Should Firms Employ Personalized Pricing?

• Published date: 01 October 2015
• Author: Noriaki Matsushima,Toshihiro Matsumura
• Date: 01 October 2015
• DOI: http://doi.org/10.1111/jems.12109

Should Firms Employ Personalized Pricing?

TOSHIHIRO MATSUMURA

Institute of Social Science

the University of Tokyo

Hongo 7-3-1, Bunkyo, Tokyo 113-0033 Japan

matsumur@iss.u-tokyo.ac.jp

NORIAKI MATSUSHIMA

Institute of Social and Economic Research

Osaka University

Mihogaoka 6-1, Ibaraki, Osaka 567-0047 Japan

nmatsush@iser.osaka-u.ac.jp

This paper theoretically considers a duopoly model in which all firms do not always employ per-

sonalized pricing. Our model incorporates the fact that firms engage in marginal cost-reducing

activities after they decide whether to employ personalized pricing. When the ex ante cost differ-

ence between the firms is large, the less-efficient firm does not employ personalized pricing even

when the fixed cost to do so is zero. This is because employing personalized pricing induces the

rival firm to engage more in reducing its costs, which is more likely to harm the less-efficient firm.

1. Introduction

Recent developments in information technology (IT) have enabled firms to employ per-

sonalized pricing (Arora et al., 2008 and references therein). In the U.K. supermarket

industry, Tesco learns a variety of facts on customer purchasing activities using profiles

developed based on information collected through loyalty cards. Tescouses these data to

send personalized coupons and other offers to all Tesco Clubcardhouseholds quarterly,

and this is a huge operation that accounts for over 6% of the annual postal volume in the

United Kingdom (Raju and Zhang, 2010, pp. 131–132). 1However, some retailers do not

employ such personalized pricing strategies despite being able to. Asda, one of the “big

three” supermarkets in the United Kingdom, introduced a loyalty card scheme in 1994,

which it discontinued in 1999 after Tesco and Sainsbury, the other two largest chains, re-

spectively,launched their loyalty card programs in 1995 and 1996 (Thanassoulis, 2009).2

This raises the following question: Should all firms employ personalized pricing? If not,

why not? The purpose of this paper is to provide a simple rationale behind why firms

do not always employ personalized pricing.

∗We thank the Editor-in-Chief, an anonymous co-editor, and two anonymous referees for their helpful and

insightful comments. We also thank seminar participants at Nagoya University and Shinshu University for

their helpful comments. The authors gratefully acknowledge financial support from Grant-in-Aid from JSPS

KAKENHI Grant Numbers 22530175, 23330099, and 24530248. Needless to say, we are responsible for any

remaining errors.

1. Recently, grocers such as Safeway and Kroger have also begun using various methods to determine

individualized prices (The New YorkTimes [August 10, 2012]). Hoping to improve razor-thin profit margins,

they are creating specific offers and prices based on shoppers’ behaviors, which could encourage shoppers to

spend more.

2. The head of research and pricing at Asda mentioned that it preferredto provide customer value in other

ways (Rohwedder, 2006).

C2015 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume24, Number 4, Winter 2015, 887–903

888 Journal of Economics & Management Strategy

The model setting is as follows. Two firms compete in a differentiated product

market, and each firm determines whether to employ personalized pricing. If a firm

employs personalized pricing, it is able to distinguish among customers and offer them

personalized prices; otherwise, it offers a standard nondiscriminatory price to all cus-

tomers. The model setting is based on the models of Thisse and Vives (1988) and Shaffer

and Zhang (2002). Our model also incorporates the fact that firms engage in marginal

cost-reducing or quality-improving activities.3More concretely, we set a model in which

the firms engage in these activities after determining their pricing policies. This assump-

tion implies that introducing personalized pricing is time-consuming because it requires

significant time to optimize, and thus, is a long-term decision.4

We explain the plausibility of the timing structure employed in this paper, that is,

firms determine their pricing policies before they engage in cost-reducing or quality-

improving activities. In the retailing context, retailers must construct electronic devices

to distribute appropriate promotional coupons using frequent-shopper programs based

on household purchase record analysis. Furthermore, to use such devices, these retailers

need to employ various items, including new cash registers, computers to gather cus-

tomer purchase records, applications for record analysis, data analysts, and customer

service offices for their frequent shopper programs.5After they set up personalized pric-

ing through these efforts, they engage in several cost-reducing activities. Tesco uses the

Clubcard information to determine which items should be stocked in a store given local

demographic and shopping habits (Thanassoulis, 2009); this reduces unnecessary stocks

of goods and lost sales opportunities. Clearly,Tesco engages in this efficiency-improving

effort by using the database after it does complex data mining, which is an expensive

intelligence source and a challenging process because of the enormous amount of data

generated (Hair Jr., 2007; Humby et al., 2007).6In other words, Tesco engaged in cost-

reducing activities after it invested in its customer data analysis by creating its loyalty

program. Moreover,as mentioned in Pancras and Sudhir (2007), in several research areas

including marketing, information systems, and computer science, research has focused

on the method by which firms should use individual browsing/purchasing data to

personalize advertising or promotions after introducing devices for personalized pro-

motions. This implies that employing personalized pricing requires significant effort to

implement promotions and has a long-term and time-consuming investment nature. We

therefore believe that the timing structure in our model is plausible.7

3. In our model, quality improvement is qualitatively equivalent with cost reduction, and is discussed in

Section 4.1.

4. This case is related to oligopoly models with cost reductions where firms engage in cost-reducing

activities before they set their prices or quantities. These models have been intensively discussed in the

literature of industrial organization (e.g., Brander and Spencer,1983). Recently, researchers have investigated

the relationship between firm asymmetry and R&D activities as discussed in this paper (Cohen and Klepper,

1996; Matsumura and Matsushima, 2010; Ishida et al., 2011). We believe that our paper can offer a novel

contribution in this context. Although most studies in this field have focused on the effects of R&D on

equilibrium pricing levels, our paper sheds light on the effect of each firm’s decision on pricing policy (i.e.,

whether to employ personalized pricing) on the strategic R&D behavior of firms, and the influence of the

degree of firm asymmetry.

5. In fact, Tesco calculated that a 1.6% uplift in sales was requiredto offset the 10 million pounds launch

cost of Clubcard and the cost of issuing membership cards and reward vouchers (Humby et al., 2007, p. 64).

6. In 2006, Tesco signed up 12 million Britons for its Clubcardprogram (Rohwedder, 2006).

7. Of course, even if a firm does not employ personalized pricing, it still engages in cost-reducing activity.

For instance, firms (retailers) incur investment costs to achieve saving in distribution costs through reduction

of the lead time, reduction of out-of-stock losses, and less falloff and damages. One might think that a firm

may seem to be able to keep its research and development (R&D) costs low by not employing personalized

pricing. Although we discuss the case in which the effort costs under the two pricing policies are the same in