florida: Harcourt Inc., 2001.
[10] Cameron, A.C., and Trivedi, P.K., Regression Analysis of Count Data, Cambridge University press, 1998.
[11] Corey, E.R., Cespedes, F.V., and Rangan, V. K., "Going to Market:
Distribution Systems for Industrial Products" Harvard Business School Press, Boston, 1989.
[12] Coughlan, Anne T., Erin Anderson, Louis W. Stern and Adel I El-Ansary, Marketing Channels, 6th ed., Upper Saddle, 2001.
[13] Dhar, S.K., Hoch, S.J., and Kumar, N., "Effective Category Management Depends on the Role of the Category", Journal of Retailing, 77 (2), 165-184, 2001.
[14] Dhalla, N.K., Yuspeh, S., "Forget the product life cycle concept", Harvard Business Review, Jan-Feb, 1976.
[15] Fader, P., Hardie, B. and Lee, Ka Lok, "RFM and CLV: Using Iso-Value Curves for Customer Base Analysis", Journal of Marketing Research, 42, 415-430, 2005.
[16] Frazier, Gary L. and Lassar, Walfried M., "Determinants of Distribution Intensity", Journal of Marketing, 60 (4), 39-51, 1996.
[17] Frazier, G., Kirti S., and Tasadduq S., “Intensity, Functions, and Integration in Channels of Distribution”, in Zeithaml, Valarie, (Ed) A.M.A. Review of Marketing, Chicago, 263 - 298, 1990.
[18] Fein, A.J. and Anderson, E., "Patterns of credible commitments: territory and brand selectivity in industrial distribution channels", Journal of Marketing 61(April),19-34,1997.
[19] Fein, A.J. and Anderson, E., "Patterns of credible commitments: territory and brand selectivity in industrial distribution channels", Journal of Marketing 61(April),19-34,1997.
[20] Gastwirth, J.L., "The estimation of the Lorenz curve and Gini index Source", The review of economics and statistics, 54 (3), 306-316, 1972.
[21] Hardy, K. G., and Magrath, A. J., Marketing Channel Management: Strategic Planning and Tactics, Illinois: Scott Publishing, 1988.
[22] Ingene, C.A., “ Productivity and functional shifting in spatial retailing: private and social perspectives,” Journal of Retailing, 60, 15–36, 1984.
[23] Jain, S., International Marketing Management, Fourth Edition, WadsworthPublishing, California, 1993.
[24] Johansson, J.K., Global Marketing, 2nd edition. USA: Irwin, 2000.
[25] Johnson, J. and Tellis ,G. J., (2008), "Drivers of Success for Market Entry into China and India," Journal of Marketing, 72, 1-13.
[26] Kotler, P. and Keller, K.L., Marketing Management, Upper Saddle River, NJ, 2009.
[27] Krugman, Paul, "Increasing Returns and Economic Geography", Journal of Political Economy, University of Chicago Press, 99(3), 483-499, 1991.
[28] Li, L., "Determinants of Exportdistribution intensity in Emerging Markets: The British Experience in China", Asia Pacific journal of management, 20(4) 501-516, 2003.
[29] Lassar, W., and Jeffrey K., “Strategy and Control in Supplier-Distributor Relationships: An Agency Perspective”, Strategic Management Journal, 17, 613 - 632,1996.
[30] Mallen, Bruce, “Marketing Channels and Economic Development: A Literature Overview”, International Journal of Physical Distribution and Logistics Management, 26( 5), May, 42 - 47,1996.
[31] Morrison, D.G. and Schmittlein, D.C., "Generalizing the NBD Model for Customer Purchases", Journal of the American Statistical Association, 6(2) ,145-159, 1988.
[32] Onvisit, S., and Shaw, J., International Marketing: Analysis and Strategy, MacMillan Publishing, New York, 1990.
[33] Pelton, L.P., Strutton, D., and Lumpkin, J.R., Marketing Channels: A Relationship Approach, McGraw-Hill Higher Education, New York, 2002.
[34] Porter, M.E., The Competitive Advantage of Nations. New York: The Free Press, 1990.
[35] Quer, D., Claver, E., and Rienda, L., "Business and management in China: A review of empirical research in leading international journals Business and management in China: A review of empirical research in leading international journals", Asia Pacific Journal of Management, 24( 3), 359-384(26), 2007.
[36] Robertson, T.S. and Gatignon, H, "How innovators thwart new entrants into their market", Strategy & leadership, 19(5), 4-12, 1991.
[37] Rungie, C.M., Laurent, G., Habel ,C.A., A New Model for the Pareto Effect (80: 20) at Brand Level, Anzmac, Melbourne, Anzmac, 2002.
[38] Rangan, V.K., "The Channel Design Decision: A Model and an Application", Management Science, 6( 2), Spring,156 - 175,1987.
[39] Rosenbloom, B., Marketing Channels, 5th Edition, Dryden Press, Chicago,1995.
[40] Sabavala, D.J., "Generalizing the NBD Model for Customer Purchases: What Are the Implications and Is It Worth the Effort?", Journal of the American Statistical Association, 6(2) ,159-161, 1988.
[41] Schmittlein, D.C., Cooper, L.G., Morrison, D.G., "Truth in Concentration in the Land of (80/20) Laws", Marketing Science, 12(2), 167-183,1993.
[42] Stigler, G.J., "Price and Non-Price Competition Source", The journal of political economy, 76 (1), 149-154, 1968.
[43] Sutton, J., Sunk Costs and Market Structure: Price Competition, Advertising, and the Evolution of Concentration, MIT Press, 1991.
[44] Sutton, J., Technology and market structure: Theory and history, Cambridge, MA: MIT Press, 1998.
[45] Tang, Ying-chen Edwin and Lee, Pui-Wan Ruby, "China Retailing in Trasition-A Dynamic and Evolutionary Perspective", Pan Pacific Management Review, 2, 1-18, 1998.
[46] Winkelmann, R., Econometric Analysis of Count Data, fifth edition, eidelberg, New York: Springer., 2008.
Appendix A 200 selected Cities in China
NO. City City Tier NO.City City Tier NO. City City Tier NO. City City Tier
1 ShangHai 1 51 LuoYang 3 101 HuaiNan 3 151 JiaoZuo 4
2 BeiJing 1 52 ChangZhou 3 102 PingDingShan 3 152 YingKou 4
3 GuangZhou 1 53 BaoTou 3 103 HuLuDao 3 153 QinZhou 4
4 ShenZhen 2 54 LinYi 3 104 PanZhiHua 3 154 XiNing 4
5 TianJin 2 55 WeiFang 3 105 ZhangZhou 3 155 LeShan 4
6 WuHan 2 56 HuZhou 3 106 ShiYan 3 156 KaiFeng 4
7 DaQing 2 57 FuShun 3 107 YiChang 3 157 BingZhou 4
8 ShenYang 2 58 HanDan 3 108 JiuJiang 3 158 CangZhou 4
9 DaLian 2 59 QuanZhou 3 109 ZiGun 3 159 TongLing 4
10 ChongQing 2 60 NanTong 3 110 EZhou 3 160 DeYang 4
11 NanJing 2 61 YueYang 3 111 JiNing 3 161 TongLiao 4
12 HangZhou 2 62 TaiAn 3 112 XiangFan 3 162 SuZhou 4
13 ChengDu 2 63 FoShan 3 113 JinZhou 3 163 PuTian 4
14 JiNan 2 64 QinHuangDao 3 114 DeZhou 3 164 NanPing 4
15 QingDao 2 65 ZhuZhou 3 115 LiaoYang 3 165 JinHua 4
16 ChangChun 2 66 JiangMen 3 116 MuDanJiang 3 166 ChaoHu 4
17 Xi'An 2 67 WeiHai 3 117 ZhaoQing 3 167 SanMing 4
18 HaErBing 2 68 ZhenJiang 3 118 SuiZhou 3 168 ChengDe 4
19 ZiBo 2 69 LiuZhou 3 119 DanDong 3 169 NeiJiang 4
20 XiaMen 2 70 MianYang 3 120 HengYan 3 170 XuanCheng 4
21 DongGuan 2 71 KeLaMaYi 3 121 LuZhou 3 171 XinYu 4
22 KunMing 2 72 ZaoZhuang 3 122 ShaoGuan 3 172 LiaoCheng 4
23 FuZhou 2 73 HuiZhou 3 123 ChangZhi 3 173 XieTai 4
24 NingBo 2 74 DaTong 3 124 LongYan 3 174 BaiYin 4
25 WuXi 2 75 YangZhou 3 125 JingZhou 3 175 HengShui 4
26 ShiJiaZhuang 2 76 BaoDing 3 126 ZhouShan 3 176 BoZhou 4
27 ChangSha 2 77 HaiKou 3 127 YongZhou 3 177 WuZhou 4
28 DongYing 2 78 NanYang 3 128 XinXiang 3 178 SongYuan 4
29 ZhengZhou 2 79 ChangDe 3 129 YinChuan 3 179 SuiNing 4
30 SuZhou 2 80 HuHeHaoTe 3 130 QuJing 3 180 YiChun 4
31 WenZhou 2 81 JiaXing 3 131 BengBu 3 181 FuXin 4
32 AnShan 2 82 QiQiHaEr 3 132 LangFang 3 182 JingDeZhen 4
33 ZhongShan 2 83 BenXi 3 133 PingXiang 3 183 ChuZhou 4
34 NanChang 3 84 ShaoXing 3 134 YiBin 4 184 GanZhou 4
35 YanTai 3 85 XiangTan 3 135 ShangQiu 4 185 ChaoZhou 4
36 TangShan 3 86 XianYang 3 136 XinYang 4 186 BeiHai 4
37 TaiYuan 3 87 ZhangJiaKou 3 137 AnQing 4 187 SuZhou 4
38 XuZhou 3 88 MaoMing 3 138 ChiFeng 4 188 XuChang 4
39 ZhuHai 3 89 RiZhao 3 139 YangQuan 4 189 JiXi 4
40 LanZhou 3 90 QinChuan 3 140 JieYang 4 190 JinCheng 4
41 WuLuMuQi 3 91 WuFu 3 141 HuaiAn 4 191 YuLin 4
42 TaiZhou 3 92 LaiWu 3 142 YanCheng 4 192 WuWei 4
43 PanJin 3 93 JingMen 3 143 BingZhou 4 193 ZiYang 4
44 JiLin 3 94 HuangShi 3 144 HuaiBei 4 194 TongHua 4
45 HeFei 3 95 LianYunGang 3 145 ZunYi 4 195 TianShui 4
46 ShanTou 3 96 PuYang 3 146 NanChong 4 196 LinFen 4
47 GuiYang 3 97 BaoJi 3 147 YangJiang 4 197 HeBi 4
48 NanNing 3 98 AnYang 3 148 YiYang 4 198 QiTaiHe 4
49 YuXi 3 99 GuiZhou 3 149 JiaMuSi 4 199 LuoHe 4
50 ZhanJiang 3 100 MaAnShan 3 150 FuYang 4 200 ZhaoTong 4
Appendix B Computing Codes By R language
--- In Table 5.3 Pareto Share example--- x_0.189
y_qgamma(.2,1+x) pgamma(y,x)
--- In Table 5.3 Gini index--- r_0.2
ob1_qgamma(0,r) ob2_qgamma(0.5,r) ob3_qgamma(1,r)
1-2*(1/6*pgamma(ob1,r+1)+4/6*pgamma(ob2,r+1)+1/6*pgamma(ob3,r+1)) --- In Figure 5.2 3C CDI/BDI scattergram example--- rates.dat<-read.table("a:rates.txt",header=T)
rates.mat<-as.matrix(rates.dat) win.graph( )
par(pty="s")
plot(rates.mat,xlab="CDI",ylab="BDI") title("HP")
--- In Figure 5.3 contour plot example--- win.graph( )
contour(rates.mat[,1],rates.mat[,2],rates.mat[,3],xlab="CDI",ylab="BDI")
By SAS
--- In Table 5.6 NBD regression example--- PROC GENMOD DATA = cdata;
MODEL CI=CD BD /DIST=NEGBIN LINK=LOG;
RUN;
PROC GENMOD DATA = cdata;
MODEL CI= /DIST=NEGBIN LINK=LOG;
RUN;