Should Mobile Marketers Collect Data Other Than Geo‐Location?*

• Author: Geza Sapi,Irina Baye
• Date: 01 April 2019
• DOI: http://doi.org/10.1111/sjoe.12275
• Published date: 01 April 2019

Scand. J. of Economics 121(2), 647–675, 2019

DOI: 10.1111/sjoe.12275

Should Mobile Marketers Collect Data

Other Than Geo-Location?*

Irina Baye

DICE, Heinrich Heine University of D¨usseldorf, DE-40225 D¨usseldorf, Germany

baye@dice.hhu.de

Geza Sapi†

European Commission DG COMP, Chief Economist’sTeam, B-1049 Brussels, Belgium

sapi@dice.hhu.de

Abstract

Wetake today’smobile marketing data landscape as a starting point and consider a duopoly model

of third-degree price discrimination in which firms can complement geo-location information

with data on consumer flexibility of varying quality. We show that, depending on consumer

heterogeneity,higher-quality flexibility data affect profits according to three different patterns. In

equilibrium, both firms tend to acquire data if the data are of high quality,while only one acquires

data if the data quality is low. Firms are likely to gain from additional data if consumers have

similar preferences and/or when data are precise. Although social welfare (weakly) improves,

consumers can be harmed.

Keywords: Customer data; location targeting; mobile advertising; price discrimination

JEL classification:D43; L11; L13

I. Introduction

Location-based advertising is one of the fastest growing and most hotly

debated areas of advertising technology since the launch of the iPhone

in 2007. In the United States, spending on mobile advertising campaigns

using location data is projected to double, growing from 16 billion US

dollars (USD) in 2017 to 32.4 billion USD in 2021 (McAdams, 2017).

Geo-targeting enabled unprecedented personalization of offers to customers

depending on their physical location obtained from smartphone signals

(Ionescu, 2010). Location data have become near perfect, as they can be

transmitted in real time by mobile applications (IAB, 2015). Yet “the critical

*Wethank S. P.Anderson,T.Athanasopoulos, J. Bouckaert, J. Claussen, S. Kovbasyuk,C. Peukert,

and other participants of the Media Economics Workshop (2016), the CRESSE Conference

(2016), the IIOC (2016), the Workshop on the Economics of Information Security (2013), and

the EARIE Conference (2013) for their very valuable comments. The views expressed in this

paper are solely those of the authors and do not represent an official position of the European

Commission.

†G. Sapi is a research affiliate at D¨usseldorf Institute for Competition Economics (DICE).

C

The editors of The Scandinavian Journal of Economics 2017.

648 Should mobile marketers collect data other than geo-location?

element when designing services that utilize mobile location is to ensure

they [advertisers] have more attributes than just location” (Howarth, 2014).

Indeed, information relating to age and income, for example, and even

health status, are typical variables in datasets that mobile marketers can

collect or obtain to complement their data on consumers’ real-time physical

locations.1These data allow advertisers to conclude on consumer flexibility,

which shows how likely consumers are to switch between competing

advertisers depending on the discounts offered. For example, consumers of

different ages, incomes, or health statuses at the same physical location

will likely differ in the incentives needed to induce them to approach

either firm. Younger and healthier consumers are likely to be more flexible,

and similarly, lower-income consumers can be reasonably expected to react

differently from higher-income consumers to the same unit of discount

offered by either firm. It should also be noted here that the techniques

employed by mobile marketing firms to collect and use customer data have

raised the concerns of consumers, regulators, and privacy advocates (see,

for instance, Tene and Polonetsky, 2012; Tode, 2013).

In this paper, we capture the most important features of today’s mobile

marketing data landscape, in which competing firms make personalized

offers to consumers based on their physical location, and firms can

(costlessly) collect or acquire data on other consumer attributes that are

(imperfect) signals of consumer flexibility. We augment the traditional

model of spatial competition, in which firms’ locations are given

exogenously, to two-dimensional consumer heterogeneity. In addition to

the standard differentiation parameter, which we interpret literally as the

physical location of consumers between the two competing firms, we

allow consumers to differ in flexibility measured by their transport cost

parameters. This modeling approach takes into account the fact that different

consumers at the same location might respond differently to discounts. Our

model realistically assumes that firms hold perfect data on the physical

locations of consumers and can acquire data of imperfect quality on

consumer flexibility. With both location and flexibility data at hand, a

firm can partition consumers at each location into groups based on their

flexibility. A firm might price discriminately between different groups, but

not within the same group. The quality of the flexibility data is higher

if they allow for the identification and targeting of a narrower group of

consumers based on transport cost parameters, with perfect-quality data

giving rise to first-degree price discrimination. We investigate how profits

change as the quality of the available flexibility data improves. We also

1See, for example, Panzarino (2012) and Small (2014). Section IV of IAB (2015) and Mobile

Marketing Association (2015, p. 6) list examples of the type of information marketing firms

combine location data with.

C

The editors of The Scandinavian Journal of Economics 2017.