Selling through referrals

Published date01 October 2018
DOIhttp://doi.org/10.1111/jems.12251
AuthorVasiliki Skreta,Daniele Condorelli,Andrea Galeotti
Date01 October 2018
669
J Econ Manage Strat. 2018;27:669–685. wileyonlinelibrary.com/journal/jems © 2018 Wiley Periodicals, Inc.
Received: 19 February 2016 Revised: 25 December 2017 Accepted: 13 April 2018
DOI: 10.1111/jems.12251
ORIGINAL ARTICLE
Selling through referrals
Daniele Condorelli1Andrea Galeotti1,2 Vasiliki Skreta3
1Department of Economics, University of
Essex (Email: d.condorelli@gmail.com;
andrea.galeotti@eui.eu)
2Department of Economics, European Univer-
sity Institute
3Department of Economics, UT Austin &
UCL (Email: vskreta@gmail.com)
Abstract
We endogenize intermediaries' choice to operate as agents or merchants in a market
where there are frictions due to asymmetric information about consumption values. A
seller has an object for sale and can reach buyers only through intermediaries. Inter-
mediaries can either mediate the transaction by buying and reselling—the merchant
mode—or refer buyers to the seller for a fee—the referral mode. When the seller has
a strong bargaining position and can condition the asking price to the intermediaries'
business model choice, all intermediaries specialize in agency. The seller's and inter-
mediaries' joint profits equal the seller's profits when he has access to all buyers. When
the seller does not have such bargaining power, the level of the referral fee and the
degree of competition among intermediaries determine the business mode adoption.
A hybrid agency–merchant mode may be adopted in equilibrium. Banning the referral
mode can decrease welfare because the merchant mode is associated with additional
allocative distortions due to asymmetric information.
1INTRODUCTION
An intermediary is an economic agent that facilitates the connection between supply and demand. Intermediaries purchase from
suppliers and resell to customers or facilitate the meeting of buyers and sellers. For example,online travel agents such as Expedia,
Priceline, and Orbitz connect buyers to hoteliers, airlines, and car rental companies. Amazon marketplace, Groupon, and alike,
connect buyers to retailers selling a variety of goods. Amazon and Apple intermediate the E-books industry. Intermediation is
also important in off-line markets and beyond traditional retailing. For instance, it plays a keyrole in trading real estate, ar t, used
cars, books, as well as in markets for professional services.1
Intermediaries predominantly operate under two business models — or hybrids of those. Under the referral mode (or
agency mode), they refer buyers to sellers, who then negotiate directly on the terms of trade. In return, intermediaries receive
referral fees for the creation of the match and/or commissions based on sales. Underthe more traditional retailer/merchant mode,
intermediaries buy goods from suppliers for resale to consumers.
In the online travel industry, Priceline makes most of its revenue through the referral mode and the rest from acting as a
merchant. Priceline's subsidiary Booking.com is an agency-based business, whereas its subsidiary Agoda is a merchant-based
business.2Expedia operates mainly under the merchant mode receiving roughly 75 percent of its revenue through the merchant
We are grateful to Ilari Paasivirta, Alkis Georgiadis, and Angelos Diamantopoulos for excellent research assistance. We thank Luis Cabral, Alex Gershkow,
Panos Ipeirotis, Navin Kartik, Ilan Lobel, Phil Reny, and Jidong Zhou for useful conversations and comments. We thank seminar participants at the 12th
Columbia/Duke/MIT/Northwestern IO Theory Conference at Duke, Brown University,CERGE-EI, Columbia University, LSE, New York University(Ster n),
Northwestern University, PennState University, PSE, University of Arizona, University College London, Universityof Essex, University of Surrey, University
of Guelph Workshop on Auction Theory,University of Helsinki, University of Leicester, University of Warwick, and WashingtonUniversity. Andrea Galeotti
is grateful to the European Research Council for support through ERC-starting grant (awardno. 283454) and ERC-consolidator grant (award no. 724356) and to
The Leverhulme Trust for support through the Philip LeverhulmePr ize. VasilikiSkreta is grateful to the European Research Council (ERC) consolidator grant
“Frontiers in Design.”
2JOURNAL OF ECONOMICS & MANAGEMENTSTRATEGY
670
mode and some 21 percent through the referral mode.3Expedia expanded its business by acquiring the agency-based online
hotel business Venere. Orbitz's net revenue stems fairly evenlyfrom its air and hotel businesses (34 and 29 percent respectively),
with its revenue from the hotel business coming mainly from the merchant mode.4
In this paper, we explicitly model the choice of an intermediary to operate as a merchant or a reseller.5Beyond providing
insights on the equilibrium structure of the industry, we also analyze the effects of the two models on market outcomes and
welfare. Finally,we draw some implications for competition policy. Crucially, weconsider markets where both options, to refer
or to resell, are available to intermediaries and not controlled by sellers.
An intermediary who chooses its business model faces the following trade-off, whichis at the core of our analysis. By operating
as an agent, the intermediary's revenue only depends on the number of consumers he refers to the seller and on the referral fee.
There is no uncertainty, nor a dependence on the market structure. Instead, by becoming a merchant, the intermediary makes a
risky profit that depends on the difference between the expected cost and revenue of buying and reselling. These variables, in
turn, depend on the market structure, not just on the number of buyers, and on the protocols that regulate trade upstream and
downstream.
We develop a simple model that captures some of the forces shaping intermediaries' decisions to operate as merchants or
agents, and allows us to perform comparative statics. In our baseline model, there is a seller with one object for sale. The
seller is in contact with a number of intermediaries, each intermediary has access to a subset of buyers, and each buyer has
private information about his valuation. The seller and the intermediaries have no consumption value for the object. The inter-
action between the seller, the intermediaries, and the buyers is captured by a three-phase game. In the business-model choice
phase, each intermediary decides whether to become an agent or a merchant. An agent-intermediary refers all her buyers to
the seller in exchange for a referral payment. In the trading phase, the choices of intermediaries become public, the seller
sells to referred buyers and merchant intermediaries. If a referred buyer obtains the object, the game ends. If a merchant inter-
mediary obtains the object, then we enter the resale phase: the merchant intermediary resells to the buyers he is in contact
with.
Our model captures salient features of severalindustries. One not able exampleis the market for hotel rooms. First, hotel rooms
are limited in number, and supply cannot be increased in the short term. Second, hotel rooms are extremely differentiated (by
city area, comfort level, etc.) and therefore their sellers may enjoy nonnegligible market power. Third, multiple intermediaries
operate in the market and both intermediaries and sellers can potentially conduct the sale. As we have mentioned earlier, both
merchant intermediaries, such as Expedia, and agents, such as Booking, operate in the market. Finally,as in our model, it is costly
for consumers to look for hotels without shopping via an intermediary, and likewise hotels have no mean to directly advertise
to potential buyers in a cost-effective way.6
Our first result (Theorem 1) shows that when the seller has full bargaining power and can choose the selling procedure,
regardless of the level of referral payments, in the unique equilibrium of the game intermediaries adopt the referral model. The
prevalence of the referral mode is a consequence of the seller's ability to tailor the minimum price to the intermediary's business
model choice. Indeed, the seller anticipating the resale value of a merchant intermediary never sets the minimum reserve price
below that value. Thus, merchants enjoy no rent from buying and reselling. As a consequence, adopting the referral mode is
strictly preferable.
Because referrals operate frictionlessly, the industry's specialization in agency implies that the aggregate equilibrium profit
of both the seller and the intermediaries equals the expected seller's profit when he has direct access to all buyers—what we
call the “integrated-industry profit.” Hence, in equilibrium mis-allocation distortions arising from intermediation via resale are
eliminated. In particular, there is no risk that the “wrong” intermediary acquires the object and resells it to a low-value buyer,
thereby excluding out of the market high-value buyers who are clients of different intermediaries.
When all intermediaries adopt the referral mode, not only industry profit but welfare, including consumer surplus, is unam-
biguously higher than in the case in which some intermediaries adopt the merchant mode. This follows from the observation in
the previous paragraph and from the fact that both the seller and a merchant intermediary choose the same reserve price.7
This finding contributes to the industrial-organization literature on vertical restraints and provides additional insights into a
policy-relevant issue.8More precisely, we provide a market-based rationale for the adoption of a referral mode and show that,
ultimately, consumers may benefit from such a vertical agreement. Because higher prices in our setup reflect a more efficient
allocation, buyers may benefit even in cases in which the final price ends up being higher than under the merchant mode. This
observation could be relevant for competition policy in view of the increasing number of intermediaries that operate as agents,
and the new challenges that intermediated markets pose for competition authorities.
As a concrete example, consider the United States v. Apple Inc. case. In 2012 the U.S. Department of Justice (DOJ) filed a
lawsuit against Apple and a number of major publishers for an alleged conspiracy, having the objective to raise e-book prices
(see Gaudin & White, 2014). At the core of the case is a shift of the e-book industry toward the referral mode. Whereas before

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT