Corporate Entrepreneurship, Disruptive Business Model Innovation Adoption, and Its Performance: The Case of the Newspaper Industry

Long Range Planning 49 (2016) 342–360

Contents lists available at ScienceDirect

Long Range Planning j o u r n a l h o m e p a g e : h t t p : / / w w w. e l s e v i e r. c o m / l o c a t e / l r p

Corporate Entrepreneurship, Disruptive Business Model Innovation Adoption, and Its Performance: The Case of the Newspaper Industry ☆ Jahangir Karimi, Zhiping Walter Recently, Internet and digitization, along with major news and information companies, have disrupted traditional newspaper companies’ business models, and raised serious concerns about the future viability of the print newspaper industry. This study provides a theoretical viewpoint, supported by empirical evidence from the newspaper industry, on how prominent corporate entrepreneurship attributes impact disruptive business model innovation adoption, and how such adoption impacts business model performance. It finds that, while autonomy, risk-taking, and proactiveness do have positive associations with the extent of adoption of disruptive business model innovation, innovativeness does not. Further, disruptive business model innovation adoption has a nonlinear association with business model performance. We conclude the paper by discussing theoretical implications of the study and by providing strategies that entrepreneurs and technology managers can use to adjust their corporate entrepreneurship activities in their effort to successfully adopt disruptive business model innovation. © 2015 Elsevier Ltd. All rights reserved.

Introduction Recently, Internet and digitization along with major news and information companies have collectively disrupted traditional newspaper companies’ business models. Subsequently, many have raised serious concerns about the future viability of the print newspaper industry (Clemons et al., 2002/3). Some industry analysts are predicting a complete eventual collapse of the century-old newspaper industry (Dumpala, 2009; Kiss, 2005)1. After analyzing many industries facing disruption, Christensen (2006) has pointed to the fundamental challenge of disruptive technologies as “a business model problem, not a technology problem” (p. 48)2. Since disruptive products and services typically promise lower profit margin than the existing ones, they create a conflict between the business model that is already established for existing technology and the one that may be required to exploit the emerging disruptive technology (Christensen and Raynor, 2003). Although the roles of business model innovation (BMI) adoption in opportunity-seeking behavior (Chesbrough, 2010; Dewald and Bowen, 2010; Doz and Kosonen, 2010; George and Bock, 2011), and in dealing with organizational challenges of the incumbent firms in response to disruptive innovations, have been emphasized in the past (Lucas, 2012), little work has been done to understand how disruptive BMI adoption3 is influenced by corporate entrepreneurship (CE) (Dewald and Bowen, 2010; George and Bock, 2011) or how such adoption can impact business model performance. Such understanding is important in developing a strategy for survival and in making and executing management decisions to respond to disruption (Lucas, 2012).

☆This

research was supported in part by a grant from the Jake Jabs Center for Entrepreneurship at the Business School, University of Colorado Denver. World Association of Newspapers and News Publishers (WAN) reported this in July 2010: “Since 2008, more than 166 newspapers in US have closed down and stopped publishing a print edition. More than 39 titles did so in 2008, and the number rose to 109 in 2009. So far in 2010, more than 18 papers have closed down or stopped publishing a print version. There have been nearly 35,000 job losses or buyouts in the U.S. newspaper industry since March 2007. From March to December 2007, more than 2,256 newspaper jobs have been reportedly eliminated or offered buyouts. The numbers increased to more than 15,992 in 2008 and were at more than 14,783 in 2009. As of May 2010, there have been more than 1,797 job losses or buyouts in newspaper companies in the country,” (http://www.sfnblog.com/industry_trends/2010/07/sfn_report_more_than_166_us_newspapers_h.php). 2 This is because, compared to commercializing sustaining technologies, commercializing disruptive products and services is an increasingly costly endeavor with greater uncertainty, longer time for significant commercial success, and the possibility of disrupting the markets of existing profitable products and services (Walsh et al., 2002). 3 In this context, adopting disruptive BMI or disruptive BMI adoption means developing and implementing new business models suitable for responding to disruptive innovations. 1

http://dx.doi.org/10.1016/j.lrp.2015.09.004 0024-6301/© 2015 Elsevier Ltd. All rights reserved.

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

343

In this study, we contribute to CE literature by investigating the effects of CE on disruptive BMI adoption and the effects of disruptive BMI adoption on business model performance. Although the need for further studies to find conditions under which CE is particularly beneficial or detrimental to performance has been recognized in the past, the vast majority of prior studies have investigated the direct effects of CE on firm performance (Miller, 2011; Rauch et al., 2009)4. However, prior research has suggested that firm performance is dependent upon its business model characteristics (Zott and Amit, 2010). Further, BMI mediates the link between technology innovation and firm performance (Baden-Fuller and Haefliger, 2013). Therefore, a richer understanding of how CE affects performance requires understanding the role of BMI adoption in this relationship, especially when firms are faced with digital disruption. We intend to provide this understanding by investigating whether and how a firm’s being entrepreneurial — which is defined by prominent CE attributes — affects BMI adoption and how BMI adoption affects business model performance. Such an investigation is important because, from a theoretical perspective, the impact of prominent CE attributes on the extent of disruptive BMI adoption has never been empirically tested in the past. Further, from a managerial perspective, entrepreneurs or technology managers need to learn how to adjust their CE activities in order to succeed in adopting disruptive BMI. To develop a full understanding of how prominent CE attributes affect business model performance through BMI adoption in response to digital disruption, we integrate disruptive BMI concepts into CE theoretical framework. We focus on the newspaper industry as an information intensive industry to provide an empirical context for our research model. According to Miller (2011), there is a need for richer characterization of entrepreneurial research through context specific studies because these studies “may enhance application and generate more fine-grained and more empirically valid knowledge … may be of great interest to practitioners and scholars alike.” (p. 881). In addition, focusing on one industry gives us the empirical context in which to develop measurements for disruptive BMI adoption and for business model performance. Two research questions are addressed: what are the prominent CE attributes that impact disruptive BMI adoption? And to what extent does disruptive BMI adoption influence business model performance? This study, therefore, makes two important contributions. First, it develops 1) measurement scales for those prominent CE attributes that facilitate disruptive BMI adoption; 2) measurement scale for the extent of disruptive BMI adoption; and 3) measurement scale for business model performance. Second, it conducts an empirical study to examine how prominent CE attributes impact disruptive BMI adoption and how disruptive BMI adoption influences business model performance. Below we provide the theoretical framework for the study, followed by hypotheses development. In the Method section, we explain why disruptive innovations pose BMI adoption challenges for newspaper companies so as to provide the grounding for our empirical analysis. We then describe an empirical study designed to test the proposed research model using a sample data from the newspaper industry. We discuss new measurement scales for prominent CE attributes for disruptive BMI adoption, for extent of disruptive BMI adoption, and for business model performance. Finally, we present the results of data analyses and the study’s implications, limitations and conclusions. Theoretical framework Prior research on disruptive innovations has suggested that firms facing disruptive innovations must be prepared to respond to their unpredicted threats and opportunities by adopting disruptive BMI, since disruption creates opportunities that are almost always associated with new products and services (Christensen and Raynor, 2003; Borrell Associates, 2007; Markides and Oyon, 2010 Lucas, 2012). Although business models are frequently mentioned, they are rarely analyzed and often poorly understood (Teece, 2010). From an entrepreneurial perspective, a business model is the design of organizational structures to enact commercial opportunities explicitly initiated by market imperfections (George and Bock, 2011; Downing, 2005; Franke et al., 2008; Cohen and Winn, 2007). BMI is the replacement of the old business model with a new one for offering products or services not previously available5 (Hwang and Christensen, 2008; Mitchell and Coles, 2004). It involves gradual transition from the old business model to the new one (Cavalcante et al., 2011)6. Even though BMI adoption is a critical determinant in realizing economic value, the right business model is rarely apparent (Achtenhagen et al., 2013). BMI requires companies to explore alternatives to current ways of doing business and to understand how they can meet customers’ needs differently (Nidumolu et al., 2009). As such, it has become more important for the success of a business than product or service innovation (Johnson et al., 2008). BMI requires strong incentives to motivate entrepreneurial activities necessary

4 For example, prior research has shown that proactiveness has a linear impact on performance while risk-taking has a curvilinear impact on performance (Miller and Leiblen, 1996; Tang et al., 2008). 5 According to Zott and Amit (2010), from an activity system perspective, a business model is the template of how a firm conducts business, delivers value to stakeholders, and links factor and product markets. This perspective captures which activities should be performed, by whom, and how activities and actors are linked together. It suggests that firm performance is a function of not only financial but also social aspects of a business model. Therefore, BMI design needs to have 1) novelty to adopt innovative content, structure, or governance; 2) built-in elements to retain and lock-in business model stakeholders; 3) complementarities to bundle activities to generate more value; and 4) efficiency to reorganize activities to reduce transaction costs. 6 According to Cavalcante et al. (2011), BMI starts with new business model creation. This requires abandoning or removing some processes or closing a business area or unit associated with the old business model. This is usually followed by extending the new business model, through adding activities or expanding core processes to an existing business model. Expansion of a new business model is usually followed by revision to the new business model, through exploring alternative ways of doing business. It is then followed by gradually removing processes that are associated with existing business model, and replacing them with new processes for the new business model. Eventually firms terminate the old business by terminating the existing processes, and completely replacing the old business model with the new one.

344

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

for gaining insights into both technology and market domains in the face of disruption. It requires entrepreneurs and managers to 1) understand the “deep truth” about the fundamental needs of customers and how competitors are or are not satisfying those needs; 2) understand all technical and organizational possibilities for improvements; 3) make many informed guesses about the future behavior of customers and competitors as well as about costs; and 4) make requisite adjustments to the existing business model only after considerable trial and error learning (Sosna et al., 2010; Teece, 2010). The key disruptive BMI adoption challenges are to find out how new products and services create and distribute value across the actors in a firm’s value network and how the different actors’ incentives need to be aligned with the new value distribution (Markides, 2006; Sandström and Osborne, 2011; Sandström, 2010). According to Covin and Miles (1999), CE activities are usually aimed at sustained regeneration, organizational rejuvenation, strategic renewal, and redefinition of organizations, their markets, or industries7. They relate to “organization’s ability to regularly introduce new products or enter new markets… and to the organization’s creation and exploitation of new productmarket arenas” (Covin and Miles, 1999, 50). The commonality among all firms that could be described as entrepreneurial is, therefore, the presence of innovation (Covin and Miles, 1999). CE activities are found “in companies where the strategic leaders and the culture together generate a strong impetus to innovate, take risks, and aggressively pursue new venture opportunities” (Dess and Lumpkin, 2005, 147). This is consistent with the view that firms need to have an entrepreneurial orientation (EO)8 to engage in successful CE (Covin and Miles, 1999; Dess and Lumpkin, 2005). In a recent review of the EO construct, however, George and Marino (2011) concluded that additional characteristics that entail action are required to capture the concept of CE9. Citing Lumpkin et al. (2009), they proposed that autonomy is a key dimension that measures CE. Accordingly, in this paper we conceptualize risk-taking, proactiveness, innovativeness, and autonomy as four prominent CE attributes. We discuss theoretical arguments for the inclusion of each dimension next. Autonomy refers to freedom of action and decision-making that are often necessary for an organizational member to bring a new venture or business concept forward and carry it to completion (Lumpkin et al., 2009). It is a driving force for entrepreneurial value creation and entrepreneurial initiatives enactment. Autonomy is especially important when the new venture is creating new growth using a new business model that may disrupt the established core business (Markides and Oyon, 2010). The key dimensions of autonomy relate to resources, processes and values for a given growth group rather than geographic separation or ownership structure (Christensen and Raynor, 2003). Without autonomy, either a low priority is likely to be assigned to new ideas or an old business model may be force-fitted onto the new opportunity (Govindarajan and Trimble, 2010). Therefore, establishing autonomous growth groups is essential for CE activities for creating new processes, capabilities, or ways of working together with responsibilities to do what needs to be done to ensure the success of new innovative projects10. Risk-taking is defined as making decisions and taking actions without certain knowledge of probable outcomes, borrowing heavily, or committing significant resources to ventures in uncertain environments (Rauch et al., 2009). Without a degree of risk-taking, firms delay or refrain from introducing innovations and from undertaking exploitative CE activities. This can result in reacting conservatively to changing market conditions and in weak performance due to missed market opportunities. Timely risk-taking has been associated with strategic decision speed and it has subsequently been linked to improved business performance (Eisenhardt, 1989). However, risk-taking has also been found to negatively affect performance at the embryonic stage of growth (Hughes and Morgan, 2007). Proactiveness is defined as opportunity-seeking, forward-looking behavior for introducing new products, services, or technological capabilities ahead of the competition in anticipation of future demand, which can lead to new venture opportunities (Lumpkin and Dess, 1996). Receptiveness to market signals, awareness of customers’ needs, careful monitoring and scanning of the environment, and extensive feasibility research are often associated with a firm’s successful proactive strategy (Barclay and Benson, 1990; Wright et al., 1995). By actively anticipating and preparing for change and mobilizing resources far in advance of rivals, proactive firms are a step ahead of not so responsive competitors in accomplishing CE activities. Innovativeness is defined as the predisposition and willingness to engage in creative behaviors, in experimentation through the introduction of new products or services, or in technological leadership via R&D in new processes (Dess and Lumpkin, 2005). It is said to be present when firms pursue active implementations of new ideas, products, or processes (e.g., Hurley and Hult, 1998). Innovativeness is suggested to be one of the most critical factors in accomplishing CE activities (Covin and Miles, 1999; Lassen and Nielsen, 2009) and corporate venture performance (Kandemir and Hult, 2005). It equally refers to

7 CE scholars generally agree that the nature of CE activities encompasses 1) product, process, or administrative innovation; 2) strategic renewal by transforming existing organizations; and 3) internal or external corporate venturing (Covin and Slevin, 1990; Dess et al., 1999; Ireland et al., 2009; Lumpkin and Dess, 1996; Pinchot, 1985; Riley et al., 2009; Sharma and Chrisman, 1999). 8 EO refers to the “processes, practices, and decision-making activities” characterized by one or more of the following dimensions: “a propensity to act autonomously, a willingness to innovate and take risks, and a tendency to be aggressive toward competitors and proactive relative to marketplace opportunities” (Lumpkin and Dess, 1996, 136–137). EO measures a firm’s strategic posture “in taking business-related risks, to favor change and innovation in order to obtain a competitive advantage for their firm, and to compete aggressively with other firms,” (Covin and Slevin, 1989, 77). 9 George and Marino (2011) argued against some authors’ (e.g, Kemelgor, 2002) view that EO dimensions of risk-taking, proactiveness, and innovativeness are a direct measure of CE attributes. Instead, they favored other authors’ (e.g., Zahra, 1991; Dess and Lumpkin, 2005) view that EO represents a firm’s orientation or proclivity towards entrepreneurship, but not actual engagement in CE activities. 10 In large organizations, one common method of promoting autonomy involves spinning-off a division, a separate unit, or a growth group with its own budgets and employees, which, in the long run, may provide sizable incentives and motivate the spinoff to succeed and subsequently become its own publicly-traded enterprise (Govindarajan and Trimble, 2010; Tchong, 2010).

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

345

both a firm’s ability to create new to the world products, processes and services and its openness to new ideas and newto-the-firm product launches (Lumpkin and Dess, 1996; Pérez-Luño et al., 2011). Without innovation, there is no CE regardless of presence of other CE attributes (Covin and Miles, 1999).

Hypotheses development Prior research across many industries has shown that autonomy is a necessary condition for creating new growth that may disrupt the core business (Christensen, 1997). It is even more important for accomplishing CE activities associated with creating a new venture, especially when the new venture needs to take an approach the old venture will dislike or reject (Gilbert, 2005). Although senior management needs to set the innovation agenda and ensure that the right ideas get handled in the right way, a “growth group” (i.e., individuals, a dedicated team, a separate unit, “heavyweight” development teams, innovation council, or entrepreneurial champions) usually defines what ideas are in or out of bounds, evaluates and gets funding for new proposals, and works with managers to solve problems and shape emerging businesses (Govindarajan and Trimble, 2010; Kim and Steven, 1992). Since individuals in autonomous growth groups do not represent their functional group interest, they can more easily create new processes or ways of working together (Kim and Steven, 1992), which is required for BMI adoption. Therefore, one would expect that the more autonomous the growth groups are, the more likely they will be able to adopt disruptive BMI in responding to digital disruption. We define the extent of disruptive BMI adoption as the extent to which a firm has successfully adopted and implemented disruptive BMI and propose the following: Hypothesis 1. The greater is the extent of autonomy of the growth group in a company, the higher will be the extent of its disruptive BMI adoption. BMI often “requires managers to experiment to discover what can work and what fails, and communicate and institutionalize learning mechanisms (incorporating new knowledge and skills) into systems, procedures and structures across all echelons of the organization” (Sosna et al., 2010, 385). It requires significant risk-taking and entrepreneurial behavior in searching for alternatives outside the existing business model to identify and adopt a new or different business model (Chesbrough and Rosenbloom, 2002). Prior studies have also found that executives tend to be risk-averse and focus on traditional short-term metrics at the expense of investments needed for long-term development (Burgelman and Valikangas, 2005; McGrath et al., 2006). Risk-taking behavior has been found to influence a firm’s tendency towards exploitative innovation adoption (March, 1991). It is also positively associated with the number of internally generated innovations (Pérez-Luño et al., 2011). Positive risk experience has also been found to moderate the relationship between the incumbent manager’s increased perception of firm opportunity and intention to adopt disruptive BMI (Dewald and Bowen, 2010). Therefore, one would expect risk-taking to be positively associated with the extent of disruptive BMI adoption. The following hypothesis is proposed next: Hypothesis 2. The greater is the extent of the risk-taking behavior in a company, the higher will be the extent of its disruptive BMI adoption. Innovative firms bring external perspectives into the innovation process by having well-defined ways of interacting with their core customers, learning from non-consumers, and scanning for new ideas from other industries (Anthony et al., 2008). Innovativeness represents a basic willingness and culture to depart from existing technologies or practices and venture into the unknown (Kimberly, 1981; Kitchell, 1995; Salavou, 2004). Innovativeness requires a corporate mindset and an innovation supportive culture to foster creative, innovative, and initiative-taking behaviors among participants (Jassawalla and Sashittal, 2002). It also requires a gradual shift of the corporate mindset, from “not invented here” to “proudly found elsewhere” (Huston and Sakkab, 2006). Since organizational innovativeness behavior is related to frequency and timing associated with generation and adoption of new innovations (Salavou, 2004), one would expect a company’s innovative tendency to be related to its engagement in CE activities associated with disruptive BMI adoption. Therefore we propose the following: Hypothesis 3. The greater is the extent of the innovative behavior in a company, the higher will be the extent of its disruptive BMI adoption. Proactiveness has been found to be associated with creating competitive advantage (Dess and Lumpkin, 2005) and with greater firm performance in a variety of industry settings (Hughes and Morgan, 2007; Lumpkin and Dess, 2001). Further, the impact of proactiveness on innovation adoption is similar to that of risk-taking on innovation adoption (Pérez-Luño et al., 2011). Proactive firms have the desire to be pioneers, to adopt innovations, and to exploit the knowledge either generated internally or developed by other companies (Pérez-Luño et al., 2007). Therefore, one would expect proactiveness to be positively associated with CE activities associated with disruptive BMI adoption. Consequently, Hypothesis 4. The greater is the extent of the proactive behavior in a company, the higher will be the extent of its disruptive BMI adoption.

346

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Size1-Size5 Auto1-Auto3

Autonomy SIZE

Risk1-Risk3

Risk-taking

Innovativeness

Inno1-Inno3

Proc1-Proc3

Proactiveness

Business Model Innovation Adoption (BMIA)

BMIA1-BMIA6

Business Model Performance (BMP)

BMIP1-BMIP4

Figure 1. Research model with constructs. BMIA: Business Model Innovation Adoption; BMP: Business Model Performance

Prior studies have established a link between BMI and capturing value from innovation (Achtenhagen et al., 2013; Amit and Zott, 2001; Chesbrough, 2010; Chesbrough and Rosenbloom, 2002). However, changing the business model is likely to be difficult for a firm and it may present a greater challenge than the technology itself (Achtenhagen et al., 2013; Aspara et al., 2013). Still, some companies such as Apple, IBM and Mozilla have successfully changed their business models to compete and have benefited from the disruption by focusing on new lines of businesses and deemphasizing their old business models (Lucas, 2012). A disruptive BMI with unique pricing and distribution strategies can generate attractive profits at discount prices for low margin businesses; it can establish a new cost structure, processes and new relationships with suppliers and with partners so as to respond profitably to different classes of customers (Chesbrough and Rosenbloom, 2002; Christensen and Raynor, 2003). Therefore, one would expect that the extent of disruptive BMI adoption will positively influence business model performance. Hence, we propose, Hypothesis 5. The greater is the extent of disruptive BMI adoption in a company, the higher will be its business model performance. Research method Guided by the literature, we developed our research model as depicted in Figure 1. The theoretical model in Figure 1 includes patterns of interaction between CE attributes, BMI adoption, and business model performance in response to digital disruption. This model was tested using a sample data collected from the newspaper industry through a web survey. The research process proceeded in two stages: measurement development and the main study for model testing. We focused our attention on the newspaper industry because digital disruption has posed formidable challenges to it as discussed below. Digital disruption and disruptive BMI adoption challenges As digital publishing took off in Mid-1990s, many newspaper companies invested huge amounts in designing and marketing their websites to protect their print franchises. They were mainly concerned about the possibility that both readers and advertisers would move online, leaving them with lower revenues to cover their fixed production and distribution costs (Gilbert, 2005). Simply replicating their print business model11 online, they tried to sell blocks of online advertising space to their usual customers (e.g., department stores, car dealers), which did nothing to reverse the continuing declines in print circulation, readership, advertising revenue, profits, or share prices (American Press Institute, 2008). To maintain or even regain market share, newspaper companies were advised to behave like disruptors themselves by 1) understanding

11 As stated by Gillin (2006), “the business model of a metropolitan daily newspaper was developed over 150 years ago to support a delivery method that is becoming increasingly irrelevant” (p. 2). It requires large editorial staffs to create proprietary content and original news, a large and expensive circulation operation, enormous capital costs, and massive subscriber lists in order to support its ad rates (Gillin, 2006). It is established on the idea that timely, local information is hard to come by. It is, however, quickly collapsing when confronted with Internet and digital products that are open, inexpensive, and that allow advertisers to get to the same audience at a fraction of the cost. However, new online media entrants (e.g., Yahoo!, Monster.com, Google’s news and search offerings) built business models that captured the Internet interactive opportunities and sold to traditional newspaper companies’ advertisers.

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

347

customers’ lives and businesses’ needs in completely new ways; 2) creating products that get customers’ information JOBS12 done better than any other solutions; and 3) moving fast, reducing risk and maximizing the likelihood of success (American Press Institute, 2008). Heeding this advice, American Press Institute (API) recommends that newspaper companies go beyond the core products and core functions of reporting news and create a wide range of digital noncore products as a response to digital disruption (American Press Institute, 2008). The API views newspaper industry’s core products as traditional print products that rely primarily on display, classified, or insert advertising sold by sales representatives to traditional advertisers and that use traditional production and distribution methods (e.g., paid daily or weekly newspapers, and all other products based on a similar business model) (American Press Institute, 2006). Specifically, these include 1) print dailies and weeklies; 2) niche publications, either content-focused or advertising-focused; 3) news websites relying mainly on cost per thousand CPM (display) advertising and classified (American Press Institute, 2006). The digital noncore products include, however, 1) niche websites targeting users’ JOBS (such as, community bulletin boards like craigslist, social networking portals, and user-ratings engines); 2) digital products that rely on new revenue streams (such as paid search, lead generation and consumer direct marketing); and 3) digital products sold through nontraditional sales channels (such as self-serve advertising via Google’s AdWords) (American Press Institute, 2006). Some industry observers now believe that newspaper companies need to be “digital first and print last” by building compelling original content on platforms of their customers’ choice (Kirchner, 2011). They also believe that such a digital first strategy can result in 1) more audience (crowd) engagement by increasing the quality of shared content; 2) more robust growth of digital platforms and audiences; 3) more efficient cost structure by harnessing “both the cloud and the crowd”; 4) more revenue opportunities; and 5) more profits by allocating resources to cost-effective creation of content and sales and not just to the legacy mode of production13. This strategy, however, depends on having the freedom to create a unique cost structure and value network14 in order to be profitable in developing and selling its earliest products15, because the cost structure and value network associated with their core business cannot justify such a strategy. Since digital noncore products and services originate and prosper in new markets and value networks, they change the way value is distributed across a newspaper company’s value network (Wirtz et al., 2010). Surviving and even prospering in the new, complex value network (“value web”) of business-to-consumer (B2C) and business-to-business (B2B) require newspaper companies to adopt disruptive BMI (Lucas, 2012; Weill and Vitale, 2001; Wirtz et al., 2010; Ziv, 2009)16. This, however, is very difficult to do especially because such adoption needs to take place while the established print business still has substantial, profitable, and sustaining potential (Chesbrough, 2010; Dewald and Bowen, 2010; Markides and Oyon, 2010; Shafer et al., 2005). Consequently, they face daunting tasks when they attempt to identify all affected actors, their activities, and incentives and target them with new products and services as required by disruptive BMI adoption17. Given the grave challenges facing the newspaper industry, it offers a prime empirical context to examine factors affecting disruptive BMI adoption and its subsequent performance impacts.

12 JOBS is defined as an acronym for Jobs-to-be-done (what fundamental problem is a firm’s customers trying to solve?), Objectives (what objectives do the customers use to evaluate the solutions?), Barriers (what barriers limit the customers’ ability to use the solution?), and Solutions (what solutions do the customers consider?) (Innosight, 2009). 13 See more details about the Journal Register Company’s CEO John Paton’s Dialogue with Digital First Media Employees and the Public on “How a Newspaper Company Becomes a Digital First Company,” which was presented at WAN IFRA International Newsroom Summit “How the crowd saved our company” (http://jxpaton.wordpress.com/2011/06/08/wan_ifra/). According to Paton, “if print dollars are becoming digital dimes then we better start chasing the dimes. And we better do it cost effectively.” To do so, it will be essential to “harness both the cloud and the crowd” to drive down the legacy model production costs (http://jxpaton.wordpress.com/2010/12/02/presentation-by-john-paton-at-inma-transformation-of-news-summit-in-cambridge-mass/). 14 “Value network” defines the context within which a firm “identifies and responds to customers’ needs, procures inputs, and reacts to competitors” (Rosenbloom and Christensen, 1995, 651), and “establishes a cost structure and operating processes and works with suppliers and channel partners in order to respond profitably to common needs of a class of customers” (Christensen and Raynor, 2003). 15 The digital first strategy has resulted in remarkable payoffs in a short period of time for the Journal Register Company with more than 350 multiplatform products reaching an audience of 21 million Americans each month. While the company was bankrupt in 2009, in 2010 its profit was $41M. Further, 1) digital audience was up 100% from 5.5M to 11M unique visitors; 2) total audience — all platforms — was up 50% from 13M to 19.5M; 3) digital revenue Q1 was up 70% vs. industry’s 10%; 4) was producing 1000 videos per week vs. 400 per month; 5) it was streaming 1.7M videos per month vs. 117K; and 6) it outsourced noncore competencies (such as printing and mailroom services, delivery, pre-press, back-end digital, IT site maintenance, and finance) (http://jxpaton.wordpress.com/2011/06/08/wan_ifra/). According to the weekly digest report published on September 5, 2012 (http://jxpaton.wordpress.com/2012/09/05/another-tough-step/), “From 2009 through 2011, digital revenue grew 235% and digital audience more than doubled at Journal Register Company. So far this year, digital revenue is up 32.5%. Expenses by year’s end will be down more than 9.7% compared to 2009”. 16 In addition to offering purchased and in-house content, they add value by offering interpretation of content, as well as by providing new innovative digital platforms through which content may be consumed more widely (e.g., “Digital Kiosk” project is to set up a new platform for digital distribution in France or New York times’ NewsStand product, which provides access to hundreds of newspapers and magazines worldwide via a generic interface). 17 For example, recently, there have been concerted efforts by The New York Times to innovate by co-creating products and services with its customers and mobile technology partners, designing its products and services on mobile platforms for differential markets, and creating a mobile “experience” by being aware of the changing behaviors and needs of its customers (Ziv, 2009). Users, for example, can send SMS messages to The New York Times to get breaking news alerts, weather updates, latest sports scores and headlines from any section of the newspaper. Further, The New York Times has also created applications for its mobile platforms both in house and by accessing innovative capabilities from partners outside its organizational boundaries in the mobile technology arena (e.g., external device partners for iPhone, Blackberry, and other high end mobile devices) (Ziv, 2009). Even so, a few large newspaper companies (e.g., New York Times, Washington Post) have started to change their business models, made a move on the Internet, and learned to sell new advertising products to new advertisers to take advantage of the ad revenues of the new categories (American Press Institute, 2008; Gilbert, 2003; Ziv, 2009). These have allowed them to discover major opportunities for growth and to race ahead in developing and testing new products and business models while reducing the risks and costs of innovation (American Press Institute, 2008).

348

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Measurement development and scales Measurement scales geared towards the newspaper industry were developed through a 4-step process. In step 1, initial items were generated by 1) adapting scales found in the literature to the extent possible for CE attributes, BMI adoption, and business model performance; and 2) developing new items through coding industry reports, N2 project case studies, and white papers published by various consulting firms targeting newspaper companies. Appendix 1 gives more details on sources of initial items. In step 2, the initial items were validated through interviewing three local newspaper senior executives and coding the interview transcripts. In step 3, items were further refined through item sorting and categorization by four “judges”(i.e., domain experts recruited for the task). In step 4, a pilot study was conducted through a web survey. Newspaper executives were solicited to participate through newspaper forums and newsletters to newspaper executives. A total of eighty responses were received. Many controls were put in place to ensure data quality and reliability as detailed in Appendix 1. As part of the control for data validity, we collected data for job titles and newspapers worked for. Table 1 shows the job titles of respondents, which indicates all respondents were well-qualified to participate in the survey. After analyzing construct validity and reliability using the pilot data, we replaced one item and reworded two others before the final data collection. The complete final list of items for the six constructs and additional items for the dependent variable appear in Appendix 2, with * denoting items dropped in the final data analysis. Detailed discussions on these items follow. A minimum of three indicators are used to measure each reflective construct (Bollen, 1989). In the CE scales, autonomy (AUTO) is measured with three reflective items that tap into the extent to which the growth group has discretion and control over the selection of “new-to-the-firm” noncore products, necessary resources for them, and their development processes. As mentioned earlier, noncore products have emerged as newspaper companies’ response to disruptive innovations (i.e., Internet and digitization), as such the questionnaire items are geared towards noncore products whenever it is called for. Innovativeness (INNO) is measured by the extent to which a newspaper company has the tendency to generate and adopt innovations (Damanpour and Wischnevsky, 2006). Three reflective items are used to tap into the extent to which a newspaper company has innovative culture, encourages people to look beyond the current practice for new-to-the-firm product development, and is open to adoption of new innovations. In contrast to the traditional three items for measuring innovativeness in Covin and Slevin (1989), which reflects the company’s inclination to buy or market products or services that are new to the world, the three items used in our scale measure the company’s tendency towards the generation and implementation of new-to-the-firm new products or business practices. Risk-taking (RISK) behavior is measured by three reflective items to size the extent to which a newspaper company is willing to 1) venture into the unknown and take risks based on gut feelings when hard data are not available; 2) fund new-to-the-firm noncore products when future growth is uncertain; and 3) develop and commercialize new-to-the-firm noncore products even when they may cannibalize the core business. INNO and RISK are all measured using a 5-point Likert scale anchored from “strongly disagree” to “strongly agree”. Proactiveness (PROA) is measured by four reflective items that tap into the propensity of top management to think ahead and anticipate future organizational needs and to introduce new-to-the-firm products and services ahead of the competitions. PROA is measured using numerical scales with opposing phrases anchored on the two ends. The extent of disruptive BMI adoption (BMIA) is measured using six items that tap into the extent to which a newspaper company’s new business model for noncore products is novel and different from its existing business model for its core products. The six items measure the extent to which the BMIA has resulted in changes over the last three years in the revenue model, value proposition, pricing structure, cost structure, resources for selling noncore products, and new formal or informal arrangement for information exchange with partners. Since these items may vary independently and they do not necessarily correlate highly with each other, it is suitable to conceptualize BMIA as a formative construct. As such, BMIA items describe different facets of the construct and are not conceptually interchangeable. Further, removing any of the items would alter the conceptual meaning of the extent of BMIA at the aggregate level. Similar to PROA, BMIA is measured using numerical scales with opposing phrases anchored on the two ends. Therefore, Business model performance (BMP) is measured using four items that assess a company’s performance toward a diverse business portfolio of core and noncore products. More specifically, the four items for BMP measure the number of noncore products on a monthly basis, revenue generated from online sources, audience reached on a weekly basis, and percentage change in total number of advertisers from three years ago. These dimensions are not expected to correlate with each other. Further, they cause BMP and determine the extent

Table 1 Job Titles of Respondents for Pilot Data Publisher, CEO, or President Vice President, General Manager, or Associate President Director Editor or Editor-in-Chief Supervisor, CFO, consultant, Manager undisclosed Total

24 11 23 9 3 8 2 80

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

349

of BMP. Therefore, BMP is measured as a formative construct with four formative indicators. Collectively, they measure the success of BMIA in commercializing and deriving value from innovative noncore products. Five control variables are used: weekday circulation, weekend circulation, number of employees exclusively for web operations, total number of employees, and annual revenue. All responses for the control variables are ranges that respondents choose from. The control variables are used as reflective indicators of SIZE, which is used as the control variable for both BMIA and BMP. Data collection procedure for the main study For our main study, the survey website was modified to reflect changes in the final questionnaire after analyzing construct validity and reliability of the pilot data. Potential respondents were contacted through email solicitations. To collect email addresses, we used a website that had a comprehensive listing and URLs of newspapers in all fifty states and the District of Columbia. Names, email addresses, and job titles of senior executives of each newspaper were collected, if its website listed such information. The email addresses were divided into four groups according to time zones. Solicitations were sent to all email addresses between 8:30 am and 9:30 am local time on weekdays. To improve the response rate, the solicitation email was sent individually to each email address. Each potential respondent received a total of three solicitations within a one-month period. The same controls as in the pilot study were used for the main data collection. After eliminating duplicates, a total of 158 responses were collected. However, due to a server problem, ten data points were lost, resulting in a total of 148 responses. All were usable. Our sample represented approximately 9% of all US newspapers since various web sources put the number of all US daily newspapers at 1500–1700 (Countries Compared by Media > Newspapers and periodicals > Number of titles > Daily. International Statistics at Nationmaster.com, 2012; Journalism.org, 2012). Data for job titles and newspapers worked for were again collected to ensure data validity. Job titles of respondents for the final data set were found to be similar to those in the pilot data. Comparing job titles and newspapers worked for data between the pilot and the main study showed no overlapping respondents between the two data collections. Descriptive statistics on the responses for the main study are listed in Table 2. Data analysis and results To test for potential non-response bias, first, following the technique used in prior research (e.g., Mani et al., 2010), we divided our sample into two groups and compared the demographics of the groups. No significant difference was found. Table 3 lists distributions in first year of online publishing. Other demographic distributions are similarly comparable. This gives reasonable assurance that non-response bias is not a concern. Second, we compared demographics of our sample as seen in Table 2 18 with demographics of a large scale NAA survey shown in Table 3 (NAA, 2007). We found that they were similar. The similarity between our data in Table 2 and NAA data in Table 3 signals that non-response bias is not a concern. This is because the NAA newspaper survey was large in scale and over 33% of all U.S. newspapers participated. Further, the demographic distributions shown in Table 3 are consistent across multiple NAA newspaper surveys that we have examined and are also consistent with other published figures that gave a complete breakdown of all US newspapers by circulation size (e.g., Press Reference, 2001), lending to the conclusion that demographic distributions in NAA surveys were representative of all US newspapers. Therefore, our sample was representative of all US newspapers. To test for common method bias, first, we performed Harman’s single factor test twice, once including all independent variables and the other time including all independent variables and the dependent variable. Exploratory factor analysis was performed with no rotation and with number of factors fixed at 1. Percent of variance explained by the single-factor model was 31.94% when the dependent variable was not included and 27.02% when the dependent variable was included. Since those percentages are not high (<50%), one cannot conclude that common method variance is a concern. Second, we added a common method factor to our model to evaluate the size of common method variance, as recommended in the literature (Podsakoff et al., 2003; Williams et al., 2010). The procedure we followed was developed by Liang et al. (2007) and used by other authors (e.g., Furneaux and Wade, 2011). Testing results show that indicator variance attributable to the common method factor ranges from <0.01% to 3%, whereas indicator variance attributable to the underlying construct ranges from 22% to 92%, with a median of 70%. The negligibly small variance attributable to the common method factor compared with that attributable to the underlying construct indicates that variance explained due to common method is very small and not a concern at all. Measurement model analysis results For the main research data set, a structural equation model (SEM) was specified. The measurement and path models were tested using a nonlinear Partial Least Squares (PLS) path modeling software, WarpPLS. We chose PLS-based SEM over covariance-based SEM because multi-level formative constructs are used in our model. WarpPLS identifies and estimates

18 Over half of the respondents represented newspapers that were the smallest — below 25,000 in circulation. This is, in fact, representative of newspaper size distribution and is typical of survey participants in any large-scale newspaper survey.

350

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Table 2 Descriptive statistics for final data set Number of Employees

Number of Responses

<100 100–400 401–700 701–1000 1001–2000 >2000 Grand Total

84 45 9 7 2 1 148

First Year Publishing Online

Number of Responses

Before 1990 1990–1996 1996–2000 After 2000 Grand Total

12 38 61 37 148

Number of Web Employees

Number of Responses

<10 11–50 51–100 Grand Total

133 14 1 148

Circulation

Number of Responses (%) Weekend Circulation

Percentage 56.76% 30.41% 6.08% 4.79% 1.35% 0.68% 100.00% Percent 8.11% 25.68% 41.22% 25.00% 100.00% Percent 89.86% 9.46% 0.68% 100.00%

Weekday Circulation

Under 25,000 25,000-under 50,000 50,000 to under 100,000 100,000 to under 250,000 250,000 to under 500,000 500,000 or over Grand Total

86 (58.78%) 30 (20.27%) 13 (8.78%) 9 (6.76%) 6 (4.05%) 2 (1.35%) 148 (100%)

88 (59.46%) 29 (19.59%) 14 (9.46%) 10 (6.76%) 6 (4.05%) 1 (0.68%) 148 (100%)

Revenue

Number of Responses

Percent

<$5M $5M-$10M $10M-$20M $20M-$70M $70M-$150M $150M-$300M >$300M Total

58 29 31 15 4 9 2 148

39.19% 19.59% 20.95% 10.14% 2.70% 6.08% 1.35% 100%

both nonlinear and linear relationships among the latent variables. It first attempts to fit an S-shaped curve to the data; failing that, it then attempts to fit a U-shaped curve to the data; failing that, it lastly fits the data to a linear relationship. Since relationships in nature and in human behavior are rarely in straight lines, WarpPLS is able to give a more accurate depiction of relationships between latent variables if it finds an S- or a U-shaped curve. If not, WarpPLS gives the same linear estimates as in linear PLS Path modeling tools such as SmartPLS and PLS-Graph. Original data were fed into WarpPLS, which

Table 3 Descriptive Statistics of Survey Respondents from a NAA National Survey. Source: NAA, 2007 Weekend Circulation

Number of Responses (%)

Under 25,000 25,000-under 50,000 50,000 to under 100,000 100,000 to under 250,000 250,000 to under 500,000 500,000 or over Grand Total

375 (56.73%) 141 (21.33%) 72 (10.89%) 55 (8.32%) 11 (1.66%) 7 (1.06%) 661 (100%)

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

351

Table 4 Comparison of First Year of Publishing to Verify No Non-Response Bias First Half of Responses

Second Half of Responses

YEAR

Number of Respondents

Percentage of Respondents

Number of Respondents

Percentage of Respondents

Before 1990 1990–1996 1996–2000 After 2000 Grand Total

6 20 31 17 74

8.11% 27.03% 41.89% 22.97% 100%

6 18 30 20 74

8.11% 24.32% 40.54% 27.03% 100%

automatically standardized all data to a mean of 0 and a variance of 1. Final research data were analyzed first by evaluating the measurement model and then by evaluating the path model. Psychometric properties of reflective latent variables for the final data set were assessed by examining internal consistency, convergent validity, and discriminant validity. Internal consistency was evaluated by examining Cronbach’s Alpha and composite reliability scores produced by WarpPLS. Table 5 shows that these values indicate internal consistency for each of the four reflective constructs. Table 5 also lists mean and standard deviation of each item. Convergent and discriminant validities were verified through confirmatory factor analysis by examining factor structure, average variance extracted (AVE), and inter-construct correlations. Table 5 shows that all factor loadings are statistically significant (p < 0.001) and above the cut-off value of 0.70. All AVE values are above the cut-off value of 0.5. Table 6 shows that all loadings are greater than all cross-loadings. Table 7 shows that all square roots of AVEs are greater than all latent variable correlations. The three tables together show strong convergent and discriminant validities for latent reflective constructs. Construct reliability of formative constructs was assessed by examining all path coefficients from predictor constructs to the dependent construct, all weights, and indicator multicollinearity. Table 8 shows that three formative indicators for

Table 5 Psychometric Properties for CE Constructs Construct

Item

Mean

STD

Loading

P Value

CR

Alpha

AVE

AUTO1 AUTO2 AUTO3

3.01 2.82 3.03

1.12 1.07 1.10

0.94 0.94 0.95

<0.001 <0.001 <0.001

0.96

0.94

0.89

0.83

0.70

0.62

PROA1 PROA2 PROA3

3.00 3.00 3.13

1.02 0.98 0.92

0.77 0.84 0.75

<0.001 <0.001 <0.001 0.89

0.82

0.74

INNO1 INNO2 INNO3

3.56 4.03 3.72

0.97 0.89 0.91

0.90 0.74 0.92

<0.001 <0.001 <0.001 0.86

0.76

0.67

RISK1 RISK2 RISK3

3.23 3.18 2.76

1.08 0.99 1.07

0.85 0.87 0.74

<0.001 <0.001 <0.001

Autonomy (AUTO)

Proactiveness (PROA)

Innovativeness (INNO)

Risk taking (RISK)

Table 6 Loadings and Cross-Loadings for CE Constructs

AUTO1 AUTO2 AUTO3 INNO1 INNO2 INNO3 PROA1 PROA2 PROA3 RISK1 RISK2 RISK3

AUTO

INNO

PROA

RISK

0.94 0.94 0.95 0.09 −0.23 0.10 0.16 −0.15 0.0001 −0.18 0.06 0.13

−0.004 0.04 −0.04 0.90 0.74 0.92 −0.23 0.17 0.05 0.05 0.02 −0.09

0.01 0.01 −0.02 0.06 −0.17 0.07 0.77 0.84 0.75 −0.003 −0.17 0.20

0.02 −0.07 0.05 −0.20 0.20 0.03 0.11 −0.11 0.001 0.85 0.87 0.74

AUTO = Autonomy; INNO = Innovativeness; PROA = Proactiveness; RISK = Risk taking.

352

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Table 7 Intercorrelations and

AUTO INNO PROA RISK

AVE of Latent Variables

AUTO

INNO

PROA

RISK

0.94 0.61 0.29 0.53

0.86 0.31 0.62

0.79 0.43

0.82

AUTO = Autonomy; INNO = Innovativeness; PROA = Proactiveness; RISK = Risk taking; Diagonal values are square roots of AVEs. Off diagonal values are correlations. All correlations are significant at p < 0.01 (2tailed) level.

BMIA are not statistically significant at p < 0.05 level. To ensure content validity, all were retained for path model testing later. The other three BMIA weights are sizable and significant. The dependent variable, BMP, was conceptualized as a latent variable with four formative indicators. All four path coefficients for BMP are sizable, significant, and with the right signs. Discriminant validity of formative constructs was assessed by examining inter-correlations, which was found to be all lower than 0.2 for all inter-correlations between BMIA items and between BMP items, satisfying the condition that those correlations need to be less than 0.70, thereby establishing discriminant validity for both BMIA and BMP. Structural model testing results As shown in Figure 2, SIZE was used as a control variable for both BMIA and BMP in the path model. We used SIZE as a latent control variable with five reflective indictors instead of using five separate control variables because all five items are indicators for SIZE, their bivariate correlations are very significant (ranged from 0.51 to 0.92, p < 0.01), and their Cronbach’s Alpha has a high value of 0.88. A Jackknifing resampling procedure was carried out to estimate p-values of path coefficients. Structural model testing results show that AUTO and RISK each has a significant, positive relationship with BMIA at

Table 8 Indicator Validity for Formative Constructs

BMIA1→ BMI BMIA2→ BMI BMIA3→ BMI BMIA4→ BMI BMIA5→ BMI BMIA6→ BMI BMP1→ BMP BMP2→ BMP BMP3→ BMP BMP4→ BMP

Indicator Weight

P Value

0.16 0.15 0.39 0.39 0.53 −0.04 0.37 0.48 0.41 0.40

0.13 0.19 <0.001 <0.001 <0.001 0.34 0.002 <0.001 0.01 0.03

BMIA = Business Model Innovation Adoption; BMP = Business Model Performance.

SIZE

Autonomy

0.17* 0.23+

Risk-taking

-0.03

ns

Innovativeness 0.25++

Proactiveness

0.37++

0.12*

BMIA R2=25%

0.28++

BMP R2=27%

*: significant at 1-tailed 0.05 level +: significant at 2-tailed 0.05 level ++: significant at 2-tailed 0.01 level Ns: not significant

Figure 2. Hypothesis testing using nonlinear PLS path modeling tool, WarpPLS. Business Model Innovation Adoption

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

353

Figure 3. Demonstration of nonlinear relationship between BMIA and BMP. *: Significant at 0.05 level (2-tailed); **: Significant at 0.01 level (2-tailed); +: Significant at 0.001 level (2-tailed)

one-tailed19, p < 0.05 level (AUTO: β = 0.17, 1-tailed p < 0.05; RISK: β = 0.12, 1-tailed p < 0.05), supporting H1 and H2. On the other hand, the relationship between INNO and BMIA is very small and insignificant (β = −0.03, 1-tailed p > 0.05); therefore, H3 is not supported. PROA has a significant positive relationship (β = 0.25, 1-tailed p < 0.005) with BMIA, supporting H4. BMIA has a strong and significant positive relationship with BMP (β = 0.28, 1-tailed p < 0.005), supporting H5. The control variable SIZE has a strong and significant positive relationship with BMIA (β = 0.22, 1-tailed p < 0.025) and with BMP (β = 0.37, 1-tailed p < 0.005). In terms of nonlinear curves, WarpPLS found nonlinear curves for all hypothesized relationships. The curve for BMIA and BMP relationship (Figure 3) indicates that when BMIA value is high, its impact on BMP is most prominent. When BMIA value is low, its impact on BMP is less prominent. When BMIA value is in the middle range, its impact on BMP is barely noticeable, meaning when BMIA is in the middle range, higher levels of BMIA produces only marginally higher BMP. WarpPLS reported three model fit indices: average path coefficient (APC) with p-value, average R-squared (ARS) with p-value, and average variance inflation factor (AVIF). APC and ARS for the nonlinear model were 0.205 (p < 0.001) and 0.260 (p < 0.001), respectively. AVIF was 1.385. All indicated good fit. We also calculated a global fit measure, GoF, computed as

communality ∗ R 2 , where communality is average of communality and R 2 is average R 2 for endogenous variables (Tenenhaus et al., 2005; Wetzels et al., 2009). GoF was computed to be 0.38, indicating large effect (Wetzels et al., 2009). To find out whether the nonlinear PLS path modeling obtained different path coefficient estimates from those obtained by a linear PLS path modeling tool, we tested the measurement and path models using SmartPLS 2.0 Beta (Ringle et al., 2005), which also automatically standardized all data to a mean of 0 and a variance of 1. Bootstrapping re-sampling procedure with 500 samples was used to estimate the significance levels, since SmartPLS did not provide Jackknifing procedure. Figure 4 shows the results of the path model estimation. The main difference between the linear and nonlinear models lies in the significance levels of paths from RISK to BMIA and from AUTO to BMIA. In the nonlinear model, both are significant at 1-tailed p < 0.05 level, with PROA to BMIA having the largest coefficient (β = 0.26, 1-tailed p < 0.005). In the linear model, the path from RISK to BMIA has a non-significant path coefficient whereas the path from AUTO to BMIA has the largest coefficient that is highly significant (β = 0.26, 1-tailed p < 0.005). The R2s for BMP are similar, 27% for both. R2s for BMIA are 27.8% and 25% for the linear and the nonlinear models, respectively. Considering that the benefits of CE activities for organization rejuvenation often take many years to fruition (Dess et al., 1999; Zahra and Covin, 1995), a medium level of impact on BMP is expected. The model indices for the linear model are 0.204 for APC (p < 0.001), 0.244 for ARS (p = 0.004), and 1.432 for AVIF, all slightly worse than the corresponding numbers for the nonlinear model.

19 1-tailed statistics are used in all hypotheses testing reporting. They are sufficient because directions of hypothesized relationships are theoretically supported.

354

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Autonomy

SIZE 0.26** 0.30+

Risk-taking

0.35+

0.11 0.28** -0.11

Innovativeness

BMIA

BMP R2=27.3%

0.22*

Proactiveness

Figure 4. Hypothesis testing using linear PLS path modeling tool, SmartPLS

Discussion and conclusions This study examined the extent to which CE has an impact on disruptive BMI adoption in response to digital disruption and the impact of disruptive BMI adoption on business model performance. We developed five research hypotheses by drawing on CE theoretical framework. Using survey data from a sample of newspaper companies, we found significant evidence for direct association of autonomy (H1), risk-taking (H2), and proactiveness (H4) with the extent of disruptive BMI adoption, as expected. However, no such evidence was found for direct association of innovativeness with the extent of disruptive BMI adoption (H3). Regarding H5, our results show that at low or high levels of disruptive BMI adoption, its impact on business model performance is high whereas at medium levels of disruptive BMI adoption, its impact on business model performance is marginal. Theoretical implications Prior literature reviews on business model have frequently pointed to the fragmentation of business model definitions and the lack of well-defined theoretical constructs for business model and business model performance (George and Bock, 2011; Teece, 2010). These have prevented rigorous and integrated research on business models and have led to inconsistent empirical findings regarding the effects of disruptive BMI adoption on business model performance, especially within an entrepreneurial context (Dewald and Bowen, 2010; George and Bock, 2011). This study has re-conceptualized the concepts of business model and BMI innovation from CE perspective and investigated the roles of prominent CE attributes in disruptive BMI adoption and the impact of disruptive BMI adoption on business model performance. It has advanced CE theory by refining the conceptualization and measurement of disruptive BMI adoption, by illustrating how prominent CE attributes can predict the extent of disruptive BMI adoption, and by demonstrating the impact of disruptive BMI adoption on business model performance. By linking prominent CE attributes to the extent of disruptive BMI adoption and by linking disruptive BMI to business model performance, this study has extended the research on business model beyond product and transaction characteristics for measuring a firm’s long-term performance. Using the newspaper industry as an example, we developed 1) new measurement scales for prominent CE attributes that clearly reflect new-to-the-firm innovation adoption; 2) new measurement scale for the extent of disruptive BMI adoption; and 3) new measurement scale for business model performance. For innovation-adopting organization, the critical innovation issue is how to manage innovation adoption in order to achieve adaptive organizational change (Damanpour and Wischnevsky, 2006). However, prior literature has found no relationship between proactiveness or risk-taking and new-to-the-firm innovation adoption (Pérez-Luño et al., 2011). Our findings suggest that while autonomy, risk-taking and proactiveness have direct significant associations with the extent of disruptive BMI adoption, innovativeness has no such association. Innovativeness is more related to a firm’s inclination to adopt new-to-the-firm products and services or to create new to the world products and services than to its ability to generate and implement new-to-the-firm ideas. As such, it is not necessarily directly related to the firm’s extent of disruptive BMI adoption. These insights are obvious only in hindsight. Using empirical data, this study has extended prior literature on business model dynamics, by showing that the linkage from extent of disruptive BMI adoption to performance is high at both low and high levels of adoption and is weak at the medium level when adoption is at extension or revision stages. This explains why most survivors of technological disruption (e.g., Apple, IBM, and Mozilla) are rarely able to completely abandon their existing business models to adopt new ones; instead they change their foci to pursue new lines of business while deemphasizing their old business models (Lucas, 2012). Further, the results also show that SIZE has a significant impact on the extent of disruptive BMI adoption and on business model performance. This suggests that abundant organizational resources and capabilities in large newspaper companies are advantageous in disruptive BMI adoption. This study further extends prior research on innovation by suggesting that, while

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

355

smaller organizations have advantage over larger ones in pursuing entrepreneurial opportunities and in generating new novel innovations (Damanpour and Wischnevsky, 2006), larger organizations have advantage over smaller ones in adopting disruptive BMI. Future research needs to investigate how to bridge the development of prominent CE attributes with disruptive BMI adoption in small firms and to identify characteristics of business models that compel or hamper the development of routines for prominent CE attributes in small firms. Managerial implications In an era of digitization, free content and citizen journalists, newspaper companies are struggling to find sustainable business models. The challenge before newspaper companies is to find ways to change their inflexible and costly print business model by taking advantage of Internet, digitization, and digital newspaper publishing in such a way that people can not only get any information anytime and anyplace but also publish their own content at will (Gilbert and Bower, 2002). The results of this study suggest that size, autonomy, risk-taking, and proactiveness have significant direct impact on disruptive BMI adoption in responding to digital disruption. Although the key dimensions of autonomy relate to resources, processes, and values rather than geographic separation or ownership structure (Christensen and Raynor, 2003), the results show that having autonomous and proactive growth group is essential in helping large firms pursue disruptive BMI adoption. To succeed in responding to digital disruption, newspaper companies need to give substantial autonomy to their growth groups and their noncore operations. They need to serve new customers, rethink content models, accept different margin structures, identify and develop noncore business concepts, and grant growth groups secure budgets and decision-making autonomy to create new revenue streams (Anthony et al., 2008). Without autonomous growth groups, force-fitting the existing business model onto the new opportunity often occurs, which rarely works because disruptive technologies normally target a new market that differs in almost all aspects from the market that the existing business model is optimized for (Christensen and Raynor, 2003). Under the condition of high CE activities, the spin-off of a proactive growth group is necessary for disruptive BMI adoption20. However, such a growth group needs to maintain the same innovative and risk-taking culture of the mainstream organization so as to carefully balance the “portfolio” of low-risk opportunities in the core and new, higherrisk growth opportunities for noncore products. Instead of risking big investments on uncertain strategies, newspaper companies need to “invest a little to learn a lot” about different business models (American Press Institute, 2008)21. The results of this study also show that the more newspaper companies act in an entrepreneurial manner, the more likely they are able to succeed in adopting disruptive BMI in response to digital disruption. It suggests that firms with a higher level of CE activities are in a better position to analyze their surrounding networks of actors and to determine how adopting disruptive BMI impacts each actor. Consequently, they are able to reduce the challenges related to commercializing such innovations. Therefore, it is important for newspaper companies to design organizational systems to monitor and assess their CE attributes and to develop and reinforce each attribute as necessary. For example, do their organizational systems reward autonomy, risk-taking, innovativeness, and proactiveness? To what extend do they allow individuals or growth groups freedom to champion new ideas or to commit significant resources to new innovative projects to ensure high returns? To what extent are their performance metrics, incentives, and innovative culture for their noncore operations different from those for their core operations? Have their employees proactively created new offerings and discovered new business models, or have they mostly made reactive decisions? How separate should their new business models be from their existing business models? What are some of the unique challenges for simultaneously pursuing both their new disruptive business models and their core business models? What digital products do customers consider and what objectives do they use in their evaluations? We also encourage future researchers to examine and extend our findings to other organizational contexts to evaluate the generalizability of our results. Limitations Like any empirical research effort, this study contains a number of methodological strengths and limitations, and we recognize specific issues that warrant caution in interpreting our results. The breadth of the sample included in this study suggests that the findings are fairly generalizable to many newspaper companies. However, the findings are limited in some important ways. Although respondents possessed a high degree of relevant knowledge, all of the measures are self-reported, and

20 However, there are “solid arguments for and against a spinout organization, and more refined insights into contingencies are emerging,” (Danneels, 2004, 256). For example, prior research (Iansiti et al., 2003) has suggested that the advantage of separated approaches is agility, but the value decreases over time. The advantage of integrated approaches is efficiency; and the value generally increases over time; when resource complementarities between the new venture and the mainstream business are critical, and these complementarities require intra-company coordination, a more integrated approach may be advised. 21 API’s N2 project report (American Press Institute, 2008) highlights twenty-four case studies of companies that have moved proactively to create new offerings for consumers and discover new business models to serve businesses, in order to generate desperately needed new revenues. These products included print, Web, email and combination products; niche products; products targeting moms, parents, Latinos, young adults, upscale demographic groups, newcomers and businesses of various kinds, and one newsroom transformation project. Most of the new products mentioned in this followed the principle of “invest a little, learn a lot,” by creating fairly simple, first-generation products that are being tweaked and improved as the companies learn what needs to be improved.

356

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

therefore subject to hindsight and other biases. Another limitation is that we captured only a snapshot of newspaper companies’ adoption of disruptive BMI. Similar to prior research on CE, this study measured CE activities at a point in time. A company’s CE activities, however, can evolve dynamically, resulting in changes in behaviors and activities of internal stakeholders and in relationships with external stakeholders. Conclusions This study focused on the links among prominent CE attributes, disruptive BMI adoption, and business model performance in order to clarify the interrelatedness among these constructs. Grounded in CE theoretical framework, it has developed 1) new measurement scales for prominent CE attributes that clearly reflect new-to-the-firm innovation adoption; 2) new measurement scale for the extent of disruptive BMI adoption; and 3) new measurement scale for business model performance. Using empirical data from the newspaper industry, it has demonstrated that autonomy, risk-taking, and proactiveness have unique contributions to the extent of disruptive BMI adoption. It has also demonstrated that the extent of disruptive BMI adoption has a nonlinear impact on business model performance. Given that CE activities often lie at the core of responding to disruptive innovations and that their importance in adopting disruptive BMI for such a response has not been demonstrated in prior research, we hope our research has provided some directions for future research in this area. Acknowledgements The authors thank the LRP Editor-in-Chief and two anonymous reviewers for their constructive comments throughout the review process. Appendix 1. Measurement development and refinement Measurement development proceeded in four steps: 1) initial item generation through an academic literature review and coding of newspaper industry reports; 2) item validation through interviews with newspaper executives; 3) item refinement through an item sorting and categorization task by four “judges” (i.e., domain experts recruited for the task); and 4) scale testing through a pilot study. Initial Measurement Generation To develop measurement items, we adapted scales found in the literature to the extent possible for CE attributes, BMI adoption, and business model performance. However, as discussed below, either the existing scales were not suitable or there were no existing scales for the majority of the measures. We therefore searched the literature on innovation management, industry reports, N2 project case studies and white papers published by various consulting firms targeting newspaper companies to identify underlying domains for the new constructs. In this process, we used domain sampling to create items and measurement scales with adequate content validity (Nunnally, 1978). To develop measurements for CE attributes, we first looked at the existing measurement scale for EO attributes. However, these were geared towards “new to the world” innovation generation and were not directly suitable for “new to the firm” innovation adoption in the context of the newspaper industry. Innovation adoption is distinct from innovation generation because it is facilitated by different organizational factors and poses different organizational challenges (Damanpour and Wischnevsky, 2006). In addition, new-to-the-firm innovation adoption is far more common because firms can adopt many “new to the firm” products regardless of their level of EO (Pérez-Luño et al., 2011). Existing measurement scale of EO is a nine-item standard scale that measures innovativeness, proactiveness, and risk-taking (Covin and Wales, 2012; George and Marino, 2011). The scale was developed by Covin and Slevin (1989), who built on the work of Miller and Friesen (1982). However, most EO studies using this scale have not distinguished among types of entrepreneurial initiatives or “new entry” (Miller, 2011). Further, a recent review of prior empirical studies of EO by George and Marino (2011) suggested “it is time to revisit these items” because of low reliability associated with one item measuring innovativeness and the ambiguity associated with the definition of risk by not including “venturing into the unknown” (p. 1004). Consequently, new items have to be generated to measure innovative, proactive, and risk-taking behaviors. These items need to clearly reflect engagement in CE activities necessary for new-to-the-firm innovation adoption. Items for the autonomy dimension were adapted from Walter and Lopez (2008). Our reviewing of prior literature on BMI suggests that disruptive BMI needs to include 1) value propositions and the opportunities to satisfy customer needs at a profit; 2) cost structure, revenue model, processes, resources; or 3) the interactions between the firm and its key stakeholders in order to measure resource, transactive, and value structures dimensions of the new business model (George and Bock, 2011; Johnson et al., 2008; Hwang and Christensen, 2008; Osterwalder et al., 2005). As suggested by Damanpour and Wischnevsky (2006), while the earliness and the rate of adoption (speed) are appropriate measures of initiation of innovation and are suitable measures for innovation generation, the extent of implementation is a suitable measure for innovation-adopting organizations and for implementation of innovation, because it represents the prevalence of the implementation across organizational units. However, no existing items were found for measuring

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

357

the extent of adoption of disruptive BMI or business model performance, as discussed above. Therefore, we again coded newspaper industry reports to generate initial items that covered the entire domain of each construct. Measurement Item Validation and Refinement Following initial item selection, we conducted interviews with senior executives from two newspapers to ensure content and face validity (Churchill, 1979) for all new items that we developed as the result of the coding process. Interview transcripts with three local newspaper senior executives were then coded to verify our earlier set of items. The above process resulted in a set of twenty-five final items, plus additional items for the control variables. Next, we recruited three business school faculty and one doctoral student for a sorting and categorization task for all constructs. The resulting set of items was then given to doctoral students enrolled in a doctoral level research methods class for the second round of sorting and categorization. Scale Testing Through a Pilot Study Following measurement development, we conducted a pilot study through a web survey to test the scale. Participants were solicited through newspaper forums and newsletters to newspaper executives. The weekly newsletter was a free service to all paid members belonging to the Digital Media chapter of NAA. The paid membership directory of the Digital Media chapter at the time showed that almost all members had job responsibilities related to digital media at a newspaper. For data control purposes, the web survey recorded IP addresses, date/time survey started, date/time survey submitted, and a session ID for each survey. The session ID was used to allow a respondent to return to an incomplete (not yet submitted) survey at a later time. Time started and time submitted were used to eliminate surveys that were completed too fast (less than five minutes). Session IDs and IP addresses were used to discern possible multiple entries from the same person. We also requested job title and newspaper name for further data control. After data were collected, we sorted all responses based on IP addresses and found five groups of responses sharing the same IP addresses within each group, for a total of eight suspect entries. We eliminated two data points that were evidently duplicates (same session ID and same answers across all questions as existing data). We eliminated two more data points because each had the same session ID, job title, and newspaper name as another data point, although answers varied slightly. The eliminated ones were submitted a few days earlier than the ones retained and presumed from the same person. No submitted survey took less than five minutes. This resulted in eighty valid responses. Analyses of job titles indicated that all respondents were well-qualified to participate in the survey. After analyzing construct validity and reliability using the pilot data, we finalized the measurement scale by replacing one item and rewording two others. Appendix 2. Survey instrument Note: Autonomy, Innovativeness, and Risk-taking are on a 5-point Likert scale, labeled with the words “Strongly Disagree,” “Disagree,” “Neutral,” “Agree,” and “Strongly Agree.” Please indicate the extent to which you agree or disagree with each of the following statements about the growth group’s autonomy with respect to developing noncore products at your newspaper company: Autonomy (AUTO) Our growth groups have substantial discretion over which noncore products to pursue. Our growth groups have control over resources necessary for developing noncore products. Our growth groups have control over development processes of noncore products. Please indicate the extent to which you agree or disagree with each of the following statements about your newspaper company’s tendency towards noncore products: Innovativeness (INNO) Our culture encourages people to look beyond the boundaries of our current business practices and our normal business model. We can accept and run with ideas that were “not invented here.” Our culture encourages the development of new, innovative products. Risk-taking (RISK) We are willing to pursue ideas for new noncore products based on gut feel, even if hard data (e.g., market research) is sketchy or unavailable. We are willing to fund new noncore products even if the initial projection of earnings maybe low compared to the core business and future growth is uncertain. We are willing to develop and commercialize new noncore products even if they are likely to cannibalize our core business.

358

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Proactiveness (PROA) In general, top management of our newspaper tends to focus on … improving the existing businesses. 12345 developing new businesses independent from existing businesses. In general, top management of our newspaper favors a strong emphasis on developing products that lie … within the boundaries of current business practice 12345 outside the boundaries of current business practice. When it comes to new product development, top management of our newspaper tends to go with products … That our current lead customers would want 12345 That would cater to potential new customers *When it comes to utilizing capabilities of the digital media, top management of our newspaper tends to… Focus on delivering news in innovative way 12345 Focus on developing new, innovative noncore products

Business Model Innovation Adoption (BMIA) How much of revenue from noncore products are generated through traditional revenue sources such as circulation, display advertising, and classified advertising? Almost all 12345 Almost none How do you sell your noncore products? Existing sales force sells both core and noncore products. 12345 Noncore products are exclusively sold through digital media sales force. How many new formal or informal arrangements for information exchange with your partners have been created in the past 3 years? No new arrangements 12345 Very many new arrangements In the last 3 years, have you changed your pricing structure for print or online products? We have made no changes to our pricing structure. 12345 We have completely changed our pricing structure. Please compare the value propositions offered by your products/services now with those offered 3 years ago. They are pretty much the same 12345 They are dramatically different Please compare the cost structure of means employed to produce the noncore products with that employed to produce the core products. Cost structure for noncore product is much higher 12345 Cost structure for noncore products is much lower

Business Model Performance (BMP) Please estimate the percentage change in total number of advertisers from 3 years ago: Increase Decrease Not Much Change If Decrease or Increase, please select a percentage: about 5%, about 10%, about 15%, about 20%, about 25%, about 30%, about 35%, about 40%, above 40% Please estimate the percentage of annual revenue from all online sources. Please select: under 1%, 1%–5%, 6%–10%, 11%–15%, 16%–20%, 21%–25%, 26%–30%, 31%–35%, over 35% On average, how many noncore products do you have on a monthly basis? Please select: 0; 1–5; 6–10; 11–15; 16–20; 21–25; 26–30; 31–50; 51–99; 100 or more Please estimate total audience reached by all your products on a weekly basis. Please select: under 10%; 10%–19%; 20%–29%; 30%–39%; 40%–49%; 50%–59%; 60%–69%; 70%–79%; 80% or over Note: * The initial results of the measurement model indicated it was necessary to drop this indicator.

References Achtenhagen, L., Melin, L., Naldi, L., 2013. Dynamics of business models — strategizing, critical capabilities and activities for sustained value creation. Long Range Planning 46 (6), 427–442. American Press Institute, 2006. Newspaper Next 2.0: A blueprint for transformation. American Press Institute, 2008. Newspaper Next 2.0: Making the Leap Beyond “Newspaper Companies”. Amit, R., Zott, C., 2001. Value creation in E-business. Strategic Management Journal 22, 493–520. Anthony, S.D., Johnson, M.W., Sinfield, J.V., 2008. Institutionalizing innovation. MIT Sloan Management Review 49 (2), 45–53. Aspara, J., Lamberg, J., Laukia, A., Tikkanen, H., 2013. Corporate business model transformation and inter-organizational cognition: the case of Nokia. Long Range Planning 46 (6), 459–474. Baden-Fuller, C., Haefliger, S., 2013. Business models and technological innovation. Long Range Planning 46 (6), 419–426. Barclay, I., Benson, M., 1990. New product development: organization and current practice. Leadership and Organization Development Journal 11 (6), 13–24. Bollen, K.A., 1989. Structure Equations with Latent Variables. John Wiley & Sons, New York. Borrell Associates, 2007. Report on Disruptive Technology and Local Media (June). (accessed 2007). Burgelman, R.A., Valikangas, L., 2005. Managing internal corporate venturing cycles. Sloan Management Review 46 (4), 26–34. Cavalcante, S., Kesting, P., Ulhoi, J., 2011. Business model dynamics and innovation: (re)establishing the missing linkages. Management Decision 49 (8), 1327–1342. Chesbrough, H., 2010. Business model innovation: opportunities and barriers. Long Range Planning 43 (2–3), 354–363. Chesbrough, H., Rosenbloom, R.S., 2002. The role of the business model in capturing value from innovation: evidence from Xerox Corporation’s technology spin-off companies. Industrial and Corporate Change 11 (3), 529–555. Christensen, C.M., 1997. The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail. Harvard Business School Press, Boston, MA. Christensen, C.M., 2006. The ongoing process of building a theory of disruption. The Journal of Product Innovation Management 23 (1), 39–55. Christensen, C.M., Raynor, M.E., 2003. The Innovator’s Solution: Creating and Sustaining Successful Growth. Harvard Business School Press, Boston, MA. Churchill, G.A., 1979. A paradigm for developing better measures of marketing constructs. Journal of Marketing Research 16 (1), 64–73. Clemons, E.K., Gu, B., Reiner Lang, K., 2002/3. Newly vulnerable markets in an age of pure information products: an analysis of online music and online news. Journal of Management Information Systems 19 (3), 17–41. Cohen, B., Winn, M.I., 2007. Market imperfections, opportunity and sustainable entrepreneurship. Journal of Business Venturing 22 (1), 29–49. Covin, J.G., Miles, M.P., 1999. Corporate entrepreneurship and the pursuit of competitive advantage. Entrepreneurship Theory and Practice 23 (3), 47–63. Covin, J.G., Slevin, D.P., 1989. Strategic management of small firms in hostile and benign environments. Strategic Management Journal 10 (1), 75–78. Covin, J.G., Slevin, D.P., 1990. New venture strategic posture, structure, and performance: an industry life cycle analysis. Journal of Business Venturing 5 (2), 123–135. Covin, J.G., Wales, W.J., 2012. The measurement of entrepreneurial orientation. Entrepreneurship Theory and Practice 36 (4), 677–702. Damanpour, F., Wischnevsky, J.D., 2006. Research on innovation in organizations: distinguishing innovation-generating from innovation-adopting organization. Journal of Engineering and Technology Management 23 (4), 269–291. Danneels, E., 2004. Disruptive technology reconsidered: a critique and research agenda. The Journal of Product Innovation Management 21 (4), 246–258.

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

359

Dess, G., Lumpkin, G.T., McGee, J.E., 1999. Linking corporate entrepreneurship to strategy, structure, and process: suggested research directions. Entrepreneurship Theory and Practice 23, 85–102. Dess, G.G., Lumpkin, G.T., 2005. The role of entrepreneurial orientation in stimulating effective corporate entrepreneurship. Academy of Management Review 19 (1), 147–156. Dewald, J., Bowen, F., 2010. Storm clouds and silver linings: responding to disruptive innovations through cognitive resilience. Entrepreneurship Theory and Practice 34 (1), 197–218. Downing, S., 2005. The social construction of entrepreneurship: narrative and dramatic processes in the coproduction of organizations and identities. Entrepreneurship Theory and Practice 29 (2), 185–204. Doz, Y.L., Kosonen, M., 2010. Embedding strategic agility: a leadership agenda for accelerating business model renewal. Long Range Planning 43 (2–3), 370–382. Dumpala, P., 2009. The Year The Newspaper Died. The Business Insider. July 4. Eisenhardt, K.M., 1989. Making fast strategic decisions in high-velocity environments. Academy of Management Journal 27, 299–343. Franke, N., Gruber, M., Harhoff, D., Henkel, J., 2008. Venture capitalists’ evaluations of start-up teams: trade-offs, knock-out criteria, and the impact of VC experience. Entrepreneurship Theory and Practice 32 (3), 459–483. Furneaux, B., Wade, M., 2011. An exploration of organizational level information systems discontinuance intentions. MIS Quarterly 35 (4), 573–598. George, B.A., Marino, L., 2011. The epistemology of entrepreneurial orientation: conceptual formation, modeling, and operationalization. Entrepreneurship Theory and Practice 35 (5), 989–1924. George, G., Bock, A.J., 2011. The business model in practice and its implications for entrepreneurship research. Entrepreneurship Theory and Practice 35 (1), 83–111. Gilbert, C., Bower, J.L., 2002. Disruptive change: when trying harder is part of the problem. Harvard Business Review 80 (5), 94–101. Gilbert, C.G., 2003. The disruption opportunity. MIT Sloan Management Review 27–32. Gilbert, C.G., 2005. Unbundling the structure of inertia: resource versus routine rigidity. Academy of Management Journal 48 (5), 741–763. Gillin, P., 2006. How the coming newspaper industry collapse will reinvigorate journalism. Paul Gillin Communications. Dec 15. Govindarajan, V., Trimble, C., 2010. Stop the innovation wars. Harvard Business Review 88 (7/8), 76–83. Hughes, M., Morgan, R.E., 2007. Deconstructing the relationship between entrepreneurial orientation and business performance at the embryonic stage of firm growth. Industrial Marketing Management 36 (5), 651–661. Hurley, R.F., Hult, G.T.M., 1998. Innovation, market orientation, and organizational learning: an integration and empirical examination. Journal of Marketing 62 (3), 42–54. Huston, L., Sakkab, N., 2006. Connect and develop: inside Proctor & Gamble’s new model for innovation. Harvard Business Review 84 (3), 58–66. Hwang, J., Christensen, C.M., 2008. Disruptive innovation in health care delivery: a framework for business-model innovation. Health Affairs 27 (5), 1329–1335. Iansiti, M., McFarlan, F.W., Westerman, G., 2003. Leveraging the incumbent’s advantage. MIT Sloan Management Review 44 (4), 58–64. Innosight, LLC, 2009. Mastering transformation. Innosight White paper (www.innosight.com), January, LLC 2008: www.innosight.com/our_approach/JOBS.html. Ireland, E.D., Covin, J.G., Kuratko, D.F., 2009. Conceptualizing corporate entrepreneurship strategy. Entrepreneurship Theory and Practice 33 (1), 19–46. Jassawalla, A.H., Sashittal, H.C., 2002. Cultures that support product-innovation processes. Academy of Management Executive 16 (3), 42–54. Johnson, M.W., Christensen, C.M., Kagermann, H., 2008. Reinventing your business model. Harvard Business Review 86 (12), 51–59. Journalism.org, 2012. Number of U.S. Daily Newspapers, 5-Year Increments, March 12, 2007. (accessed 05.10.12.). Kandemir, D., Hult, G.T.M., 2005. A conceptualization of an organizational learning curve in international joint ventures. Industrial Marketing Management 34 (5), 430–439. Kemelgor, B.H., 2002. A comparative analysis of corporate entrepreneurial orientation between selected firms in the Netherlands and the USA. Entrepreneurship and Regional Development 14 (1), 67–87. Kim, B.C., Steven, C.W., 1992. Organizing and leading heavyweight development teams. California Management Review 34 (3), 9–28. Kimberly, J.R., 1981. Managerial innovation. In: Nystrom, P.C., Starbuck, W.H. (Eds.), Handbook of Organizational Design, vol. 1. Oxford University Press, New York, pp. 84–104. Kirchner, L., 2011. John Paton’s Big Bet: Will “Digital First” Bring Home the Bacon? Columbia JournalismReviewJuly/August . Kiss, J., 2005. Newspapers: an industry in crisis. . Kitchell, S., 1995. Corporate culture, environmental adaptation, and innovation adoption: a qualitative/quantitative approach. Journal of Academy of Marketing Science 23 (3), 195–205. Lassen, A.S., Nielsen, S.L., 2009. Corporate entrepreneurship: innovation at the intersection between creative destruction and controlled adaptation. Journal of Enterprising Culture 17 (2), 181–199. Liang, H., Saraf, N., Hu, Q., Xue, Y., 2007. Assimilation of enterprise systems: the effect of institutional pressures and the mediating role of top management. MIS Quarterly 31 (1), 59–87. Lucas, H.C., Jr., 2012. The Search for Survival: Lessons from Disruptive Technologies. Praeger, Santa Barbara, CA. Lumpkin, G.T., Dess, G.G., 1996. Clarifying the entrepreneurial orientation construct and linking it to performance. Academy of Management Review 2 (1), 135–172. Lumpkin, G.T., Dess, G.G., 2001. Linking two dimensions of entrepreneurial orientation to firm performance: the moderating role of environment and industry life cycle. Journal of Business Venturing 16 (5), 429–1085. Lumpkin, G.T., Cogliser, C.C., Schneider, D.R., 2009. Understanding and measuring autonomy: an entrepreneurial orientation perspective. Entrepreneurship Theory and Practice 33 (1), 47–69. Mani, D., Barua, A., Whinston, A., 2010. An empirical analysis of the impact of information capabilities design on business process outsourcing performance. MIS Quarterly 34 (1), 39–62. March, J.G., 1991. Exploration and exploitation in organizational learning. Organization Science 2 (1), 71–87. Markides, C.C., 2006. Disruptive innovation: in need of better theory. The Journal of Product Innovation Management 23 (1), 19–25. Markides, C.C., Oyon, D., 2010. What to do against disruptive business models (when and how to play two games at once). MIT Sloan Management Review 51 (4), 25–32. McGrath, R.G., Keil, T., Tukiainen, T., 2006. Extracting value from corporate venturing. Sloan Management Review 48 (1), 50–56. Miller, D., 2011. Miller (1983) revisited: a reflection on EO research and some suggestions for the future. Entrepreneurship Theory and Practice 35 (5), 873–894. Miller, D., Friesen, P.H., 1982. Innovation in conservative and entrepreneurial firms: two models of strategic momentum. Strategic Management Journal 3 (1), 1–25. Miller, K., Leiblen, M., 1996. Corporate risk–return relations. Academy of Management Journal 39 (1), 64–90. Mitchell, D.W., Coles, C.B., 2004. Business model innovation breakthrough moves. Journal of Business Strategy 25 (1), 16–26. Nationmaster.com, 2012. (accessed 10.07.12.). Newspaper Association of America (NAA), 2007. Nielsen//NetRatings “Data Highlights Attractiveness of Newspaper Web Site Audience.” First published May 7.

360

J. Karimi, Z. Walter / Long Range Planning 49 (2016) 342–360

Nidumolu, R., Prahalad, C.K., Rangaswami, M.R., 2009. Why sustainability is now the key driver of innovation. Harvard Business Review 87 (9), 57–64. Nunnally, J., 1978. Psychometric Theory. McGraw Hill, New York. Osterwalder, A., Pigneur, Y., Tucci, C.L., 2005. Clarifying business models: origins, present, and future of the concept. Journal of Association for Information Systems 16 (1), 1–25. Pérez-Luño, A., Valle Cabrera, R., Wiklund, J., 2007. Innovation and imitation as sources of sustainable competitive advantage. Management Research 5 (2), 67–79. Pérez-Luño, A., Wiklund, J., Valle Cabrera, R., 2011. The dual nature of innovative activity: how entrepreneurial orientation influences innovation generation and adoption. Journal of Business Venturing 26 (5), 555–571. Pinchot, G., III, 1985. Intrapreneuring. Harper & Row, New York. Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology 88 (5), 879–903. Press Reference, 2001. (accessed 10.06.10.). Rauch, A., Wiklund, J., Lumpkin, G.T., Frese, M., 2009. Entrepreneurial orientation and business performance: an assessment of past research and suggestions for the future. Entrepreneurship Theory and Practice 33 (3), 761–787 (accessed 10.06.2010.). Riley, D., Kalafatis, S., Manoochehri, F., 2009. The impact of entrepreneurial orientation and managerial resources on corporate venture outcomes. International Journal of Innovation Management 9 (3), 242–266. Ringle, C.M., Wende, S., Will, S., 2005. SmartPLS 2.0 (M3) Beta, Hamburg. . Rosenbloom, R.S., Christensen, C.M., 1995. Technological discontinuities, organizational capabilities and strategic commitments. Industrial and Corporate Change 3 (3), 655–685. Salavou, H., 2004. The concept of innovativeness: should we need to focus? European Journal of Innovation Management 7 (1), 33–44. Sandström, C., Osborne, R.-G., 2011. Managing business model renewal. International Journal of Business and Systems Research 5 (5), 461–474. Sandström, C.G., 2010. A Revised Perspective on Disruptive Innovation — Exploring Value, Networks and Business Models (PhD Thesis). Chalmers University of Technology, Göteborg, Sweden. Shafer, S.M., Smith, H.J., Linder, J.C., 2005. The power of business models. Business Horizons 48 (3), 199–207. Sharma, P., Chrisman, J.J., 1999. Toward a reconciliation of the definitional issues in the field of corporate entrepreneurship. Entrepreneurship Theory and Practice 23 (3), 11–27. Sosna, M., Trevinyo-Rodriguez, R.N., Velamuri, S.R., 2010. Business model innovation through trial-and-error learning: the Naturhouse case. Long Range Planning 43 (2–3), 383–407. Tang, J., Tang, Z., Marino, L., Zhang, Y., Li, Q., 2008. Exploring an inverted U-shape relationship between entrepreneurial orientation and performance in Chinese ventures. Entrepreneurship Theory and Practice 32 (1), 219–239. Tchong, M., 2010. Inventing from the Inside Out. (accessed 10.06.10.). Teece, D.J., 2010. Business models, business strategy and innovation. Long Range Planning 43 (2–3), 172–194. Tenenhaus, M., Vinzi, V.E., Chatelin, Y.-M., Lauro, C., 2005. PLS path modeling. Computational Statistics & Data Analysis 48 (1), 159–205. Walsh, S.T., Kirchhoff, B.A., Newbert, S., 2002. Differentiating market strategies for disruptive technologies. IEEE Transactions on Engineering Management 49 (4), 341–351. Walter, Z., Lopez, M.S., 2008. Physician acceptance of information technologies: role of perceived threat to professional autonomy. Decision Support Systems 46 (1), 206–215. Weill, P., Vitale, M.R., 2001. Place to Space: Migrating to eBusiness Models. Harvard Business School Press, Boston, MA. Wetzels, M., Odekerken-Schröder, G., Van Oppen, C., 2009. Using PLS path modeling for assessing hierarchical construct models: guidelines and empirical illustration. MIS Quarterly 33 (1), 177–195. Williams, L.J., Hartman, N., Cavazotte, F., 2010. Method variance and marker variables: a review and comprehensive CFA marker technique. Organizational Research Methods 13 (3), 477–514. Wirtz, B.W., Schilke, O., Ullrich, S., 2010. Strategic development of business models: implications of the Web 2.0 for creating value on the internet. Long Range Planning 43 (2–3), 272–290. Wright, P., Kroll, M., Pray, B., Lado, A., 1995. Strategic orientations, competitive advantage, and business performance. Journal of Business Research 33 (2), 143–151. Zahra, S., Covin, J.G., 1995. Contextual influences on the corporate entrepreneurship-performance relationship: a longitudinal analysis. Journal of Business Venturing 10 (1), 43–58. Zahra, S.A., 1991. Predictors and financial outcomes of corporate entrepreneurship: an exploratory study. Journal of Business Venturing 6 (4), 259–285. Ziv, N., 2009. The impact of Wireless on the Newspaper Industry: The New York Times as a Case in Point. Proceedings of Eighth International Conference on Mobil Business, June 26–28. Dalian University of Technology, China. Zott, C., Amit, R., 2010. Business model design: an activity system perspective. Long Range Planning 43 (2–3), 216–226.

Biographies Jahangir Karimi received the Ph.D. degree in Management Information Systems from the University of Arizona. He serves as the discipline director for Information Systems Program at the School of Business, University of Colorado Denver. His research interests include information technology management in national and international environments, IT-enabled E-business transformation, and new E-business models. Dr. Karimi has published in MIS Quarterly, Information Systems Research, Communications of the ACM, Journal of Management Information Systems, IEEE Transactions on Software Engineering, IEEE Transactions on Engineering Management, Decision Sciences, Journal of Systems and Software, Information and Software Technology, Concurrency: Practice and Experience, several books and conference proceedings. He is on the editorial board of International Journal of Electronic Commerce, and is a member of the Association for Information Systems. E-mail: [email protected] Zhiping Walter received the Ph.D. degree in Business Administration, specializing in Management Information Systems, from the Simon School of Business, University of Rochester. She is currently an Associate Professor of Management Information Systems at the Business School, University of Colorado Denver. Dr. Walter’s research interests are in the areas of online information seeking, online consumer behavior, and disruptive innovations. Her research has appeared in Communications of the ACM, Decision Support Systems, International Journal of Electronic Commerce, and European Journal of Operational Research, among others. E-mail: [email protected]