The Role of Coordination Bias in Platform Competition

• Author: Yaron Yehezkel,Hanna Hałaburda
• DOI: http://doi.org/10.1111/jems.12163
• Published date: 01 April 2016
• Date: 01 April 2016

The Role of Coordination Bias in Platform

Competition

HANNA HAŁABURDA

Bank of Canada, Currency Department, Ottawa, ON, Canada; and

NYU-Stern, Department of Management and Organizations, New York,NY

hhalaburda@gmail.com

YARON YEHEZKEL

Faculty of Management, Tel-Aviv University P.O.Box 69978 , Tel-Aviv, Israel

yehezkel@post.tau.ac.il

This paper considers platform competition in a two-sided market that includes buyers and sellers.

One of the platforms benefits from a favorable coordination bias in the market, in that for this

platform it is less costly than for the other platform to convince customers that the two sides will

coordinate on joining it. We find that the degree of the coordination bias affects the platform’s

decision regarding the business model (i.e., whether to subsidize buyers or sellers), the access

fees, and the size of the platform. A slight increase in the coordination bias may induce the

advantaged platform to switch from subsidizing sellers to subsidizing buyers, or induce the

disadvantaged platform to switch from subsidizing buyers to subsidizing sellers. Moreover, in

such a case the advantaged platform switches from oversupplying to undersupplying sellers, and

the disadvantaged platform switches from undersupplying to oversupplying sellers.

1. Introduction

When platforms compete in a two-sided market, each platform’s attractiveness to cus-

tomers depends on intrinsic quality and its ability to convince customers that it can

attract the other side as well. The market may have a “coordination bias in favor or

one,” advantaged, platform when customers expect that the other side joins it rather

than the other, disadvantaged, platform. In this paper, we analyze how the degree of

coordination bias affects platformsˆ

a strategies and equilibrium outcome.

In many markets, we observe that platforms face varying degrees of coordination

bias. For example, at the time Apple launched the iPhone 5, most application developers

have not yet developed new applications that can support the new features of the device

(such as the wider screen, for example). Nevertheless, the iPhone 5 preorders topped

2 millions, only 1 day after its launch, and analysts predicted that 50 million users would

For helpful comments, we thank Talia Bar, Alessandro Bonatti, Marc Bourreau, Ramon Casadesus-Masanell,

Michael Cusumano, Jim Dana, Geoff Dunbar, Kobi Glazer, Avi Goldfarb, Tanjim Hossain, Doh-Shin Jeon,

Bruno Jullien, Laurent Linnemer,Jean Tirole, Thomas Tregouet, Patrick Rey,Dick Schmalensee, Yossi Spiegel,

Glen Weyl, and participants at HBS Strategy Conference, IO workshop in Tel-Aviv University and Bar-Ilan

University, marketing seminar at Tel-AvivBusiness School, and Searle Center Conference on Internet Search

and Innovation at Northwestern University, CREST, EARIE, IIOC, The Conference on the Economics of ICT

in Paris, the conference on The Economics of Intellectual Property, Software and the Internet in Toulouse,

and ZEW Conference on the Economics of Information and Communication Technologies in Mannheim.

Hałaburda is grateful to the HBS Division of Research. Wegratefully acknowledge financial support from the

United States-Israel Binational Science Foundation (BSF), the NET Institute (www.netinst.org),and the Henry

Crown Institute of Business Research in Israel.

C2016 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume25, Number 2, Summer 2016, 274–312

Coordination Bias in Platform Competition 275

buy the new smartphone within 3 months of its launch (Faughnder and Satariano, 2012).

This suggests that at the time, the market expected that the iPhone 5 will attract users

and developers. In other words, Apple benefited from a strong coordination bias. New

operating systems for tablets and smartphones, such as Microsoft’s WindowsPhone and

RIM’s Blackberry 10, may not benefit from such favorable coordination bias. Analysts

report that some application developers doubt Microsoft can catch up to Google and

Apple (Ovide and Sherr, 2012). Consequently, developers prefer to focus their energies

on developing new applications for Apple and Android. Obviously,such disadvantaged

position makes it difficult for Microsoft to attract users as well, as users may expect that

developers are more likely to continue focusing on Apple and Android. RIM is facing

a similar problem: when launching its new operating system, the Blackberry 10, some

application developers where skeptical concerning its ability to attract users, making it

unattractive to developers as well.1As one developer remarked, “If this is a horse race,

RIM is two laps behind and has a lame leg.” This is again a problem for RIM, when competing

to attract users. As RIM’s vice president remarked: “If we launch without applications, well,

it will be slow.” Similarly, in the battle between HD DVD and BluRay, it mattered not only

which format provides better experience of high-definition movies (one of the sides of

the market), but also how many movies would be released by the movie studios in a

given format (the other side of the market).

We can think of coordination bias advantage as a more recognizable brand name

or better track record in the past. Although these may come from higher quality offered

in the past, they do not need to be related to the quality offered currently. In some cases,

a platform can be disadvantaged in market coordination bias even though it offers a

product of equal, or even superior, quality than that of an advantaged platform. For

example, the hardware of Blackberry’s new smartphones, running on Blackberry 10

operating system, received very good reviews from market analysts.2However, the lack

of applications, current and expected, is crucial. A reviewer commented that: “4 months

after launching, BB10 [Blackberry 10] is still struggling to attract developers of quality

apps”(Bunton, 2013). This evidence suggests that the main competitive disadvantage of

Blackberry is coordination bias, and not the quality of its hardware.

In their competition to attract the two sides of the market, platforms should take

into account the degree of their coordination bias, as it may affect a platform’s strategic

decisions. In particular, a platform that enjoys favorable coordination bias, needs to

identify how to translate it into competitive advantage over its rival. Likewise, a platform

that suffers from unfavorable coordination bias needs to identify how to choose its prices

in order to overcome its competitive disadvantage and win the market, and under which

conditions winning the market is profitable.

The main research question of this paper is: How the coordination bias affects the

platform’s pricing strategies? Platforms usually compete by setting different prices to

the two sides of the market. In particular, a platform may offer a low, perhaps negative

price to one of the sides, and then charge a high price to the other side. For example,

videogame consoles like Xbox or PlayStation, often sell at a loss in retail, but they make

profits by charging the game developers who sell games to be played on the consoles.

1. Austen (2012). It is also the source of all quotes in this paragraph.

2. PC Magazine compared between the technical specifications of the Blackberry Z10 smartphone and the

iPhone 5 and found that “...it’s a pretty close match” (Colon, 2013). In another review,the Z10 is said to have

a faster processor,higher screen resolution and longer battery life than the iPhone 5 (Holly,2013). In a review

of another Blackberry’s smartphone, the Q10, it is said that the Q10 has a comfortable keyboard and design,

with a superior battery and signal strength than the iPhone (Bunton, 2013).

276 Journal of Economics & Management Strategy

In contrast, Samsung offered prizes to support application developers in developing

applications to Samsung’s new tablets.3Likewise, Microsoft hosted more than 850 ses-

sions worldwide to coach application developers on Windows Phone, as well as other

means for helping developers to profit from developing applications for its new operat-

ing system. Although other firms were also known to offer some support to developers,

the scale of Microsoft’s efforts is unprecedented (Ovide and Sherr, 2012). We therefore

specifically ask how the coordination bias affects the platforms’ pricing strategies in

terms of (i) the side to attract and (ii) the number of sellers to attract. Notice that we raise

this question for both the platform with favorable coordination bias and the platform

with unfavorable bias. This is because, as we show, a platform with unfavorable coordi-

nation bias can still win the market if it has sufficiently high quality, and if the platform

correctly chose its pricing strategies in accordance with the coordination bias against it.

To answer this question, we consider a model with the following features. There

are two sides of a market, buyers and sellers, which cannot interact without a platform.

Once they join a platform, sellers compete among themselves for buyers. This means

that sellers can make positive profit from joining a platform only if buyers indeed joined

the same platform and only if not too many other sellers joined the platform.4

There are two platforms that compete by setting access fees to sellers and buyers.

The platforms differ in two respects. First, they may differ in the quality. We allow for

cases where either platform is of higher or lower quality than the other. Second, one

platform benefits from a favorable coordination bias of the market, in that it is less costly

for it to convince the two sides that all agents will coordinate on joining it. The degree

of coordination bias in our analysis varies along a continuous spectrum, such that it can

be weak or strong. As the degree of coordination bias increases, it becomes cheaper for

the advantaged platform to convince the customers that both sides would join it. We

analyze how the firms’ strategies and the market outcome are affected by the degree of

coordination bias.

We find that the platforms’ pricing strategies can be characterized by two distinct

business models. When neither of the platforms enjoys a favorable coordination bias

and the sellers’ fixed costs are very low, then both platforms will choose a business

model that relies on the revenue from the buyers while fully subsidizing the sellers’

fixed costs. We refer to it as a buyers-pay business model. When, however, the sellers’

fixed costs are very high, both platforms adopt a business model that relies mainly

on the revenues from sellers, while competing to attract buyers—a sellers-pay business

model. For the case where sellers have intermediate values of the fixed costs, and the

degree of coordination bias is high enough, the advantaged platform adopts a sellers-

pay business model, whereas the disadvantaged platform adopts a buyers-pay business

model. Therefore, an increase in the degree of coordination bias can have two effects

on the platforms’ business models. First, it may motivate the advantaged platform to

switch from a buyers-pay to a sellers-pay business model. Second, it may motivate the

disadvantaged platform to switch from a sellers-pay to a buyers-pay business model. This

is because the profitability of a sellers-pay business model relies on the platforms’ ability

to attract buyers, which in turn relies on the degree of coordination bias toward the

advantaged platform. We also find that a platform’s business model affects not only

3. See: http://www.smartappchallenge.com/eng/challenge/awardsJudging.do

4. Our results depend on such an asymmetry of the two sides. This asymmetry is common in many real-

life two-sided markets. However, there are also two-sided markets to which our model does not apply, for

example, online dating.