Managing buzz

• Author: Jiwoong Shin,Arthur Campbell,Dina Mayzlin
• DOI: http://doi.org/10.1111/1756-2171.12173
• Date: 01 March 2017
• Published date: 01 March 2017

RAND Journal of Economics

Vol.48, No. 1, Spring 2017

pp. 203–229

Managing buzz

Arthur Campbell∗

Dina Mayzlin∗∗

and

Jiwoong Shin∗∗∗

Wemodel the incentives of individuals to engage in word of mouth (or buzz) about a product, and

how a firm may strategicallyinfluence this process through its information release and advertising

strategies. Individuals receive utility by improving how others perceive them. A firm restricts

access to information, advertising may crowd out word of mouth, and a credible commitment not

to engage in advertising is valuable for a firm.

1. Introduction

Word of mouth is an important driverof consumers’ purchase decisions.1However, despite

its importance, the existing literature typically treats word of mouth as a costless and mechanical

process and offers little insight into an individual’s incentive to engage in word of mouth. In this

article, we introduce a framework to model how consumers’ concerns about social status drive

their incentives to engage in word of mouth. We focus on a particular reputational motive—word

of mouth as “self-enhancement” (Baumeister, 1998) or the idea that an individual engagesin word

of mouth to improve howshe is perceived by the listener. In contrast to a setting where consumers

mechanically undertake word of mouth, we show that a firm may improve the incentives for

individuals to engage in word of mouth by restricting early access to product information.

∗Monash University; arthur.campbell@monash.edu.

∗∗University of Southern California; mayzlin@marshall.usc.edu.

∗∗∗Yale University; jiwoong.shin@yale.edu.

We thank Dirk Bergemann, Alessandro Bonatti, Odilon Camara, Florian Ederer, Johannes Horner, Yuichiro Kamada,

Duncan Simester, Birger Wernerfelt, and seminar participants at Brown University, CKGSB, Columbia University,

Indiana University,MIT, Monash University,Paris School of Economics, Purdue University, UC Berkeley,University of

Cambridge, University of Florida, University of Melbourne, University of Toronto, Wharton, Yale University, the 2012

SICS Conference at Berkeley,and the 2011 Conference on the Economics of Advertising in Moscow.

1Word of mouth has been shown to affect purchasing behavior in restaurant choices (Luca, 2016), book sales

(Chevalier and Mayzlin, 2006), banking (Keaveney, 1995), entertainment (Chintagunta, Gopinath, and Venkataraman,

2010), technological products (Herr, Kardes, and Kim, 1991), and appliances and clothing (Richins, 1983). These studies

are also consistent with recent industry research: for example, according to Wordof Mouth Marketing Association (2011),

54% of purchase decisions are influenced by word of mouth. Also, “Word of mouth is the primary factor behind 20 to 50

percent of all purchase decisions” in McKinsey Quarterly (Bughin, Doogan, and Vetvik, 2010).

C2017, The RAND Corporation. 203

204 / THE RAND JOURNAL OF ECONOMICS

As a real-life illustration of our finding that firms may benefit from actively restricting

early access to product information, consider the US launch of the European music streaming

site Spotify in July 2011. At first, Spotify’s free US version was available by invitation only.

Interestingly, obtaining an invitation was nontrivial, and direct invitations werelimited to cer tain

groups: consumers could receive one either through current users or through other channels. For

example, the company sent invitations to users whointeracted with Spotify on Twitter, and Coca-

Cola gave out invitations to users whosubmitted their email address.2After a few weeks, anyone

could download the free version of Spotify through the company’s website.3By November 2011,

Spotify was able to attract four million users, although it undertook almost no advertising. Media

sources speculated that the initial exclusivity surrounding the site contributed to early buzz and

high adoption rates.4

The Spotify example is by no means unique. Many marketing practitioners recommend and

use similar strategies, which limit access to information in order to spur word of mouth. For

example, Hughes (2005) states, “Sometimes withholding can work better than flooding. Limit

supply and everybody’s interested. Limit those in the know of a secret, those not in the know

want the currency of knowing—they want to be part of the exclusive circle.” David Balter, the

founder of the buzz marketing firm BzzAgent, considers exclusivity to be one of the necessary

ingredients for a successful word-of-mouth campaign. “Exclusivity is the velvet rope of social

media: everyone wants to be special enough to be on the right side of it.”5Also, Sernovitz

(2011) observes that, “Many people are more likely to talk about a product if there is some

kind of insider access or privileged status” and provides a number of examples where firms

use exclusivity strategy to increase word of mouth about their products. For instance, retailers

sometimes offer private shopping hours for their select customers the night before new products

are available to the public, and software companies send prerelease versions of new software to

active message board users. Our model explains why purposely limiting the number of people

who are initially exposed to product information may in fact maximize overall product adoption.

We develop a model where consumers meet one another at a Poisson rate over time. Con-

sumers are heterogeneous in that they are either high or low type. The most straightforward

interpretation of high type here is being knowledgeable about a particular product area or having

category expertise: for example, having good taste in wine, being technologicallysavvy, knowing

the best restaurants and bars, or having good taste in music. The key element of the model is that

the utility an individual receives during a social interaction is an increasing function of her peer’s

belief that she is the high type. Prior to meeting others, individuals choose whether or not to

undertake a search to acquire information about the firm’s product at a certain individual-specific

cost. Then, during each social interaction, individuals decide whether or not to engage in costly

word of mouth. We focus on a signalling equilibrium where word of mouth about a product serves

as a credible signal of high type. The central focus of our analysis is how the firm can manage

the extent of the information diffusion in this context.

Webroadly consider two types of strategies bythe fir m. First,we consider information release

strategies where the firm imposes differential costs for information acquisition on different types

of consumers. When the costs of acquiring information for the low type are high enough, there

exists a signalling equilibrium where individuals acquire information and then pass it on through

2news.cnet.com/8301-13845_3-20081418-58/get-a-quick-and-easy-invitation-to-spotify/

3One possible reason for a firm’sinitial limited release could be a beta version of the product in a test market for

the purposes of collecting feedback from users about the product’sfunctionality before its wide release. This explanation

is less applicable to the Spotify case, given its presence and operational volume in Europe bythe time it launched in the

US market in 2011—Spotify had already become the most popular service of its kind in the world; it had 1.6 million paid

subscribers and more than 10 million registered users in total (www.nytimes.com/2011/07/14/technology/spotify-music-

streaming-service-comes-to-us.html).

4See, for example, “Spotify’s ascension can be largely attributed to word of mouth” (http://www.theverge.

com/2013/3/25/4145146/spotify-kicks-off-ad-blitz-as-rumors-hint-of-video-service).

5blog.hubspot.com/blog/tabid/6307/bid/6183/4-Social-Media-Methods-for-Generating-Word-of-Mouth.aspx

C

The RAND Corporation 2017.