Combating incumbency advantage of network effects: The role of entrant’s decisions and consumer preferences

Abstract

This article studies the competitive dynamics between an incumbent and a late entrant in a market governed by network effects. Classifying markets based on decision criterion used by majority of consumers in a market, an analytical model for the likelihood of incumbent displacement in these markets is developed and corroborated using simulations. Using product quality offered and product price charged by the entrant as managerial levers, the impact of managerial decision-making on the expected market share of the entrant is discussed. Efficacy of seeding and delayed entry by the entrant in an attempt to perfect its product is also examined.

Keywords

Network effects product quality product price consumer heterogeneity late entry

Introduction

Easy and cheap access to the Internet has led to the proliferation of high-tech products that are amenable to network effects, that is, the attractiveness of these products¹ rises as more and more consumers adopt them. As a result in the case of such products, the firm to enter the market first, that is, the first mover, is likely to enjoy the benefits of early entry. While an extensive body of research is devoted to various issues associated with the first mover advantage (see Lieberman & Montgomery, 2013, for an excellent overview), very little work examines the potential of success for late entrants (Besharat, Langan, & Nguyen, 2015; Fosfuri, Lanzolla, & Suarez, 2013). Furthermore, how consumer preferences, that is, the dimensions that govern consumers’ purchase decisions (Clark & Chatterjee, 2015; Tellis, Yin, & Niraj, 2009), affect the competitive dynamics between the first mover or incumbent firm and the late entrant firm have not been sufficiently explored. Therefore, in this research, we model the product decisions of the entrant, with respect to quality offered and price charged relative to the incumbent (Lee & O’Connor, 2003), in a market for a product amenable to network effects and study how these decisions impact the market share garnered by the entrant when the market comprises consumers with heterogeneous preferences. We also contribute to extant strands of literature on seeding (Dou, Niculescu, & Wu, 2013) and timing of entry (Schilling, 2002) in markets with network effects, by incorporating the effect of consumer preferences on the efficacy of late entry.

Conventional wisdom suggests that navigating in markets with network effects is essentially characterized by two thumb rules—getting in early and getting big fast (Halaburda & Oberholzer-Gee, 2014). The pioneer or the first mover advantage refers to the phenomenon in which these organizations can get competitive advantage by virtue of being the first to market. This could be due to a variety of factors including acquiring expertise, technological leadership, or control of distribution channels (Finney, Lueg, & Campbell, 2008); gaining influence on customers (Carpenter & Nakamoto, 1989); or establishing entry barriers as well as switching costs (Gao & Knight, 2007). Despite the persistent interest among management scholars on issues related to the first mover advantage, an upcoming stream of literature has started investigating the contingencies that allow the late entrant to enjoy advantage over the first mover (Besharat et al., 2015; Fosfuri et al., 2013). Even big companies have struggled to compete against the fist mover advantage afforded by network effects. For example, despite the technological prowess and deep pockets, Google was unable to compete against the dominant social network, Facebook, which enjoyed first mover advantage. Eventually, Google decided to shut down Google plus amid reports of low consumer adoption.² On the contrary, one of the most celebrated cases in recent times is the rapid growth enjoyed by telecom player Jio in the Indian market. Jio was able to create disruption in the marketplace displacing a number of well-established players. Its playbook was so successful that it plans to employ the same strategy in the broadband market as well. Similarly, the Chinese-based UC browser has been able to make a dent in South Asian countries in spite of being a very late entrant.³ Considering that these late entrants are far more plentiful than the pioneers, it is all the more important that management scholarship analyzes various facets associated with the success (and failure) of such organizations. In particular, research on managerial decisions with respect to market entry by late entrants could guide both research and practice in understanding how these decisions affect the survival/success rates of these late entrants.

In markets that are governed by network effects, the subscribed user base, the quality of entrant’s product relative to incumbent’s product quality, and the price of entrant’s product relative to incumbent’s product price not only affect product adoption directly but also through interaction effects (Lee & O’Connor, 2003). While quality and price are two important managerial decision points with respect to product attributes at the time of market entry, their interaction with network effects can lead to nontrivial competitive dynamics, especially due to heterogeneity in consumers’ decision-making with respect to product adoption (Ge, 2002; Tellis et al., 2009). Surprisingly, literature has hitherto paid little attention to the prospective consumers’ heterogeneous preferences and decision-making behavior, which play a critical role in a successful market entry (Sharma, Gupta, & Sarkar, 2015). The accentuating (or inhibiting) role of behavioral preferences of the target market on the success of market entry in the presence of interaction effects among product attributes and subscribed user base is the subject of this research.

First, this article mathematically models a limiting case of a market with infinite population to study the expectation of incumbent displacement for different values of relative product quality offered, and relative product price charged by the entrant, in markets comprising consumers with heterogeneous preferences. Consumers’ preferences are modeled with respect to their preference for product quality, product price, and network effects. Thereafter, a similar analysis is performed for markets with finite population, comprising of consumers with heterogeneous preferences for product quality, product price, and network effects. Since the finite population case is sensitive to initial conditions, it is important to understand the variance associated with the expectation of incumbent displacement. We, therefore, conduct discrete event simulations that enable an analysis of the probability of the entrant displacing the incumbent in counterfactual markets with heterogeneous consumer preferences.

Through this research, we highlight the role of consumers’ heterogeneity in decision-making related to product adoption and, thereby, suggest that firms must account for the characteristics of their target market before entering with a product. We also discuss how the entrant’s managerial levers of relative product quality to offer and relative product price to charge affect its expectation of displacing the incumbent in markets consisting of consumers with different preferences. In addition, the strategies of (a) seeding and (b) delaying market entry in trying to develop a higher quality product, which might be adopted by an entrant, are discussed.

The rest of this article is organized as follows. Section 2 reviews prior work and we frame a research agenda from extant literature. Section 3 presents a stylized analytic model. To get a more nuanced understanding, a simulation is developed in sections 4 and 5. Section 6 discusses the implications of our findings. Section 7 concludes the article by reviewing the implications for research and practice and offering directions for future work.

Related literature

High-tech markets exhibit the presence of network effects, wherein the utility derived from a product increases as more number of people adopt the product (Katz & Shapiro, 1985). The phenomenon of network effects has intrigued the interests of scholars from strategy, economics, and marketing to explore different facets of the phenomenon. The impact of network effects on the evolution of markets (Gupta, Jain, & Sawhney, 1999), their influence on customer lock-in (Benedek, Lublóy, & Vastag, 2014; Liebowitz & Margolis, 1995), and competition in the market (Katz & Shapiro, 1985; Ohashi, 2003; Shankar & Bayus, 2003) along with their importance in protecting the market position of the incumbent (Barua & Mukherjee, 2011) are some of the streams that have been studied in literature.

The presence of network effects complicates a firms’ decision with respect to timing of entry into the market, wherein the firm needs to balance the risk of premature entry or missed opportunity due to late entry (Klingebiel & Joseph, 2016; Lilien & Yoon, 1990). Each of these strategies is fraught with its own set of benefits and trade-offs²⁴ and it is difficult to determine ex ante the better of the two strategies. However, it is certain that entry decisions become more complex and risky for the followers in markets for products that are amenable to network effects (Ratchford et al., 2009; Tellis et al., 2009). While, on the one hand, network effects provide an additional advantage to the first mover (Lieberman & Montgomery, 2013; Zhu & Iansiti, 2012), research has also shown that network effects significantly decrease the survival duration of incumbents as the benefits derived from an installed base are outweighed by the increase in marginal utility over time and wait-and-watch behavior of customers (Srinivasan, Lilien, & Rangaswamy, 2004). In order to counter the incumbent’s network effects, various strategies are utilized by the entrant, such as seeding the market by freely distributing a certain number of units of their product to start off network effects (Marsden, 2006) or deciding against major product innovation and entering the market as early as possible (Ali, 1994).⁵ However, we believe that extant literature that has looked at strategies that can be used by an entrant when challenging an incumbent with some preexisting users can be extended along three dimensions. We discuss each of these next.

Contextualizing entrant–incumbent competition for different types of markets

It would be naive for managers to consider entering a market without understanding the preferences and decision-making behavior of their potential consumers. Consumer decisions often involve a choice among several alternatives, each described by a set of attributes (Bettman, Johnson, & Payne, 1991; Huang, Leng, & Parlar, 2013; Narasimhan, Mendez, & Ghosh, 1996). Product quality and product price have been identified as important factors influencing consumer decision-making (Levin & Johnson, 1984; Wolinsky, 1983). Research has also suggested that on digital mediums, different types of consumers exhibit different search patterns and that their response to ads is also different (Animesh, Viswanathan, & Agarwal, 2011). It is argued that price-seeking consumers search deep compared to quality conscious consumers who have higher willingness to pay for the same quality (Animesh, Viswanathan, & Agarwal, 2011). It can, therefore, be argued that price-seeking customers are more likely to search for sellers providing a lower price product than those providing a higher price product, and quality-seeking customers are more likely to search for sellers providing a higher quality product than those offering a lower quality product. Furthermore, network effects have also been found to influence consumer decisions for products amenable to such effects, such as telephones, video games, computer hardware and software, and so on (Katz & Shapiro, 1994). A typical market may comprise a population mix with varying proportions of quality-, price-, or network effects-seeking people. Extant theoretical research when modeling competition between entrant and incumbent has considered markets (situations) where equal number of people prefer different levels of quality (Sun, Xie, & Cao, 2004). However, research has shown that consumers from different regions can exhibit varied decision-making styles (Zhou, Arnold, Pereira, & Yu, 2010), implying that the proportion of consumers with heterogeneous preferences that comprise the market and their decision-making style is important. For example, consumers from developing economies and developed economies are likely to be less value conscious (Sharma, 2011). The different sensitivities of consumers to product attributes, say price or quality, and the proportion of consumers who value a product based on different product attributes are, therefore, expected to affect adoption; more so in the case of products that exhibit network effects. We believe that this population mix influences the eventual outcome of market entry for an entrant trying to displace an entrenched incumbent in a market. For example, in a market with higher proportion of quality-seeking individuals, network effects will kick in faster for a higher quality product in comparison to a lower quality one. Therefore, in this study, we explore the following research question:

RQ1: How do markets that exhibit different proportion of people who are sensitive to price, quality, and network effects influence incumbent displacement?

Incorporating entrant’s product pricing and quality decisions in understanding entrant–incumbent competition

While the population mix of the market and network effects is beyond the control of managers (Bental & Spiegel, 1995), price and quality are important levers in the hands of product managers. Product quality has been identified as a source of durable competitive advantage for new products (Cooper & Kleinschmidt, 1987; Dietl, Grutter, & Lutzenberger, 2009; Lawless & Fisher, 1990) and has been found to imply an aspect of comparison with competing products, thereby influencing ultimate success or failure of the product (Calantone & Di Benedetto, 1988). The effect of product quality on growth is significant even after controlling for the size of the installed base (McIntyre, 2011a) and it positively influences market share of high-tech products (Tellis et al., 2009). In fact, product quality and product price are two integral components of the 4 Ps of the marketing mix (McCarthy, 1978). The influence of price on the perceived quality of goods is well established, with higher priced goods being perceived as possessing better quality (see Rao & Monroe, 1989; Völckner & Hofmann, 2007). Product price has been found to significantly influence the effectiveness of market entry strategies such that a late entrant can still displace the incumbent by pricing a less distinctive product at a comparatively higher price point (Besharat et al., 2015).

An emerging strand of literature is focusing on the price and quality decision in markets with network effects (Dietl et al., 2009). It has been found that not only can firms’ pricing strategies strengthen network effects in early stages of product introduction (Ohashi, 2003), but network effects can in turn positively influence the price that firms charge (Chakravarty, Dogan, & Tomlinson, 2006). Similarly, it is argued that a superior technological product can eliminate the benefits of lock-in and network effects, often referred to as technological leapfrogging (Schilling, 2003). A documented example, in this regard, is that of the video-gaming industry where in each generation successful competitors have offered at least three times the clock speed of the fastest system in the previous generation (Schilling, 2003). While it is argued that network effects can enhance the effect of quality of a product, it is also contended that product quality is essential in order to derive benefits of network effects (Etzion & Pang, 2014; Tellis et al., 2009). In the presence of such varied academic stances, it can be argued that there exists a complex interplay of network effects, quality, and price on entry strategy and successful market entry, be it measured in terms of profit, market share, or survival (Lieberman & Montgomery, 2013; Zhu & Iansiti, 2012). Consequently, the quality offered by the entrant and the price charged become important decision variables for an entrant when entering a market with network effects, a particular population mix (proportion of quality-, price-, and network effects-seeking people), and an already entrenched incumbent. Since it is often easier for managers to understand quality and price in terms relative to competition (Calantone & Di Benedetto, 1988), rather than in absolute terms, this research uses the relative product quality of the entrant (to the incumbent) and relative product price of the entrant (to the incumbent) as constructs to model the success probabilities of the entrant. Market entry, therefore, requires a composite strategy that includes deciding product quality and price, considering compromise on quality for a reduced time to market, contemplating market seeding, and tuning the strategy to leverage consumers’ preferences that guide decision-making, all the while accounting for network effects. Specifically, this research seeks answers to the following questions:

RQ2: In a market with network effects and a given population mix, how does the relative product quality offered by the entrant influence incumbent displacement?

RQ3: In a market with network effects and a given population mix, how does the relative product price charged by the entrant influence incumbent displacement?

Combating incumbency advantages with market seeding

Literature dealing with the effect of installed base on enabling an incumbent to counter an entrant remains inconclusive. While it is argued that an incumbent’s installed user base may have preemptive effects on entry similar to that of investment in capacity (Fudenberg & Tirole, 2000), it is also recognized that the installed base does not provide a safety shield to the incumbent in quality-driven markets (Zhu & Iansiti, 2012). In practice, entrant firms often resort to charging a lower relative price (market penetration) or offering a higher relative quality in order to displace the incumbent. Pricing, however, is an important signal of product quality and higher price tends to signal higher quality (Bagwell & Riordan, 1991). Thus, entering a market with low price may send adverse signals about quality. Entrants, therefore, have little choice but to invest in a higher quality product before entering the market. Nevertheless, an alternative available to the entrant that preempts investment in quality and still allows countering network effects is that of offering a few units of its product for free to a certain proportion of the population, often referred to as seeding (Galeotti & Goyal, 2009; Jiang & Sarkar, 2009). Seeding gets the ball rolling and assists in initial product adoption. Seeding and network effects are often considered related phenomena (Lehmann & Esteban-Bravo, 2006), and seeding by the entrant can decrease the strength of incumbent’s network effects. However, even if the entrant adopts a seeding strategy, it is unclear as to what will be the effect of different seeding proportions on the relationship between relative quality offered and relative price charged and probability of incumbent displacement in markets with different population mixes. This is explored by the next research question:

RQ4: In a market with network effects and a given population mix, how does seeding proportion influence incumbent displacement for different product decision of the entrant with respect to relative product quality and relative product

Compromises between product quality and entry time

The limitations of incumbency advantage are often explained through the concept of “creative destruction” (Schumpeter, 1950), that is, an innovative firms displacing the incumbent by entering with higher quality products. However, developing a higher quality product may often entail significant delays in entering the market. A firm’s need to balance the performance offered by the product and the time to enter the market has been an important debate in product innovation literature (Cohen, Eliasberg, & Ho, 1996). While the time-based competition school of thought (Stalk, 1988) argues in favor of early entry for better firm performance (Green & Ryans, 1990) and suggests that firms steeped in quality consciousness can miss the bus when it comes to displacing an already entrenched incumbent, the product performance school argues that a new product’s success depends critically on the product’s performance (Zirger & Maidique, 1990). This trade-off becomes even graver in markets that are governed by network effects, where literature argues that an entrant might be able to overcome the incumbent’s advantage in a market neither by entering too early nor by entering too late (Schilling, 2002). On the one hand, a firm that invests more time and effort in developing a product that is high on quality ends up ceding more customers to the incumbent, strengthening the incumbent’s network effects, and, consequently, making it difficult for itself to compete with the incumbent. On the other hand, a firm that enters too early with an underdeveloped product is not able to meet market expectations and attract buyers. Nonetheless, the outcomes of the compromise for markets with different population mixes cannot be judged ex ante. This makes higher quality a double-edged sword, motivating this research to investigate:

RQ5: In a market with network effects and a given population mix, what is the effect of delay in entry time attributed to development of a higher quality product on incumbent displacement?

Analytical model

The model developed here mimics a simple oligopolistic market consisting of two sellers and a buying population with heterogeneous consumers who decide based on different attributes of a product. The sellers deal in similar high-tech products or services that are amenable to network effects. They individually decide the quality and price of the product or service to be offered before entering the market. For simplicity, it is assumed that once this decision is made, it cannot be changed. One of the two sellers enters the market first and is called the first mover/pioneer/incumbent and the other follows being the late mover/follower/entrant. That is to say that by the time the entrant enters the market, the incumbent has already garnered a certain proportion of the market called the incumbent’s base, which may influence subsequent buying behavior of buyers by affecting the perceived quality of the product and giving the incumbent an advantage in driving network effects in its favor. By design, the entrant can observe the quality and price being offered by the incumbent before it enters the market and can decide the price and quality to offer accordingly.

The model is based on the assumption that each individual in the buyer population buys exactly one of the two products or services being sold in the market. An individual’s preferences influence which of the two products she or he buys. Research has shown that the purchase decisions made by consumers are complex tasks due to bounded rationality and limits on memory of consumers. As a result, buyers often resort to various simple heuristics while making purchase decisions (Bettman et al., 1991). In order to keep the simulation simple, yet as close to real-world decision-making as possible, this research models the lexicographic heuristic where the consumer “determines the most important attribute, and then examines the values of all alternatives on that attribute. The alternative which provides best value on the most important attribute is selected” (Bettman et al., 1991, p.59). This is in consonance with previous research which argues that consumers often seek different attributes in the product that they purchase, like quality or price (Animesh et al., 2011). For products which exhibit network effects, it is posited that value is derived from two distinct sources—first is the value derived due to the network and second is the value derived from the inherent physical attributes, in other words, quality of the product (Bental & Spiegel, 1995). Many scholars also contend that the value derived from product quality is retained even in the absence of network effect (Brynjolfsson & Kemerer, 1996; McIntyre, 2011b). Such a belief calls for an analysis that models network effects and quality as separate variables that influence choice rather than an integrated utilitarian model (Lee, Song, & Yang, 2016) commonly used in literature so far.

Following this line of reasoning, this article considers the buyer population consisting of consumers with different preferences for different product attributes, namely, product quality, product price, and network effects⁶ (Wind, 1978). We assume that selection of a particular attribute by consumers to base their decision of buying the product is stochastic and that the probability of choosing a particular product is determined by the aggregate preferences of the target consumers. This assumption helps us model demographic differences that may exist between two or more geographical regions for the same product as larger macroeconomic factors of a region could dictate how many individuals in a given population would be inclined to buy a product at the given price. For example, the same product may have to be priced differently in developed and developing countries. In other words, we define a market by the propensities of individuals favoring different product attributes. This formulation for choice between competing goods and services departs from the commonly used additive or multiplicative rule ascribed in literature pertaining to market entry (Clark & Chatterjee, 2015; Peng, Fan, & Dey, 2011; Zhu & Iansiti, 2012). By using additive and multiplicative rules, such models assume that not only is it possible to assign absolute values to network effects, quality, and price, but also it is possible to reconcile the differences in their units and scale in a single equation. These rules are often far from reality, where people are known to resort to noncompensatory heuristics (such that poor value for the most important attribute is not compensated by excellent value for a less important attribute) while making consumer decisions, due to their ease of execution without having to resort to difficult value trade-offs (Bettman et al., 1991).⁷ In addition, modeling buyers’ decisions using noncompensatory lexicographic heuristics based on the relative values of network effects, quality, and price of the two products on offer helps us avoid the need to reconcile units and scales of measurement.

Let p, q, and n be the probability of a person selected from the target population to make the buying decision based on product price, product quality, and current installed user base (contributing to network effects), respectively, where p > 0, q > 0, n > 0 and p + q + n = 1. Individuals in the buyer population (ξ) are picked up sequentially and are expected to decide which of the two competing goods/services to buy. Each individual makes a probabilistic choice between the two competing products, based on the attribute that she or he prefers. The basis for stochastic modeling of the consumer decision process lies in Luce’s choice axiom. Luce (1959) argues that “choice behaviour is best described as a probabilistic, not an algebraic, phenomenon” (p. 2). Accordingly, Luce’s choice axiom defines the probability P(a) of choosing an object a over object b as P(a) = a′ / (a′ + b′), such that a′ and b′ signify the consumer’s measure of preference for alternatives a and b, respectively. Previous research in psychology, sociology, and economics has widely applied this axiom in empirical, as well as, theoretical research (see Luce, 1977, for a review). Now, consider the decision of an individual after k people have bought the product.

If individual j is quality favoring and the quality of the incumbent’s and entrant’s products are ω and θ units, respectively, then the probability of j choosing the incumbent’s product I is

P_{j}^{I} = {\frac{ω}{(ω + θ)}}

Similarly, the probability of quality favoring individual j choosing the entrant’s product E is

P_{j}^{E} = {\frac{θ}{(ω + θ)}} = β

If individual j is price favoring and the price of the incumbent’s and entrant’s products are χ and λ units, respectively, then the probability of j choosing the incumbent’s product I is

P_{j}^{I} = {\frac{λ}{(χ + λ)}}

The probability of price favoring individual j choosing the entrant’s product E is

P_{j}^{E} = {\frac{χ}{(χ + λ)}} = α

Lastly, if individual j is network effects favoring and makes a decision when k people have already made their decisions (over and above the initial conditions), that is, the number of individuals already subscribed to the incumbent’s and entrant’s product are X_ak and X_bk, respectively, then the probability of j choosing the incumbent’s product is

P_{j}^{I} = {\frac{X_{a k}}{(X_{a 0} + X_{b 0} + k)}}

The probability that network effects favoring individual j chooses the entrant’s product E is

P_{j}^{E} = {\frac{X_{b k}}{(X_{a 0} + X_{b 0} + k)}}

Here, $X_{a k} \geq X_{a 0}$ and $X_{b k} \geq X_{b 0}$ , $X_{a 0} and X_{b 0}$ being the incumbent’s and entrant’s base, respectively, since the individuals who are committed to either the incumbent’s or the entrant’s product/service are assumed to purchase the same immediately after its introduction into the market, before any of the noncommitted buyers make a purchase decision. Due to the assumption about existence of incumbent’s base ( $X_{a 0}$ ), entrant’s base ( $X_{b 0}$ ), and single-homing, only the remaining population that is neither subscribed nor committed to either of the goods/services makes a buying decision. Further, we assume that when the kth consumer makes the choice, she or he is aware of the decisions of the k-1 previous consumers, that is, the installed base of the two products at the time of decision is public knowledge.

Let X_bk be the number of people who have bought the entrant’s product E after k people have made their decisions, such that l_k be the proportion of consumers that have chosen the entrant:

l_{k} = \frac{X_{b k}}{(X_{a 0} + X_{b 0} + k)}

Since both X_bk and l_k are random variables, let π_k be the expected proportion of entrant’s product base after k buyers have made their purchase decisions. Thus,

π_{k} = E [l_{k}]

We define a random variable Z_k that captures the choice of the kth consumer with respect to the entrant’s product. Z_k is defined as:

Z_{k} = {\begin{cases} 1 if entrant′s product is chosen by the buyer \\ 0 otherwise \end{cases}}

Based upon the choice exhibited by the kth consumer, the entrant’s base after k people have exercised their choice can be written as

X_{b k} = X_{b (k - 1)} + Z_{k}

where $X_{b (k - 1)}$ is the entrant’s base before the kth consumer makes her choice.

The conditional expectation of X_bk after (k − 1) people have made their purchase decision will be

E [X_{b k} | F_{k - 1}] = X_{b (k - 1)} + E [Z_{k} | F_{k - 1}]

where $F_{k - 1}$ denotes user bases of incumbent and entrant when the kth person makes the purchase decision. Now,

E [Z_{k} | F_{k - 1}] = 1 \times P [Z_{k} = 1 | F_{k - 1}] + 0 \times P [Z_{k} = 0 | F_{k - 1}] = P [Z_{k} = 1 | F_{k - 1}]

The probability that $[Z_{k} = 1 | F_{k - 1}]$ is the probability that the kth consumer chooses the entrant’s product. From equations (1) to (3), we can write $P [Z_{k} = 1 | F_{k - 1}]$ as

\begin{array}{l} = p \times P [Z_{k} = 1 | price favoring] + q \times P [Z_{k} = 1 | quality favoring] + \\ n \times P [Z_{k} = 1 | network favoring] \\ = p \times α + q \times β + n \times \frac{X_{b (k - 1)}}{(X_{a 0} + X_{b 0} + k - 1)} \end{array}

Thus, equation (6) can be expressed as

E [X_{b k} | F_{k - 1}] = X_{b (k - 1)} + p \times α + q \times β + n \times \frac{X_{b (k - 1)}}{(X_{a 0} + X_{b 0} + k - 1)}

E [X_{b k} | F_{k - 1}] = C + C_{k - 1} X_{b (k - 1)}

where

C = (p \times α + q \times β)

and

C_{k - 1} = (\frac{1 + n}{(X_{a 0} + X_{b 0} + k - 1)})

Recursively writing $E [X_{b (k - 1)} | F_{k - 2}]$ using equations (9) and (10), we get

E [X_{b k} | F_{k - 2}] = C + C_{k - 1} (C + C_{k - 2} X_{b (k - 2)})

E [X_{b k} | F_{k - 2}] = C (1 + C_{k - 1}) + C_{k - 1} C_{k - 2} X_{b (k - 2)}

Proceeding in a similar situation, we can write $E [X_{b k}]$ as a function of initial bases of the two competitors and the product attributes (product quality and product price) along with the consumer heterogeneities in their preference for different attributes.

E [X_{b k}] = E [X_{b k} | F_{0}] = C (1 + C_{k - 1} + C_{k - 1} C_{k - 2} + ... + \prod_{i = 1}^{k - 2} C_{k - i}) + (\prod_{i = 1}^{k - 1} C_{k - i}) X_{b 0}

If $X_{b 0} = 0,$ that is, entrant does not have any supporting base, then

E [X_{b k}] = C (1 + C_{k - 1} + C_{k - 1} C_{k - 2} + ... + \prod_{i = 1}^{k - 2} C_{k - i})

E [X_{b k}] = f (p, q, n, α, β, k, X_{a 0})

That is, the expected number of people who have chosen to adopt the entrant’s offering when a total of k people have made their adoption decision is a function of, among other things, the initial consumer base of the incumbent. The expected proportion of entrant’s population after k people have decided would be $\frac{E [X_{b k}]}{(X_{a 0} + X_{b 0} + k - 1)}$ .

Results

Theoretical results: Infinite population

If the number of people in the target population is very large, then it is reasonable to assume that after some time, the expected proportion of buyers who have purchased the entrant’s product will tend to converge. In other words, $π_{k - 1} \to π_{k}$ .

Under the condition of convergence, $π_{k} = π_{k - 1} = π$ .

Hence,

(X_{a 0} + X_{b 0} + k) \times π = (X_{a 0} + X_{b 0} + k - 1) \times π + p \times α + q \times β + n \times π

π = \frac{p \times α + q \times β}{p + q} = \frac{q \times (\frac{θ}{(ω + θ)}) + p (\frac{χ}{(χ + λ)})}{p + q}

From the above equation, it can be seen that in the case of infinite population, the expected value of the entrant’s market share stabilizes around a value which is independent of the value of X_a0 and X_b0, that is, if there is an extremely large market, then the proportion of the incumbent base when the entrant decides to enter does not affect the final expected market share of the entrant, and the expected market share of the entrant asymptotically reaches the value expressed by equation (14).

Assuming the quality of the entrant to be η_q times that of the incumbent, that is, $θ = η_{q} \times ω$ and the price of the entrant’s product to be η_q times the product of the incumbent, that is, $λ = η_{q} \times χ$ , we plot the expected market share of the entrant in terms of η_q and for various values of p, q, and n. The setup discussed above allows simulation of different counterfactual markets, each corresponding to a different population mix of quality-, price-, and network effects-favoring individuals who stochastically choose between the incumbent’s and entrant’s offering. Each unique combination of p, q, and n is a distinct market that can be imagined on a plane in a three-dimensional space with the probability of choosing the product on quality, price, and network effects as the three axes (Figure 1). In order to keep the results of this research comprehensible, the plane representing the markets is divided into four regions. Depending on whether the region in the plain is dominated by probability of choosing on quality, price, or network effects, three regions can be delineated as quality dominated or quality seeking (blue), price dominated or price seeking (green), and network effects dominated or network effects seeking (red). The fourth region may be called the moderate region (yellow) consisting of markets where no single attribute is given a higher preference by individuals. Each of these regions forms a triangle as shown in Figure 1. A market lying close to the centroid of each of these triangles is considered a typical market for that region. There are, therefore, four typical markets representing markets where consumers are more attracted toward quality (quality dominant, “Q” market), price (price dominant, “P” market), network effects (network effects dominant, “N” market), and no affinity to any particular attribute (moderate, “M” market), indicated as Q, P, N, and M, respectively, in Figure 1.

Figure 1.

Four typical markets.

From equation (14), we know that the managerial selection of entrant’s product quality and entrant’s product price has an important bearing on the entrant’s market share. Of interest to the entrant are the conditions under which it would be able to displace the incumbent and thereby get a market share of greater than 50%. In a moderate “M” market, the entrant’s decisions with respect to relative quality to be offered and the relative price to charge (as signified through the values of η_q and η_p) that result in a market share greater than 50% for the incumbent are shown in Figure 2. The shaded region of Figure 2 represents the desirable decisions that would displace the incumbent.

Figure 2.

η_q and η_p for incumbent displacement in M market.

Among the feasible set of decision options, an entrant might actually be more interested in knowing the limiting values of price that the entrant can afford to charge, given a certain product quality such that she or he is still able to displace the incumbent. For the four markets discussed above, the limiting values of relative price that entrant can afford to charge for different values of the relative product quality are shown in Figure 3. It is observed that as the probability of quality being the consumer decision variable increases, the ability of the entrant to charge higher price compared to the incumbent continuously increases.

Figure 3.

Limiting values of η_q and η_p for typical markets.

Simulation results

Need for simulation

While the theoretical case of considering an infinite populations helps in providing an analytical solution, most real-world markets are finite, often constrained by the target segment of the product. Appreciating this fact, we resort to simulations as a method to illustrate the case of finite markets.

For a finite population, the initial conditions of the incumbent base would have an important bearing because the condition for stationarity cannot be assumed. In such a case, the recursive formula developed in equation (12) would give the expected market share for the entrant. That said, it is expected that as the target population for the product becomes larger, the expected market share should asymptotically reach the values of π as suggested by equation (14). Apart from the expected market share, an important measure is the probability associated with this expected market share. Given that the initial conditions matter, it is important to understand the variance associated with the expectation. Hence, we conducted discrete event simulations⁸ to understand the probability of incumbent displacement for different values of p, q, and n. In the simulation, each individual made a probabilistic choice between the two competing products, based on the attribute that she or he prefers. This probabilistic decision-making was built into the simulation in order to reflect the stochastic nature of real-life decisions. The simulation ended when all individuals in the population had subscribed to either of the two products. The details of the simulation are given in Table 1.

Table 1.

Details of simulation parameters.

Parameter	Description	Range of parameter values	Remarks
ξ	Market population	ξ ∈ {1000, 2000, 3000, 4000}
q	Percentage of population favoring quality	0 ≤ q ≤ 1; step size 0.05
p	Percentage of population favoring price	0 ≤ p ≤ 1; step size 0.05
n	Percentage of population favoring network effects	1 − p − q	Every individual in the population belongs to q, p, or n
Ω	Incumbent’s product/service quality	1
Χ	Incumbent’s product/service price	1
ρ	Incumbent’s base	Base case: 0 < ρ ≤ 0.3; step size 0.1
θ	Entrant’s product/service quality	1 ≤ θ ≤ 5; step size 0.1	Modeled as a multiple of ω
λ	Entrant’s product/service price	1 ≤ λ ≤ 5; step size 0.1	Modeled as a multiple of χ
σ	Entrant’s base	0 ≤ σ ≤ 0.9; step size 0.05	Modeled as a percentage of ρ

In all, we simulated 231 different markets representing different values of p, q, and n. For each market, we performed 1000 simulation runs to estimate the probability associated with the entrant gaining a market share of more than 50%. We present the results only for the four typical markets. By doing so, a balance between parsimony and detail can be achieved. While results will be explicated for the typical markets in each region, it will not be difficult for practitioners to extrapolate the results for any market that lies on the plane. Table 2 presents the proportion of quality-, price-, and network effects-favoring individuals in each of these four markets.

Table 2.

Definition of four typical markets.

Market designation	Proportion of individuals favoring quality	Proportion of individuals favoring price	Proportion of individuals favoring network effects
Q	0.6	0.2	0.2
P	0.2	0.6	0.2
N	0.2	0.2	0.6
M	0.4	0.3	0.3

To keep the findings brief yet realistic, results are presented only for the simulations where the incumbent had already captured 10% of the market at the time of entry of the entrant, and the market population was 1000. It is important to note here that for other values of incumbent base and entrant base, the nature of the results is qualitatively the same.

Investigating RQ1, RQ2, and RQ3: Effect of entrant’s product quality and product price on expected market share and probability of incumbent displacement

Earlier in the example of technological leapfrogging mentioned in the theoretical background, we noted that in the video-gaming industry, successful competitors have offered at least three times the clock speed of the fastest system in the previous generation. While three times the product quality was able to displace the incumbent in the video game industry in the United States, the same may not work for different demographics. Therefore, to investigate the effect of entrant’s product quality and product price on the expected market share of the entrant in markets comprising different demographics, we numerically varied the value of entrant’s relative product quality η_q and relative product price η_p. Both η_q and η_p were varied from 0.5 to 5. In Figure 4, the region depicted in dark blue represents the combination of η_q and η_p for which the entrant’s expected market share is greater than that of the incumbent in the four typical markets P, Q, M, and N. The region depicted in light blue represents the combination of η_q and η_p for which the entrant’s expected market share is less than 50%, that is, it is unable to displace the incumbent. The black line shown in the four graphs in Figure 4 indicates the limiting values of η_q and η_p for an infinite population market such that the entrant’s market share is 50%.

Figure 4.

$η_{q}$ and η_p for incumbent displacement in typical markets.

We observe that there is a difference between the black line and the values of η_q and η_p that separate the area into light and dark blue, and this difference is larger in the case of markets M and N, compared to P and Q markets. The reason for this observation can be attributed to the dominance of network effects based on adoption in M and N markets as they comprise of individuals who are more likely to buy the product based on network effects. Additionally, we observe that in an N market, the only way to displace the incumbent is by undercutting the incumbent (almost offering the product for free) and offering a very high-quality product. While in a P market there are diminishing returns of investing in a higher quality product (an entrant in unable to get a proportionate increase in relative price), there are increasing returns of investing in higher quality in a Q market (as the entrant is able to charge a much higher premium in terms of price).

Next, we shift our focus to understanding the probability of entrant displacing the incumbent, that is, the probability of the entrant obtaining a market share of more than 50%. As discussed previously, we carried out discrete event simulations for the aforementioned analysis. To bring out the effect of product quality on probability of success of the entrant, the relative quality offered by the entrant was varied from 1 (i.e. equal to that of the incumbent product) to 5 times (i.e. five times superior to the incumbent product). The price charged by the entrant was kept constant and at par with the price charged by the incumbent. The chance of the entrant displacing the incumbent was analyzed in the four typical markets and is illustrated in Figure 5.⁹

Figure 5.

Probability of incumbent displacement for different values of η_q.

An analysis based on the probability of incumbent displacement allows us to estimate the risk associated with entering different markets rather than just looking at the expected market shares. For example, while we see from Figure 4, in a P market, the entrant never gets an expected market share of 50% when the entrant’s price is equal to that of the incumbent, the probability analysis reveals that the probability of incumbent displacement is not zero. Figure 5 reveals it is much more risky for the entrant to enter a “P” market compared to an “M” market because the probability of displacing the incumbent is always lower in a “P” market, despite the entrant charging the same price as that of the incumbent. As a consequence, an entrant with limited resources and having an option to enter either a “P” or an “M” market would be better off with an investment in the moderate “M” market.

Similar to Figure 4, we find that it is only in a “Q” and an “M” market that an entrant has more than 50% chance of displacing the incumbent when the entrant’s price is equal to that of the incumbent. However, even in a Q market, there exists a minimum additional quality over and above that provided by the incumbent that the entrant must provide in order to have even a small chance of displacing the incumbent (if it intends to charge a price equivalent to that of the incumbent). The minimum quality that the entrant must provide is even higher in markets “M,” “P,” and “N.” It is observed that in market “P,” the probability of displacing the incumbent never reaches 50%. This low probability of success is surprising because the price charged by the entrant is same as that charged by the incumbent. A probable reason for this observation is that the entrant is unable to counter the incumbency advantage of initial installed base enjoyed by the incumbent.¹⁰ In an “N” market, an entrant is never able to displace the incumbent for the different simulation parameters. This happens because despite offering even five times the incumbent’s product quality, the market characteristics are such that too few people favor higher quality, making it extremely difficult to offset the incumbent’s advantage of initial base. A firm is thus advised to avoid entering a market in which users are disproportionately tilted toward buying a product just based on network effects.

Next, to understand the effect of price charged by the entrant on the probability of incumbent displacement, η_q was set to 3 (drawing once again from the video game industry example), and η_p was varied from 1 to 5. Figure 6 shows the probability of incumbent displacement for different prices charged by the entrant for the four typical markets when $η_{q} = 3$ . It is expected that, as the price of the entrant’s product increases, the proportion of price favoring people in a market is less likely to buy the entrant’s product. Moreover, this decrease indirectly affects the choice of product made by network effects favoring people. As a result, the probability of incumbent displacement by the entrant is anticipated to be highest when the price charged by the entrant is equal to that charged by the incumbent. However, from an entrant’s view point, it is essential to understand how incremental price affects its chances to displace the incumbent because if the entrant is able to charge a higher price without hugely affecting probability of success, the high prices increase the firm’s profit.¹¹

Figure 6.

Probability of incumbent displacement for different values of η_p for $η_{q} = 3$ .

As expected, the probability of winning is almost zero in “P” and “N” markets. This is tied to the poor response at launch of the Play Station 3 console of Sony, which was expensively priced (even though it was much more technologically sophisticated). In an “M” market, the probability of displacing the incumbent decreases almost exponentially with price charged by the entrant, but, in a quality dominant “Q” market, this probability decreases almost linearly with the price charged by the entrant. A faster decline is seen in market “M,” since it comprises a higher proportion of people favoring price and people favoring network effects in comparison with a “Q” market. The higher fraction of people favoring price in market “M” makes the entrant’s product (which is relatively priced higher than the incumbent) a less favorable alternative. This in turn makes the entrant’s product a less favored alternative for people favoring network effects in market “M” as well. Consequently, an amplified response to increase in price is seen in market “M.”

Investigating RQ4: Effect of seeding initial customers

Market seeding is often used as a strategy to counter the initial customer base advantage held by the incumbent. The effect of seeding was, therefore, studied in all the typical markets. As mentioned earlier, it is assumed that the incumbent enjoys 10% market share when the entrant enters the market. In order to investigate the effect of seeding, we considered different proportions for seeding by the entrant. For brevity, we are reporting the results for two cases, namely, no seeding and 3% seeding by the entrant (i.e., 30% of the incumbent’s initial base $X_{a 0}$ at the time if entry); 3% seeding by the entrant means that when the entrant enters the market, it is immediately able to garner a market share of 3% by seeding, while the incumbent possesses a market share of 10% at that time. Figure 7 presents whether the incumbent or the entrant dominates the market for different values of η_q and η_p. We see that in all the four typical markets, seeding allows the entrant to dominate for more number of combinations of η_q and, compared to no seeding by the entrant. That is to say that the expectation of market share for the entrant is more than 50% for a larger number of combination of η_q and η_p. This is because garnering an initial market share due to seeding allows the entrant to take advantage of the network effects arising from the seeded population, leading to a better chance of displacing the incumbent for a given combination of η_q and η_p.

Figure 7.

Effect of seeding on η_q and η_p for incumbent displacement in typical markets.

We also observe that the effect of seeding is more prominent in markets M and N, in comparison with markets Q and P. That is, the number of combinations of η_q and η_p for which the expected market share of the entrant crosses 50% mark is much more numerous for the M and N markets, compared to the Q and P markets. This may simply be attributed to the higher proportion of network effects favoring people in the M and N markets, compared to the Q and P markets. Interestingly, the advantage of seeding is highest in an M market, compared to an N market, even though the proportion of network effects favoring people is more in an N market. A probable reason for this may be that in an N market, the entrant can only take advantage of network effects, since the proportions of people favoring quality and price are much lower compared to the proportion of people favoring network effects. But, in an M market, moderate proportions of people favoring quality, price, and network effects allow the entrant to take advantage of the interaction effects between higher quality offered and network effects for high values of η_q and interaction effects between lower price charged and network effects for lower values of η_p. This enables the entrant a better chance to displace the incumbent for a larger number of combinations of η_q and η_p. It, therefore, follows from this result that market seeding as a strategy is specifically beneficial for the entrant in the M and N markets, over Q and P markets. It may be noted, however, that despite seeding (at even 30% of the current incumbent base at the time of entry) in an N market, the entrant cannot expect to displace the incumbent for $η_{p} \geq 1$ , which means that the expected market share of the entrant can exceed the incumbent only if the entrant undercuts the incumbent in an N market.

We now discuss the probability of the entrant displacing the incumbent in an M market (Figure 8). Results are shown for the base case (no seeding), 1% seeding, and 3% seeding. For all cases, price charged by the entrant is equal to that charged by the incumbent, that is, $η_{p} = 1$ . As expected, we see that as the percentage of population seeded increases, the probability of displacing the incumbent increases. This also implies that, for the same positive chance of displacing the incumbent, as the percentage of seeding increases, the quality required to be offered lowers. Thus, it may be advantageous to incur initial costs (often leading to initial losses) by giving away free products, as they can substantially increase the probability of the entrant to be the market leader without having to offer a relatively higher quality product (needed if the entrant decides not to seed the market). As an example, consider the case of telecom entrant Jio in the highly competitive market of India. At launch, it ran a promotional offer whereby it gave free SIM cards and data (all calls were also linked to data and were thus free). This strategy proved to be a smashing success as it enabled Jio to garner millions of subscribers in a very short span of time. The impact of its entry shook up the market leading to some notable exits and mergers.

Figure 8.

Effect of seeding on probability of incumbent displacement for different values of η_q and $η_{p} = 1$ .

Another possibility afforded by seeding is the ability to charge a higher price for the same product quality without affecting the probability of winning adversely. Figure 9 shows the effect of seeding strategy on the maximum price $({η_{p}}^{*})$ that the entrant can charge for different values of product quality such that it can still displace the incumbent in market “M.”

Figure 9.

Effect of seeding on Maximum price $({η_{p}}^{*})$ at which incumbent displacement probability > 0.5.

From Figure 9, the following two observations can be made. Firstly, there is a leftward shift in the minimum quality that is required to have more than 50% chance of displacing the incumbent. That is to say that the minimum quality that must be offered by the entrant to displace the incumbent progressively decreases as the percentage of population seeded increases. Secondly, as the product quality increases, the difference between maximum price that can be charged under the three scenarios (no seeding, 1% seeding, and 3% seeding) increases. This is interesting since it implies that an entrant would economically benefit more by seeding when it enters the market with a higher quality product. Intuitively, one would imagine that since the product quality is already high, there is no need to invest in seeding. The explanation for this result is twofold. At high product quality, the proportion of quality favoring people who buy the entrant’s product increases. This increase indirectly makes the entrant’s product more appealing to network effects favoring people. As a result, the effect of seeding is amplified as the quality of the entrant’s product increases and the probability of entrant displacing the incumbent would be higher. An entrant can, therefore, trade off this amplified higher probability of winning for a higher price that it can charge for the product, consequently, leading to the observed result.

Investigating RQ5: Delaying market entry for better quality

While seeding might compensate for the time and effort required to provide higher levels of quality, seeding might not always be a viable strategy for the firm, specifically, when the technology platforms or underlying technology has itself not been fully developed. In such cases, the firm must strive to develop a higher quality product. However, in a market where innovation and value derived increases with adoption, entering too late albeit with a much advanced product may not be the best strategy. It might happen that while the entrant works on product quality, the incumbent firm enhances its market share. In such a case, is it wise to wait and enter with a higher quality? This question motivates our analysis on appropriate quality and delay in entry.

The model developed, hitherto, considered that the entrant could enter with any quality level that it chose at the start of the simulation. However, in reality, a firm may actually have to invest more time (and effort) to come up with a product of high quality. During such time, the incumbent may capture additional market share, thus, making entry with a higher quality a doubled-edged sword. The timing effects associated with better quality were incorporated in the model by considering that a firm requires more time to enter with a higher quality. Assuming a linear relationship between product quality and time taken to develop that product and a linear product adoption rate for the incumbent with time, we write the following equations: $θ = ω + a t$ , and $ρ = ρ_{0} + f t$ , where a and f are the entrant’s innovation rate and incumbent’s product adoption rate, respectively, and ω and ρ₀ are the incumbent’s quality and incumbent’s base at the time the entrant enters the market, respectively. Solving these equations by eliminating time (t), we get the incumbent’s base when the entrant is ready with a product quality θ as $ρ = ρ_{0} + \frac{f}{a} (θ - ω)$ .

The term $(θ - ω)$ is the net differential of the quality difference between entrant and incumbent and f/a is defined as the net effect of late entry on incumbent’s product adoption. While f might be an exogenous parameter for the entrant,¹² it may still be able to influence a. For the purpose of this research, we focus on the aggregate effects of both these parameters as reflected in f/a values. Using simulation, to understand the effect of late entry, the quality that the entrant decides to enter the market with, on the probability of winning the market, $η_{p} = 1$ and η_q, is varied from one to five. The probability of incumbent displacement for different values of f/a for all the typical markets was simulated. An f/a value of 0 indicates that the entrant is already ready with the product and the incumbent does not get any additional advantages of increased product adoption. A higher value of f/a indicates that either the product innovation rate of the entrant is low or the incumbent’s product adoption rate is high. Both these would translate into a higher incumbency advantage for the incumbent.

In a Q market (Figure 10), it is found that as f/a increases the probability of displacing, the incumbent does not increase monotonically with increase in the quality offered by the entrant. For higher (lower) values of incumbent’s adoption rate (entrant’s innovation rate), the probability of displacing the incumbent increases only until the quality offered by the entrant is of a moderately high level, after which it starts decreasing. In other words, the probability of displacing the incumbent follows an inverted U relationship with the quality offered by the entrant for moderately high values of f/a. However, if the f/a rate further increases, it is extremely difficult to dislodge an incumbent. Conversely, for low values of f/a, the probability of the entrant displacing the incumbent is higher for higher values of η_q. Interestingly, in M, N, and P markets, we do not observe the inverted U shape for the probability associated with a successful incumbent displacement for the simulation parameters discussed above. For all the positive f/a values considered (0.025, 0.050, and 0.075 shown in Figure 10 for Q market), the probability of incumbent displacement is 0.

Figure 10.

Effect of f/a on probability of incumbent displacement for different values of η_q.

Discussion

Early research exploring network effects has been grounded largely in neoclassical economics and has focused primarily on the role of installed base to understand competition between early and late entrants (Fudenberg & Tirole, 2000; Zhu & Iansiti, 2012). More recent research has started to move beyond the size of installed base and has explored the role of network structure (connectivity, structural holes, the roles of members) and the conduct of actors in the network (opportunistic behavior, reputation signaling, perceptions of trust) for a more nuanced understanding of success in markets governed by network effects (Afuah, 2013; Lee et al., 2016). While this move has added to our understanding of the phenomenon of network effects, extant literature has mostly looked at market entry under the influence of network effects from the perspective of product quality, product price, and an initial installed base and argued for the relative importance of one over the other. The influence of prospective consumers’ decision-making behavior has largely remained unexplored in this literature. Our contribution with respect to this stream of literature lies in showing that consumer heterogeneities exhibited toward buying a product based on product price, product quality, or size of installed base can significantly affect the expectation and probability of an entrant displacing the incumbent. In doing so, we offer a more contextual understanding of network effects and incumbent–entrant competition that is grounded in consumer preferences.

Extending current literature that has dealt with interaction effects between product quality offered, price charged, and network effects (McIntyre, 2011a; Ohashi, 2003; Tellis et al., 2009), this research shows that the outcomes of late entry are also governed by the heterogeneities in buyers decision-making style constituting the market. Firms need to modulate their late entry strategy in terms of the three factors, based on the characteristics of the market that they plan to enter. The findings of the study also speak to the debate about the salience of early entry in markets governed by network effects being beneficial and providing strong incumbency advantage (Arthur, 1989a, 1994) or not being as durable (Katz & Shapiro, 1985, 1986). Extant literature had emphasized timing of entry as a double-edged sword where premature entry with low product quality needed to be balanced with missed opportunity due to late entry (Lilien & Yoon, 1990). It has also been proposed that an entrant’s probability of displacing the incumbent has an inverted U relationship with product quality (Schilling, 2002). This research shows that the inverted U relationship does indeed exist, but only for high (low) values of incumbent’s adoption (entrant’s innovation) rate in a Q market where quality-favoring people dominate. In P, N, and M markets, however, the late entrant is not able to displace the entrant even for low valued of incumbent’s adoption (entrant’s innovation) rate, and hence the inverted U relationship does not exist. Thus, we show that in markets governed by network effects, a much higher product quality might not ensure incumbent displacement, and an entrant should adequately account for the customers’ preferences in a market, its own innovation rate, and the incumbent’s product adoption rate before entering a market.

Concluding remarks

Implication for research and practice

Considering the plethora of factors which dictate the success of market entry, it is not surprising that devising a successful entry is described as a difficult strategic problem (Carpenter & Nakamoto, 1990). This research extends extant literature on three counts. Firstly, this work brings to fore the effect of the consumers’ preferences, or population characteristics of the market, in terms of quality, price, and network effects favoring people, on success of the entrant. Secondly, taking a departure from the current literature that uses multiplicative and additive utilitarian models, this research treats network effects as distinct from quality. This is especially significant for theorizing in highly networked contexts. For example, consider the case of WhatsApp and WeChat as instant messaging apps. Both WhatsApp and WeChat are free to use products. While WeChat dominated the Chinese market, it was not successful in making inroads into the Indian market. It is argued that among others the two reasons that worked against WeChat were (a) that it did not have the same number of features in the app that was launched in India but lesser, thus, creating difference in perceived product quality and (b) a lot of users had already adopted the competitor WhatsApp. Thirdly, this work introduces the probability of displacement of the incumbent by the entrant as the dependent variable. The authors believe that given the uncertain environments under which firms operate, this dependent variable can better inform market entry decisions of firms. In other words, being free products, the above example becomes a case of the incumbent (WhatsApp) being challenged by an entrant (WeChat) in an “N” market and failing to succeed or dominate the incumbent.

This research also contributes to practice by, first of all, elaborating that seeding a market before entry might compensate for quality offered and may become the only possible way to displace an incumbent. Secondly, though counterintuitive, this work suggests that in certain scenarios, it might, in fact, serve a firm better to offer moderate levels of quality rather than high levels of quality. Another important indication from the study is that organizations need to get a better grasp of the relative preferences of the target population, so that they may devise strategies that could provide a better chance of succeeding in the marketplace. Lastly, managers might not only find the possibility of entering multiple markets with the same product exciting but also benefit from such endeavors that can substantially reduce product development costs.

Limitations and future research

Given the increased focus on the type of problems analyzed in this article, there is a lot of scope for extending this research. Including the underlying network topology and density (Lee et al., 2016; Weitzel, Beimborn, & König, 2006) could be an interesting area for future research that aims at guiding an entrant’s strategy (in terms of product quality, price, time of entry, etc.) for different network structures. Additionally, future research could also focus on counter strategies that could be employed by the incumbent when an entrant does enter and explore how, if at all, these strategies are affected by population characteristics. Impact of switching costs, which also play a crucial role in adoption decisions, could also be investigated. Lastly, in this research, we assume that the target market comprises consumers who buy one of the two products. This assumption does not allow us to understand social welfare. More sophisticated decision rules and selection decisions could help us understand social welfare associated with the competition in network effect settings.

Supplemental material

Supplemental Material, Appendix_RNR - Combating incumbency advantage of network effects: The role of entrant’s decisions and consumer preferences

Supplemental Material, Appendix_RNR for Combating incumbency advantage of network effects: The role of entrant’s decisions and consumer preferences by Agam Gupta, Arqum Mateen, Divya Sharma, Uttam K. Sarkar and Vinu Cheruvil Thomas in Competition and Regulation in Network Industries

Footnotes

Author contributions

All authors have contributed equally in this work.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental material

Supplemental material for this article is available online.

Notes

References

Afuah

(2013). Are network effects really all about size? The role of structure and conduct. Strategic Management Journal, 34, 257–273.

Ali

(1994). Pioneering versus incremental innovation: Review and research propositions. Journal of Product Innovation Management, 11, 46–61.

Animesh

Viswanathan

Agarwal

(2011). Competing ‘creatively’ in sponsored search markets: The effect of rank, differentiation strategy, and competition on performance. Information Systems Research, 22, 153–169. Retrieved June 4, 2014, from http://pubsonline.informs.org/doi/abs/10.1287/isre.1090.0254

Arthur

(1989a). Competing technologies, increasing returns, and lock-in by historical events. The Economic Journal, 99, 116–131. Retrieved August 23, 2012, from https://www-jstor-org.web.bisu.edu.cn/stable/10.2307/2234208

Arthur

(1989b). “Silicon valley” locational clusters: When Do increasing returns imply monopoly? Mathematical Social Sciences, 19, 235–251.

Arthur

(1994). Increasing returns and path dependence in the economy. Ann Arbor: University of Michigan Press.

Bagwell

Riordan

M. H.

(1991). High and declining prices signal product quality. The American Economic Review, 81, 224–239. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/10.2307/2006797

Barua

Mukherjee

(2011). The incumbency protection power of network effects: Hype or reality? In ICIS 2011 Proceedings. Paper 9. Retrieved from https://aisel.aisnet.org/icis2011/proceedings/breakthroughideas/9

Benedek

Lublóy

Vastag

(2014). The importance of social embeddedness: Churn models at mobile providers. Decision Sciences, 45, 175–201. Retrieved from http://doi.wiley.com/10.1111/deci.12057

10.

Bental

Spiegel

(1995). Network competition, product quality, and market coverage in the presence of network. The Journal of Industrial Economics, 43, 197–208.

11.

Besharat

Langan

R. J.

Nguyen

C. G.

(2015). Fashionably late: Strategies for competing against a pioneer advantage. Journal of Business Research. Retrieved from http://linkinghub.elsevier.com/retrieve/pii/S0148296315003471

12.

Bettman

J. R.

Johnson

E. J.

Payne

J. W.

(1991). Consumer decision making. Handbook of Consumer Behavior, 6, 50–84.

13.

Brynjolfsson

Kemerer

C. F.

(1996). Network externalities in microcomputer software: An econometric analysis of the spreadsheet market. Management Science, 42, 1627–1647.

14.

Calantone

R. J.

Di Benedetto

(1988). An integrative model of the new product development process an empirical validation. Journal of Product Innovation Management, 5, 201–215.

15.

Carpenter

G. S.

Nakamoto

(1989). Consumer preference formation and pioneering advantage. Journal of Marketing, 26, 285–298.

16.

Carpenter

G. S.

Nakamoto

(1990). Competitive strategies for late entry into a market with a dominant brand. Management Science, 36, 1268–1278.

17.

Chakravarty

Dogan

Tomlinson

(2006). A hedonic study of network effects in the market for word processing software. Decision Support Systems, 41, 747–763.

18.

Clark

B. H.

Chatterjee

(2015). The evolution of dominant market shares: The role of network effects. Journal of Marketing Theory and Practice, 7, 83–96.

19.

Cohen

M. A.

Eliasberg

(1996). New product development: The performance and time-to-market tradeoff. Management Science, 42, 173–186. Retrieved from http://pubsonline.informs.org/doi/abs/10.1287/mnsc.42.2.173

20.

Cooper

R. G.

Kleinschmidt

E. J.

(1987). New products: What separates winners from losers? Journal of Product Innovation Management, 4, 169–184.

21.

Dietl

H. M.

Grutter

Lutzenberger

(2009). Defensive competitive strategies in two-sided markets: The example of the mail industry. Competition and Regulation in Network Industries, 10, 3–15.

22.

Dou

Niculescu

M. F.

D. J.

(2013). Engineering optimal network effects via social media services engineering optimal network effects via social media features and seeding in markets for digital goods and services. Information Systems Research, 24, 164–185.

23.

Etzion

Pang

M. S.

(2014). Complementary online services in competitive markets: Maintaining profitability in the presence of network effects. MIS Quarterly, 38, 231–248.

24.

Finney

R. Z.

Lueg

J. E.

Campbell

N. D.

(2008). Market pioneers, late movers, and the resource-based view (RBV): A conceptual model. Journal of Business Research, 61, 925–932.

25.

Fosfuri

Lanzolla

Suarez

F. F.

(2013). Entry-timing strategies: The road ahead. Long Range Planning, 46, 300–311. Retrieved from http://linkinghub.elsevier.com/retrieve/pii/S0024630113000368

26.

Fudenberg

Tirole

(2000). Pricing a network good to deter entry. The Journal of Industrial Economics, XLVIII, 373–390.

27.

Galeotti

Goyal

(2009). Influencing the influences: A theory of strategic diffusion. RAND Journal of Economics, 40, 509–532.

28.

Gao

Knight

(2007). Pioneering advantage and product-country image: Evidence from an exploratory study in China. Journal of Marketing Management, 23, 367–385. Retrieved from https://www-tandfonline-com.web.bisu.edu.cn/doi/abs/10.1362/026725707X196422

29.

Gao

S. Y.

Lim

W. S.

Tang

(2016). The impact of the potential entry of copycats: Entry conditions, consumer welfare, and social welfare. Decision Sciences, 00, 1–31.

30.

(2002). Value pricing in presence of network effects. Journal of Product & Brand Management, 11, 174–185.

31.

Goldsmith

R. E.

Newell

S. J.

(1997). Innovativeness and price sensitivity: Managerial, theoretical and methodological issues. Journal of Product & Brand Management, 6, 163–174.

32.

Green

D. H.

Ryans

A. B.

(1990). Entry strategies and market performance causal modeling of a business simulation. Journal of Product Innovation Management, 7, 45–58. Retrieved from http://doi.wiley.com/10.1111/1540-5885.710045

33.

Gupta

Jain

D. C.

Sawhney

M. S.

(1999). Modeling the evolution of markets with indirect network externalities: An application to digital television. Marketing Science, 18, 396–416. Retrieved from http://pubsonline.informs.org/doi/abs/10.1287/mksc.18.3.396

34.

Halaburda

Oberholzer-Gee

(2014). The limits of scale. Harvard Business Review, 92, 94–99.

35.

Huang

Leng

Parlar

(2013). Demand functions in decision modeling: A comprehensive survey and research directions. Decision Sciences, 44, 557–609.

36.

Jiang

Sarkar

(2009). Speed matters: The role of free software offer in software diffusion. Journal of Management Information Systems, 26, 207–240.

37.

Katz

M. L.

Shapiro

(1985). Network externalities, competition, and compatibility. The American Economic Review, 75, 424–440.

38.

Katz

M. L.

Shapiro

(1986). Technology adoption in the presence of network externalities. Journal of Political Economy, 94, 822–841.

39.

Katz

M. L.

Shapiro

(1994). Systems competition and network effects. Journal of Economic Perspectives, 8, 93–115.

40.

Kim

Srinivasan

Wilcox

R. T.

(1999). Identifying price sensitive consumers: The relative merits of demographic vs. purchase pattern information. Journal of Retailing, 75, 173–193. Retrieved from https://www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0022435999000032

41.

Klingebiel

Joseph

(2016). Entry timing and innovation strategy in feature phones. Strategic Management Journal, 37, 1002–1020.

42.

Lawless

M. W.

Fisher

R. J.

(1990). Sources of durable competitive advantage in new products. Journal of Product Innovation Management, 7, 35–44.

43.

Lee

Song

Yang

(2016). Network structure effects on incumbency advantage. Strategic Management Journal, 37, 1632–1648.

44.

Lee

O’Connor

G. C.

(2003). New product launch strategy for network effects products. Journal of the Academy of Marketing Science, 31, 241–255.

45.

Lehmann

D. R.

Esteban-Bravo

(2006). When giving some away makes sense to jump-start the diffusion process. Marketing Letters, 17, 243–254.

46.

Levin

I. P.

Johnson

R. D.

(1984). Estimating using price-quality tradeoffs comparative judgments. Journal of Consumer Research, 11, 593–600.

47.

Lieberman

M. B.

Montgomery

D. B.

(2013). Conundra and progress: Research on entry order and performance. Long Range Planning, 46, 312–324. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1016/j.lrp.2013.06.005

48.

Lieberman

M. B.

Montgomery

D. B.

(1988). First mover advantages. Strategic Management Journal, 9, 41–58. Retrieved from http://www.springerlink.com/index/T438820842320653.pdf

49.

Liebowitz

S. J.

Margolis

S. E.

(1995). Path dependence, lock-in, and history. Journal of Law, Economics, & Organization, 11, 205–226.

50.

Lilien

G. L.

Yoon

(1990). The timing of competitive market entry: An exploratory study of new industrial products. Management Science, 36, 568–585.

51.

Luce

R. D.

(1959). Individual choice behavior: A theoretical analysis. New York, NY: Wiley.

52.

Luce

R. D.

(1977). The choice axiom after twenty years. Journal of Mathematical Psychology, 15, 215–233.

53.

Marsden

(2006). Seed to spread: How seeding trials ignite epidemics of demand. In Kirby

Marsden

(Eds.), Connected Marketing: The Viral, Buzz and Word of Mouth Revolution (pp. 3–23). U.K.: Taylor & Francis.

54.

McCarthy

E. J.

(1978). Basic marketing: A managerial approach. Homewood, Illinois: R. D. Irwin.

55.

McIntyre

D. P.

(2011a). In a network industry, does product quality matter? Journal of Product Innovation Management, 28, 99–108.

56.

McIntyre

D. P.

(2011b). Where there’s a way, is there a will? Installed base and product quality in a network industry. The Journal of High Technology Management Research, 22, 59–66.

57.

Narasimhan

Mendez

Ghosh

(1996). An examination of the effect of continuous quality improvements on optimal pricing for durable goods. Decision Sciences, 27, 389–413. Retrieved from http://doi.wiley.com/10.1111/j.1540-5915.1996.tb01820.x

58.

Ohashi

(2003). The role of network effects in the US VCR market, 1978–1986. Journal of Economics and Management Strategy, 12, 447–494.

59.

Peng

Fan

Dey

(2011). Impact of network effects and diffusion channels on home computer adoption. Decision Support Systems, 51, 384–393.

60.

Rao

A. R.

Monroe

K. B.

(1989). The effect of price, brand name, and store name on buyers’ perceptions of product quality: An integrative review. Journal of Marketing Research, 26, 351–357.

61.

Ratchford

B. T.

Shugan

S. M.

Reibstein

D. J.

Rossi

P. E.

Brown

Morgan

Tellis

G. J.

Yin

Niraj

(2009). Commentaries and rejoinder to “does quality win? Network effects versus quality in high-tech market. Journal of Marketing Research, 46, 150–151.

62.

Schilling

M. A.

(2002). Technology success and failure in winner-take-all markets: The impact of learning orientation, timing, and network externalities. The Academy of Management Journal, 45, 387–398.

63.

Schilling

M. A.

(2003). Technological leapfrogging: Lessons from the U.S. video game console industry. California Management Review, 45, 6–32. Retrieved from https://www-scopus-com.web.bisu.edu.cn/inward/record.url?%20eid=2-s2.0-0037835775&partnerID=40&md5=0e41e0e55b7e480fa49cd53ea68c9e2a

64.

Schumpeter

J. A.

(1950). Capitalism, socialism, and democracy (3rd ed.). New York, NY: Harper.

65.

Shankar

Bayus

B. L.

(2003). Network effects and competition: An empirical analysis of the home video game industry. Strategic Management Journal, 24, 375–384.

66.

Shankar

Carpenter

G. S.

Krishnamurthi

(1998). Late mover advantage: How innovative late entrants outsell pioneers. Journal of Marketing Research, 35, 54–70. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/10.2307/3151930

67.

Sharma

Gupta

Sarkar

U. K.

(2015). Challenging the incumbent: The impact of product quality and network effects. In ECIS 2015 Research-in-Progress Papers. Paper 47. Retrieved from https://aisel.aisnet.org/ecis2015_rip/47

68.

Sharma

(2011). Country of origin effects in developed and emerging markets: Exploring the contrasting roles of materialism and value consciousness. Journal of International Business Studies, 42, 285–306.

69.

Srinivasan

Lilien

G. L.

Rangaswamy

(2004). First in, first out? The effects of network externalities on pioneer. Journal of Marketing, 68, 41–58.

70.

Stalk

(1988). Time—The next source of competitive advantage. Harvard Business Review, 66, 41–51. Retrieved from http://ezp-prod1.hul.harvard.edu/login?%20url=http://search.ebscohost.com/login.aspx?%20direct=true&db=heh&AN=8800022437&site=ehost-live&scope=site

71.

Sun

Xie

Cao

H. H.

(2004). Product strategy for innovators in markets with network effects. Marketing Science, 23, 243–254.

72.

Tellis

G. J.

Yin

Niraj

(2009). Does quality win? Network effects versus quality in high-tech markets. Journal of Marketing Research, 46, 135–149. Retrieved from http://journals.ama.org/doi/abs/10.1509/jmkr.46.2.135

73.

Timsit

Castiaux

Truong

Athaide

G. A.

Klink

R. R.

(2015). The effect of market-pull vs. resource-push orientation on performance when entering new markets. Journal of Business Research, 68, 2005–2014. Retrieved from http://linkinghub.elsevier.com/retrieve/pii/S0148296315000867

74.

Völckner

Hofmann

(2007). The price-perceived quality relationship: A meta-analytic review and assessment of its determinants. Marketing Letters, 18, 181–196.

75.

Wang

Chen

Xie

(2010). Survival in markets with network effects: Product compatibility and order-of-entry effects. Journal of Marketing, 74, 1–14.

76.

Weitzel

Beimborn

König

(2006). A unified economic model of standard diffusion: The impact of standardization cost, network effects, and network topology. MIS Quarterly, 30, 489–514.

77.

Weitzel

Wendt

von Westarp

(2000). Reconsidering network effect theory. In ECIS 2000 Proceedings. Paper 91. Retrieved from https://aisel.aisnet.org/ecis2000/91

78.

Wind

(1978). Issues and advances in segmentation research. Journal of Marketing Research, 15, 317–337.

79.

Wolinsky

(1983). Prices as signals of product quality. The Review of Economic Studies, 50, 647.

80.

Zhang

(2007). Agent-based simulation of consumer purchase decision-making and the decoy effect. Journal of Business Research, 60, 912–922.

81.

Zhou

J. X.

Arnold

M. J.

Pereira

(2010). Chinese consumer decision-making styles: A comparison between the coastal and Inland regions. Journal of Business Research, 63, 45–51. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jbusres.2009.01.010

82.

Zhu

Iansiti

(2012). Entry into platform-based markets. Strategic Management Journal, 33, 88–106.

83.

Zirger

B. J.

Maidique

M. A.

(1990). A model of new product development: An empirical test. Management Science, 36, 867–883.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.76 MB