Advertising as a search deterrent

• Author: Chengsi Wang
• DOI: http://doi.org/10.1111/1756-2171.12209
• Date: 01 December 2017
• Published date: 01 December 2017

RAND Journal of Economics

Vol.48, No. 4, Winter 2017

pp. 949–971

Advertising as a search deterrent

Chengsi Wang∗

This article examines a monopoly firm’s incentive to disclose information through advertising

when consumers can choose between buying immediately and searching for additional informa-

tion. Because sales drop when search reveals low match values to consumers, the firm has an

incentive to deter search. We show that partial information disclosure emerges as a useful tool

for search deterrence when search costs are low. Informative advertising and consumer search

can be viewed as complements in producing information. Although transparency policies reduce

search expendituresand improve purchase decisions, whether they are sociallydesirable depends

on the magnitude of search costs.

1. Introduction

In many markets, especially those characterized by rapid product innovation, prices, sales,

and the induced efficiency largely depend on the availability of product information. Such infor-

mation can be produced by both sides of the market through advertising and search. Firms tend to

have better knowledge about their products than consumers, and through informative advertising

they can reveal the information to guide consumers towardbetter purchase decisions. Consumers

can engage in various searching activities to acquire additional information. For instance, they

can call the company’s service line to make inquiries, go to the brick-and-mortar store to inspect

the product, attend a promotional campaign to obtain a free trial, or spend time reading reviews

written by other consumers. Whether consumers engage in such searching activities depends on

both the price and the information revealed by advertisements (hereafter, ads). In turn, firms’

advertising choices are constrained by the search activities. The aim of this article is to study the

interaction between informative advertising and consumer search in a market equilibrium and its

implications for advertising precision, price, and welfare.

Advertising helps inform consumers about the match between their tastes and the product

attributes. Although most theoretical work predicts that firms selling horizontally differentiated

∗University of Mannheim and Monash University; chengsiwang@gmail.com.

An earlier version of this article was titled “Informative advertising, consumer search, and transparency policy.” I am

grateful to the Editor, Mark Armstrong, and three anonymous referees for their valuable comments and suggestions. I

would like to thank Hodaka Morita and Carlos Pimienta for extensive advice and support. Formany helpful comments,

I also thank Simon Anderson, Malin Arve, Heski Bar-Isaac, Jay Pil Choi, Curtis Eaton, Heiko Gerlach, Arghya Ghosh,

Stephan Lauermann, Marco Haan, Maarten Janssen, Flavio Menezes, Andras Niedermayer, Martin Peitz,Regis Renault,

Nicolas Schutz, Jianfei Shen, Ralph Winter, John Wooders, and Julian Wright as well as participants at various seminars

and conferences for their comments. I acknowledge the financial support from the Deutsche Forschungsgemeinschaft

through SFB-TR 15. All remaining errors are my own.

C2017, The RAND Corporation. 949

950 / THE RAND JOURNAL OF ECONOMICS

products prefer either perfect disclosure or no disclosure of product information (see Lewis

and Sappington, 1994; Johnson and Myatt, 2006; Ivanov, 2013), there is abundant evidence

from marketing research showing that most real-world advertising only partially reveals product

information.1In this article, we show that, when the cost of making purchases is negligible

but the cost of acquiring product information remains significant, such as in online markets, a

firm may partially disclose information in order to induce consumers to buy more “quickly.”2

Moreover, in contrast to the common view that advertising and search are substitutes, we show

that advertising and search can be seen as complements.3Finally, our model also sheds light on

the welfare implications of transparency policies. Imposing such policies may reduce consumer

surplus when search costs are relatively low.

We consider a monopoly firm who sells products with value vto consumers, which is ex

ante unknown to both parties. The firm first chooses how much information to include in the

ads, along with a price p. Consumers then decide whether to buy immediately or search for

more information. Search is costly but allows consumers to be perfectly informed about v.A

more informative advertising enables consumers to formulate a more precise estimate about v.

This information transmission process is captured by a signal realization xthat each consumer

independently observes after seeing the ads. Consumers who observe a high xrevise upward

their expectation about vand are likely to make immediate purchases. Other consumers revise

downward their expectation and might search for additional information or leave the market

immediately. The consumers who choose to search will buy only if v≥p. The setup can be more

broadly interpreted. For example, consumers may choose between buying a book from an online

bookshop and from a brick-and-mortar store. Although traveling to the brick-and-mortar store

is costly, consumers can better evaluate the product by physically inspecting it (see Loginova,

2009).4

When production is costless (costly production is considered in Section 6), the ideal situation

for the firm is to keep all consumers completely uninformed and fully extract surplus by selling

at the price equal to E(v). Consider the trade-off that the firm faces when deciding how much

information to disclose via ads. If the ads perfectly reveal vto consumers, those consumers with

v<pdo not buy. However, if the ads convey no information and search costs are low enough,

consumers will search and become perfectly informed about v, which leads to the same outcome.

The firm needs to design an advertising and pricing scheme to reduce consumers’ temptation to

search and induce them to buy immediately at a relatively high price.

One way to lower consumers’ search inclination is to reduce price. Facing a lower price,

the risk of making a “bad” purchase decision is reduced and the willingness to buy immediately

increases. This strategy is optimal to the firm when search costs are not too low, because a small

reduction in price is sufficient to induce all consumers to buy immediately without searching.

However, using price reduction to deter search is no longer appealing when search costs are

very small, because the price reduction has to be significant. In this case, the firm can partially

deter search by revealing some, but not all, information to consumers. Partial disclosure reduces

the search temptation of those consumers who hear “good news” (high x). These consumers

are willing to buy immediately. Consumers who hear “bad news” (low x) will still participate

1Abernethy and Butler (1992) found that, in US television advertising, the mean number of “information cues”

is 1.06, with only 27.7% having two or more cues, and 37.5% havingno cues. For newspaper ads, they found that only

39.6% have four or more cues.

2There could be many nonstrategic reasons that ads only partially conveyinformation. For example, firms may be

constrained by advertising space, or consumers may only have limited attention spans. We consider only the strategic

origins for partial disclosure.

3Previous workon adver ting in homogeneousproduct markets has examined this relation. Janssen and Non (2009)

showed that higher advertising intensity crowds out consumer search, whereasRober t and Stahl (1993) found that high

advertising intensity coexists with high search intensity.

4Another interpretation is to think of a situation in which the consumer’s choice is between making an advanced

purchase and making a late purchase. Although the late purchase causes delayed consumption, it allowsthe consumer to

resolve demand uncertainties (see Bhatt, 1989).

C

The RAND Corporation 2017.