International Journal of Management Vol. 28 No.

1 Part 2 Mar 2011 387

Relationship between Marketing Efforts and Sales of

Consumer Brands: An Empirical Study in India
Mehir Baidya
Indian Institute of Social Welfare and Business Management, India
Bipasha Maity
Future Institute of Engineering and Management, India

This study examines the relationship between the marketing efforts of firms, how much
time and money, they put into the marketing of particular products and the returns they
get from these products. The study was done on two Indian manufacturers of different
brands of consumer product, in this case hair-oil. Data on physical sales and different
marketing efforts (advertising, sales force, promotion and distribution) were have
collected for two brands (Keo-Karpin and Navratna) of the two firms. A multiplicative
regression model was then hypothesized and fitted on the data to estimate the response
coefficients of individual marketing efforts. Results suggest that there exists a constant
return to scale of sales (in units) to marketing efforts for both the brands. Findings
should assist managers of such brands to allocate resources to marketing in such a way
as to increase revenues.
1. Introduction
The concept of sales response to individual marketing efforts is a central concern of
the theory and practice of marketing management (Lilien and Rangaswamy, 2005).
Marketers have long been interested in developing the sales response function model for
describing the complex relationship between their firms and markets (Manchanda and
Chintagunta, 2004). The relationship between sales (physical or nominal as the case may
be) and different marketing efforts (e.g. advertising, sales force, promotion, distribution,
etc) is usually estimated by the managers in order to improve their understanding of how
different activities influence sales (Parsons and Abeele, 1981).
Several studies in the area have investigated the effects of advertising (Lodish et al., 1995;
Dekimpe and Hanssens, 1995; LaBahn and Biehal, 1991; Bass, 1969; Hanssens, 1980;
Aaker, 1975), sales force (Parsons and Abeele, 1981; Flaherty et al., 2007; Piercy et al.,
2007), promotion (Blattberg and Wisniewski, 1989; Ailwadi and Neslin, 1998; Neslin,
1990; Neslin et al., 1985) and distribution (Reibstein and Farris, 1995; Anderson, 1985)
on the sales of different bands in particular category. Findings of these studies have been
utilized as: to determine optimal levels of marketing efforts (e.g., Erickson, 1992) and
allocation of funds on different efforts (Chintagunta and Vilcassim, 1994). However,
few if any studies have examined dealt with the aggregate effect of different marketing
efforts on sales (in units). Further, existing literature in quantitative marketing suggests
that maximum number of studies have been carried out in western countries. Here, it is
to be noted that the findings of these studies might not be equally applicable in India as
in their respective countries.
388 International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011

To be more specific in Indian scenario, an exploratory study has been conducted in the
hair-oil product sector in West Bengal, India. It has been observed that the managers
have been investing a huge sum of money in different marketing efforts. Further, they
have been doing so without having a clear idea about the response function of sales to
these efforts (aggregate or individual as the case may be) for their brands. The purpose
of this study to develop a sales response to different marketing efforts function for two
established brands (Keo-Karpin and Navratna) of two firms in the hair-oil product sector
in West Bengal, India. However, there are 11 firms are being in operation in the sector.
All the firms have been approached to participate. Only 2 firms out of 11 agreed to share
the relevant data for this research.
In the above backdrop, the following basic, yet managerially important research question
has been identified: What is the brand-wise return to scale of sales to individual marketing
efforts? Does the estimated brand-wise return to scale significant? If yes, do they differ
from unity?
2. Research Design
2.1. Model
A double-log function has been considered for the research question as mentioned earlier
and postulated as:
β β β β
Yt = αX 1t 1 X 2t 2 X 3t 3 X 4t 4 u t , For t = 1,2,.....,24 . (1)
This form has been chosen because there are some supports for a convex-concave or
S-shaped total response function with respect to individual marketing efforts. Many
empirical studies support diminishing marginal return to each individual marketing effort
(White et al., 2001; Basu and Batra, 1988; Simon and Arndt, 1980). In this context, an
additive model cannot capture this non-linear response characteristic because it assumes
straight-line response function and this may be a reasonable simplifying assumption only
within a relatively narrow range (Lilien and Kotler, 1983). Power function has been widely
used in many other multivariate and bivariate studies relating to impact of individual
marketing efforts on sales in units (Mesak, 2002; Stone and Duffy, 1993; Neslin, 1990;
Varadarajan, 1984; Naik and Raman, 2003; Belch and Belch, 1998; Schultz, 1993).
The coefficients of the model in equation (1) are the measures of sales (in units) elasticity
for each marketing efforts. Coefficient β i is the percent change is sales in units resulting
from a one percent change in the i th marketing effort. In a Cobb-Douglas sales function,
4
the sum of these coefficients, ∑β
i =1
i , is the degree of homogeneity, which measures
whether the sales (in units) function is constant, increasing, or decreasing returns to
scale. Three possibilities exist:
4
• If ∑β
i =1
i = 1 , there is constant return to scale of sales to marketing efforts.
4
• If ∑β
i =1
i < 1 , there is decreasing return to scale of sales to marketing efforts.
International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011 389

4
• If ∑β
i =1
i > 1 , there is increasing return to scale of sales to marketing efforts.
4
To test the significance of ∑β
i =1
i
, the equation (1) has been rearranged by multiplying

β1 β2 β
and dividing by X 3t X 3t X 3t 4 .
β β β4
β β β β4 X 3t 1 X 3t 2 X 3t
Yt = αX 1t 1 X 2t 2 X 3t 3 X 4t β β β4
ut (2)
X 3t 1 X 3t 2 X 3t
Rearranging the terms in equation (2) yield:
β1 β2 β4
X  X  β +β +β +β  X 
Yt = α  1t   2t  X 3t 1 2 3 4  4t  u t (3)
 X 3t   X 3t   X 3t 
Let δ = β 1 + β 2 + β 3 + β 4 , then equation (3) can be rewritten as:
β1 β2 β4
X  X  δ X 
Yt = α  1t   2t  X 3t  4t  u t (4)
 X 3t   X 3t   X 3t 
The model presented as equation (4) can be used for estimating the degree of homogeneity
directly and for testing its statistical significance.
The final model has been presented below by taking natural log in both the sides of the
equation (4).
 X 1t  X  X 
ln Yt = β 0 + β1 ln
ln ln   + β 2 ln
ln  2t  + δ ln
ln X 3t + β 4  4t  + vt (5)
 X 3t   X 3t   X 3t 
where, Yt = Sales in units for period t; X 1t = Advertising expenditures in terms of
rupees for period t; X 2t = Sales force expenditures in terms of rupees for period t;
X 3t = Sales promotion expenditures in terms of rupees for period t; X 4t = Distribution
expenditures in terms of rupees for period t; u t = Random disturbance term for period
t; β 0 = ln
ln α and
vt = ln
ln u t
2.2. Data and Methodology
To answer the research question as mentioned earlier, quarterly time-series data on
sales in units (200 ml bottles), marketing efforts (advertising, sales force, promotion
and distribution) have been collected from unpublished sources, namely documents and
records for both the brands over the period 2003-2008. These efforts have been considered
since only discretionary expenditures that directly affect sales are of concerned. All the
data have been deseasonalised by seasonal indices except for sales force (since it does not
exhibit any seasonal patterns) (Makridakis et al., 1998). Further, data on advertising, sales
390 International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011

force and distribution have been divided by promotion to use as inputs with promotion
to perform a double-log regression analysis between sales (in units) and these regressors.
Hence, the estimated regression equation directly identifies the returns to scale of sales
as a regression coefficient of promotion (equation 5). Further, an additional ‘ t ’ test has
been performed to identify whether the elasticity (returns to scale of sales in this case)
of sales to promotion is significantly differing from unity.

3. Results and Discussion

To estimate the relationship of sales in units with individual marketing efforts (advertising/
promotion, sales force/promotion, promotion and distribution/promotion), multiple
regression analyses with ordinary least squares (OLS) algorithm have been performed
on the quarterly data for both the brands. The regression results are presented in Table
1 for both the cases.

Results of positive elasticity coefficient ( δ ) of promotion indicate that the overall

marketing effort has a positive relationship with sales in units for both the brands. That
is, 1 percent increase in marketing efforts would lead to increase sales (in units) by 0.83
and 1.10 percent for Keo-Karpin and Navratna respectively.
To test whether the estimated elasticity of sales to overall marketing has significant
positive effect on physical sales, the following hypothesis has been framed:
H1: Overall marketing effort would have a positive effect on physical sales
That is, in mathematically, H 0 : δ = 0 against H a : δ > 0

Table 1 Results of the Tests on Elasticity for Keo-Karpin and Navratna

Null Hypotheses of Null Hypotheses of

Zero Elasticity Unitary Elasticity
Elasticity
Coefficient Keo- Keo-
Navratna Navratna
Karpin Karpin
Keo- Navratna t-statistic t-statistic t-statistic t-statistic
Karpin
β0 1.70 -18.63 0.82 -2.38***

β1 -0.24 -06.90 -1.04 -0.91

β2 -0.70 10.22 -1.80*** 2.57**

δ 0.83 01.10 5.22* 3.51** -1.04 0.32

β4 1.83 12.13 7.52* 1.34

Notes: *Significant at .0001 percent level (one-tail), ** at 1 percent level (one-tail), ***at
5 percent level (one-tail).
International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011 391

Results suggest that both the values of elasticity of sales for both the brands are significant
at a desired level. That is, the hypothesis (H1) has been confirmed for both the cases. One-
tailed tests have been used in this context due to some universal notion that the overall
marketing effort should have at least a positive effect on sales. Hence, the hypothesis
has been formulated and tested on the basis of one-tailed tests.
At the same time, it would be interesting to find out whether the elasticity of sales (in
units) is showing unitary elasticity with respect to marketing efforts for both the brands.
In this context, the following hypothesis has been formed:
H2: Elasticity of physical sales to marketing efforts is unitary
That is, in mathematically, H 0 : δ = 1 against H a : δ ≠ 1
To test this hypothesis, a ‘t test’ has been carried out and results are shown in Table 1.
Results of these tests suggest that the elasticity of physical sales to marketing efforts is
not significantly different from one on even two-tailed tests for both the cases. Hence,
the hypothesis (H2) mentioned above has been confirmed for both the brands. That is,
response function of sales (in units) to marketing efforts shows a constant return to scale
for both the cases over the study period 2003-2008.
The results of validation statistics of Keo-Karpin and Navratna are presented in Table 2.
2
It is shown that the Radjusted (since the number of observations was not equal for both the
brands) values are very high for both the brands. That is, sales in units as a double-log
function of marketing efforts provide a good fit to the observed data for both the cases.
2
Marginally higher Radjusted value for Keo-Karpin depicts that the marketing efforts have
explained more variance in sales than for Navratna. Further, the results indicate, both
the values of DW statistic are insignificant, which would lead to the conclusion that the
error terms are not serially correlated for both the cases.
4. Conclusion and Implications
In this work, a sales (in units) response to marketing efforts function has been estimated
to measure the return to scale of sales for two brands in India. The results show that the
physical sales exhibit constant return to scale to marketing efforts for both the cases.
That is, proportionate increase in all the marketing efforts would lead to increase sales
by equal proportion. The findings have several implications: Managers can allocate a

Table 2 Validation Statistics for Regression of Sales on Marketing Efforts

Brand R2 F-value R2adjusted DW
Keo-Karpin 0.99 425.9* 0.98 1.92
Navratna 0.91 158.50* 0.89 1.81
Notes: Significant at .0001 percent level and DW= Durbin-Watson Statistic
*
392 International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011

fixed budget to different marketing effort on the basis of estimated coefcients.to utilize
marketing efforts more effectively to achieve better return to scale of sales in future.
This work could not incorporate the carry-over effect of marketing efforts in the model
presented here in view of small number of observations. It would be of great interest to
expand the current study from a two-brand analysis in a single sector to a multi-brand
one covering a number of sectors.
References
Aaker, D. A., (1975). ADMOD: An advertising decision model. Journal of Marketing
Research 12: 37-45.
Ailwadi, K. L., & Neslin, S. A., (1998). The effect of promotion on consumption: Buying
more and consuming it faster. Journal of Marketing Research 35: 390-398.
Anderson, E., (1985). The salesperson as outside agent or employee: A transaction cost
analysis. Marketing Science 4: 234-254.
Bass, F. M., (1969). A simultaneous equation regression study of advertising and sales
of cigarettes. Journal of Marketing Research 6: 291-300.
Basu, A. K., & Batra, R., (1988). ADSPLIT: A multi-brand advertising budget allocation
model. Journal of Advertising 17: 44-51.
Belch, G. E., & Belch, M. A., (1998). Advertising and Promotion: An Integrated
Marketing Communications Perspective. Boston, MA: Irwin-McGraw Hill Inc.
Blattberg, R.C., & Wisniewski, K.J., (1989). Price-induced patterns of competition.
Marketing Science 8, 291-309.
Chintagunta, P.K., & Vilcassim, N.J., (1994). Marketing investment decisions in a
dynamic duopoly: A model and empirical analysis. International Journal of Research
in Marketing 11, 287-306.
Dekimpe, M.G., & Hanssens, D.M., (1995). The persistence of marketing effects on
sales. Marketing Science 14, 1-21.
Erickson, G. M., (1992). Empirical analysis of closed-loop duopoly advertising strategies.
Management Science 38, 1732-1749.
Flaherty, K. E., Arnold, T. J., & Hunt, C. S., (2007). The influence of the selling situation
on the effectiveness of control: Toward a holistic perspective. The Journal of Personal
Selling and Sales Management 27: 221-234.
Hanssens, D. M., (1980). Market response, competitive behavior and time series analysis.
Journal of Marketing Research 17: 470-485.
LaBahn, D. W., & Biehal, G., (1991). Empirical Research in Strategic Market
Planning: Current Progress and Future Directions. Chicago, IL: American Marketing
Association.
Lilien, G. L., & Kotler, P., (1983). Marketing Decision Making: A Model-Building
Approach. New York, NY: Harper & Row.
International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011 393

Lilien, G. L., & Rangaswamy, A., (2005). Marketing Engineering. Singapore: Pearson
Education, Inc.
Lodish, L.M., Magid, A., Stuart, K., & Stuart, L.L., (1995). How T.V. advertising works:
A meta-analysis of 389 real world split cable T.V. advertising experiments. Journal of
Advertising Research 32, 125-139.
Makridakis, S., Wheelwright, S., & Hyndman, R.J., (1998). Forecasting: Methods and
Applications. England: John Wiley & Sons, Ltd.
Manchada, P., & Chintagunta, P.K., (2004). Responsiveness of physician perception
behavior to sales force effort: An individual level analysis. Marketing Letters 15: 129-
145.
Mesak, H. I., (2002). On the impact of initial performance on the effectiveness advertising
pulsation policies. The Journal of the Operational Research Society 53: 1247-1255.
Neslin, S. A., (1990). A market response model for coupon promotions. Marketing
Science 9: 125-145.
Neslin, S. A., Henderson, C., & Quelch, J., (1985). Consumer promotions and the
acceleration of product purchases. Marketing Science 4: 147-165.
Parsons, L. J., & Abeele, P.V., (1981). Analysis of sales call effectiveness. Journal of
Marketing Research 18:107-113.
Piercy, N. F., Cravens, D. W., & Lane, N., (2007). Enhancing salespeople’s effectiveness.
Marketing Management 16: 18-25.
Reibstein, D. J., & Farris, P.W., (1995). Market share and distribution: A generalization,
a speculation and some implications. Marketing Science 14:G190-G202.
Schultz, D. E., (1993). Integrated marketing communications: Maybe definition in the
point of view. Marketing News 27: 17
Simon, J. L., & Arndt, J., (1980). The shape of the advertising response function. Journal
of Advertising Research 20: 11-20.
Stone, R., & Duffy, M., (1993). Measuring the impact of advertising. Journal of
Advertising Research 33: RC9-RC12.
Varadarajan, P. R., (1984). Consumer response to small business coupon sales promotions.
American Journal of Small Business 9: 17-26.
White, J. B., Miles, M. P., & Smith, M. F., (2001). The interrelationship of marketing
productivity and financial performance for global SMEs. Irish Marketing Review 14:
55-60.

Contact email addresses: [email protected] [email protected]

International Journal of Management Vol. 28 No. 1 Part 2 Mar 2011 395

Salomão Farias Mehir Baidya

Francisco Moura Indian Institute of Social Welfare
Universidade Federal de Pernambuco and Business Management
Av. Prof. Moraes Rego, 1235 Management House,
Cidade Universitária College Square (W)
50670-901 - Recife, PE Kolkata-700073, W.B.
Brazil India

Daniel Gerhard Bipasha Maity

Alexander Brem Department of MBA
Christian Baccarella Future Institute of Engineering
Kai-Ingo Voigt and Management
Dept of Industrial Engineering Sonarpur Station Road
University of Erlangen-Nuremburg Kolkata-700150
Lange Gasse 20 India
90403 Nuremburg
Germany

Richard J. Cebula
College of Business
Jacksonville University
Jacksonville
Florida 32211 USA

Daewoo Park
Mina Lee
James Turner
Lynda Kilbourne
College of Business
Xavier University
Cincinnati
OH 45207 USA

Adele Parmentola
International Management
Università degli Studi di Napoli
Parthenope
Dipartimento di Studi Aziendali
Via Medina 40
80133 Napoli
Italy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.