The Two Faces of Open Business Models Francesco Sandulli and Henry Chesbrough Introduction All companies have a business model. Some companies are not really aware of the existence of their business model, so they usually do not define it in a formal way. Other companies do grasp the characteristics of their business model which is often deployed in a formal and structured way. These companies tend to be in a better position to adapt their business model over time and thus continue to create value even when the competitive landscape changes. Business models that are open, in ways we will describe below, generally create more value, and enable greater adaptation over time. Yet these open business models are not well understood, nor is it clear how to implement such models. In this article, we seek to outline two different and equally important facets of open business models: a “buy side” perspective that incorporates external resources into one’s own business model, and a “sell side” view that places one’s own resources into others’ business models. Although there have been many different definitions of business model, usually there is some consensus on its two main functions: to define mechanisms for creating value and mechanisms to capture a certain proportion of that value (Chesbrough, 2006). The history of strategic management reveals how the limits within which business models were embedded have been expanded over time. A few years ago, when companies defined their business model they focused on value creation through the use of internal resources of the company. Over time firms initiated a process of opening up the boundaries of their business model as they began to analyze the contribution of suppliers, customers and other members of the firm’s ecosystem to the value created by internal resources of the company. However, at the present time, companies are beginning to ask why limit the use of their resources only to their business model. Why not include external resources as key components of the business model of the company? 1 Electronic copy available at: http://ssrn.com/abstract=1325682 Following this new approach, companies are beginning to share their internal resources with a third party to create value, or the reverse, companies are beginning to incorporate external resources in their own business model. These new business models have been defined by Chesbrough (2006) as open business models. The reason that companies are beginning to adopt open business models is to foster the creation of value. As we will describe later, for the company that shares its resources the open business model will improve the utilization rate of its resources and therefore the return on the required investment to develop them. Moreover, companies that incorporate other’s resources to its business obtain advantages in terms of speed of development of new products, easier access to markets, to knowledge and ultimately to resources that otherwise would have been more costly or difficult to develop. Here is the outline of the remainder of this article. First we explore the role of key resources, and how they can be aggregated into open business models. Next, we examine how the nature of key resources influences the development of these models. We pay particular attention to whether the resources are rivalrous or non-rivalrous, and whether they are excludable or not. We then turn to the two faces of open business models, considering the buy side and the sell side of how resources are accessed and leveraged respectively in such models. We then examine one organization in more detail, the El Bulli restaurant, to provide an indepth illustration of both sides of that firm’s business model, and show its adaptability. We conclude with summary remarks. The Open Business Model Companies create value using resources to perform a number of activities that produce utility to its customers. Not all business resources are equally important: a company may capture a share of this value to the extent that her resources are rare, difficult to imitate and have no close substitutes (Wernerfelt, 1984; Rumelt, 1984; Barney 1986; Barney 1991 among others). The sustainable competitive advantage of a company relies on its ability to identify, build, exploit, maintain and transform valuable resources (Teece, 2007). The firm’s business model articulates what is the value proposition of valuable resources for a target market, how to define the value chain and the network of relationships within the ecosystem of the company to allow the construction of these 2 Electronic copy available at: http://ssrn.com/abstract=1325682 valuable resources, what are the revenue mechanisms that enables the capture of the value generated by those valuable resources, and ultimately what is the competitive strategy that allows sustain over time the value of the resources of the company (Chesbrough and Rosenbloom, 2002). Following the traditional conception of a closed business model, most companies have tried to internally develop their resources and to exploit them within the boundaries of their own business model. In recent years in many industries, this approach is beginning to show some limits. On one hand, companies are increasingly aware that they do not have and cannot develop internally all the resources needed to succeed in increasingly complex environments. On the other hand, firms have also realized that the difficulties in fully capturing the value of their resources which remain underexploited. The management of innovation and knowledge assets has been one of the first areas where the firms realized the limits of closed business models. On one side, the increasingly rapid pace of scientific discoveries has increased the complexity of innovating with the firm’s own internal R&D resources, while at the same time, a paradoxically high share of the innovation done by the firms was not commercially exploited (Chesbrough, 2003). Resource Rivalry and Excludability Value capture is easier when the company shares resources that are excludable (Romer, 1986). However, sometimes the company shares resources such as open source software that are not excludable. In these cases, the company must seek alternative mechanisms to capture some of the value generated by the shared resource. For instance, West (2007) suggests a few of these alternative mechanisms in the case of open source software such as the dual license approach (a free license and a fee based license) or selling complements. Chesbrough and Appleyard (2007) describe the creation of open source business models, where value is also appropriated through lowering the cost of inputs or complements. Following the previous approach, we can deduce that opening the business model is easier when the resources shared by the firm are not rivalrous, as the capacity of the 3 Electronic copy available at: http://ssrn.com/abstract=1325682 resource is not a restriction, and excludable, as the firm is able to capture some of the value created by the third (Table 1). For this reason, it will be more likely to observe open business models based on assets such as brand, tacit knowledge and intellectual property. In relation to rival resources, the greater its excess capacity and its scalability greater the likelihood that they can support open business models. An example of a rival resource that supports an open business model is the technological infrastructure of Unience, a social network for investors where financial firms can build and offer financial assets management tools to those investors. As it is based upon the Amazon Cloud Computing concept, Unience's infrastructure is easily scalable, so that capacity restrictions are lower. However other resources less scalable as the logistics infrastructure or the human resources may face more limitations when used in open business models. As discussed above, supporting an open business model on non-rival resources presents the difficulty of capturing value, and therefore it would be only feasible if the company is able to detect valid mechanisms to capture value. However, this task is not easy as shown by the so far unsuccessful attempt by the recording industry to identify alternative mechanisms to capture value on P2P networks (Sandulli, 2007). Finally, rival and nonexcludable resources are characterized by limited capacity and the difficult to capture value, so they are hardly applicable to an open business model. Table 1. Feasibility of opening the business model in light of the nature of the assets. Resource Rivalry Resource Excludability Feasibility of an Open Business Model EXCLUDABLE High RIVAL Low NON RIVAL High Logistics Infrastructure Distribution Channels Human Resources IT Infrastructure Manufacturing Capabilities Information Market Knowledge Intellectual Property Brand 4 NON EXCLUDABLE Low Open Wireless Networks Public Goods, e.g., public infrastructure Open Source Software Unprotected Knowledge Secondly, the shared resource should not be competitive. This implies that sharing the resource with another party does not reduce the value created by that resource in the company’s own business model. The exception to this principle is when sharing an asset reduces the value it generates in the business model of the company, but at the same time the company is able to offset this loss through the capture of a proportion of the value that the third party is creating. In the case of the shared network in the Telecom Industry we describe below, when a carrier opens its APIs, some of the third party applications may come into direct competition with the carrier’s own applications. In this case, the shared resource is competitive as it is eroding the carrier normal revenue sources. However at the same time, sharing the network may result in increased revenues from the network traffic and the royalties the telecom operator gets from sharing its resource with the application provider. As long as these new revenues offset the loss in terms of traditional revenue sources, the carriers will be interested in opening up the business model. Having described the role of resources in constructing an Open Business Model, we will move on to describe some challenges related to the implementation of an Open Business Model, both from the standpoint of the company that decides to share its resources with others, and from the point of the company that decides to use third-party resources in her business model. Use of External Resources in the Firm’s Business Model: The Buying Side of the Open Business Model. Opening the business model in the stages of identification and validation of the opportunity is closely related to the practices of Open Innovation. People and organizations outside the company are the source of new ideas and new business opportunities for the firms. An increasing number of organizations are using external 5 sources of knowledge for their R & D projects. For example, in 2006 IBM opened its Innovation Jams, a forum created in 2001 in which employees contribute ideas on which to base future research and development programs of the company, to individuals outside the organization. The new open Innovation Jam involved 13,366 people, generated 8,674 ideas, of which 31 became Great Ideas finalists, and finally 10 received seed funding and were included into the IBM portfolio of research programs (Helander et al., 2007). IBM now offers clients the opportunity to host their own Innovation Jams through its IBM Global Services unit. And IBM is far from alone in this. Airbus, Google, Microsoft, or the Spanish companies the Vocento Group or the building company Visesa, have launched Ideas Challenges for identifying new areas of innovation. Moreover, Innocentive, Innovationxchange, Ninesigma, Big Idea Group or Conectainnova in Spain are a bunch of new ventures whose business model is based on intermediating between organizations that need ideas and people and organizations that may be a potential source of ideas. Crowdsourcing has not been used in the Innovation field until recently. Spanish Savings Banks must give away through the trust the net profit which is not allocated to reserves. Caja Navarra has recently started a pioneering project in Civic Banking where the customers decide in which social projects the savings bank must invest. Startup companies like Threadless.com utilize the crowd to submit designs for T shirts, and then produce the ones receiving the highest approval from that crowd. Not all organizations that are adopting Open Innovation to identify new ideas do Crowdsourcing. Some prefer to have more control over the dynamics of ideas sourcing and the relationship with the innovation partners. Philips has opened its Innovation Campuses in Eindhoven and Shanghai. Where once stood a restricted series of facilities accessible only to Philips staff, now stands an open industrial park, where the company shares ideas with a selected and limited group of companies. Another example is the small user focus groups employed by Shimano in the development of the new bike Coasting Component Group. There is a movement toward the creation of science parks within Spain, intended to increase the utilization of university technologies in Spain. Kock and Torkkeli (2008) 6 have found some initial empirical evidence (based on an unrepresentative sample) that Spanish firms are more prone to lead to an efficient and effective transfer of knowledge. Another way to open the business model is working with others to validate and test the new ideas or products of the firm. Those companies that adopt this practice can expand the scope and scale of their experiments and also reduce the resources required and the product development time. This has been a powerful advantage of open source software for many years. This concept is now moving rapidly outside of software development to other areas of business. New mobile application test is one of the product development tasks which absorbs more time and resources in the mobile telecommunication industry (Nesse, 2008). Telefonica, O2, BT, Telenor and Verizon among other carriers have established communities of users to validate the future applications and expect to get significant reductions in the validation times. Philips has partnered with NH Hoteles to conduct experiments on new lighting systems in Hoteles. Facebook users are being asked to determine the packaging design for Lancashire Tea. In the Telecom Industry several carriers decided to open their Application Program Interfaces (APIs). Open MovilForum at Telefonica Moviles, 21C project at British Telecom, Telenor Content Provider Access, Telecom Italia nextTime, SingTel Partners Program, ODI Verizon, Sprint Business Mobility Framework or AT & T devCentral are some of these open API projects. The goal of these carriers is to provide for the easy, rapid and cost-effective development of mobile applications. A potentially large base of application developers have now access to the carrier network, which is a valuable resource as most of them could not develop by themselves such a distribution channel for their services. On the other side, telecom companies expect both to share with the application providers the revenues generated by these new services and to increase the network traffic, since a network with more services provides more value to the operator’s customers. Open business models can even lead to cooperation with competitors. This cooperation can be justified by the need to obtain synergies between resources (Das and Teng, 2000). A recent example of this phenomenon is the cooperation between Trek, Giant 7 and Raleigh, competitors in the U.S. bicycle market, to design new bikes with the new Shimano Automatic Coasting system. The bike manufacturers had to share ideas and experiences due to the high technical complexity of the new system. The cooperation between competitors is also explained by the need to share costs. For example, it is common that in the initial stages of market entry small businesses develop collective brands to reduce the costs of communication and marketing (Falcone, 2007). For instance, using the social network Hi5, 29 small textile companies in Costa Rica have created the fashion brand Abril to commercialize all their products. Absorptive Capacity and Organizational Inertia Companies that decide to use third-party resources into its own business model faces a number of challenges associated with both their absorptive capacity and their organizational inertia. The absorptive capacity is a concept developed in the literature that analyzes the exchange of knowledge among firms (Cohen and Levinthal, 1990). Absorptive capacity is related to three capabilities: resource search, integration and exploitation. Applying this concept to our open business model framework, we can say that the success of a company that decides to use third-party resources depends on her ability to identify resources that may create value, her ability to integrate external and internal resources, and her ability to exploit the external resource. The search for new resources is an activity with diminishing returns: beyond one point, broader and deeper resources search does not improve the results in terms of better resources found (Ahuja and Katila, 2004). This means that the search for resources can not be indiscriminate and that companies should set limits on the breadth and depth of their search (Laursen and Salter, 2006). In fact, excessive attention to the search for external resources can undermine the development of internal resources. For instance Lou and Round (2006) show that in the pharmaceutical industry the search for inlicense compounds may adversely affect the development of self-originated compounds. This external resources search process should not only consider the resource potential value, but must also take into account other aspects such as partner commitment and trust, the complementarity as it will impact the company's ability to integrate the resource within her operations, the complexity of managing the partnership, the 8 outcomes uncertainty and the ability to measure the exact outcomes of the partnership (Shah and Swaminathan, 2008). So, companies will be more likely to integrate into their business models external resources that are owned by trustful and committed partners, that complement internal resources and fit the firm’s strategy, and whose value contribution is not uncertain and can be measured. The need for low uncertainty, complementarity and partner trust and commitment may explain why the approach to open innovation of Spanish firms relies more on repeated interactions with the same partners, instead of pursuing an unbounded partner search (Kock and Torkkeli, 2008)1. This argument may also explain why in the initial stages of development of an open business model, most telecom carriers or Google in its Android project, have opened the APIs only to a selected and small group of application providers. The second important component of the absorptive capacity is the firm’s ability to integrate external resources. Within the open business model framework, resource exchanges are quasi-market mechanisms. These mechanisms will be more inefficient than hierarchies the higher the complexity in the management of the resources exchange (Roper et al., 2007). For this reason, the more complex the resource integration, the less appropriate will be an open business model for the firm. So far current research has given just a few clear clues about the capabilities a firm needs to integrate internal and external resources. Acha (2007) noted the importance of the ability to design the points of interaction between firms and the task partitioning between the firms. Zahra and George (2002) suggested the importance of the ability to develop information systems that support the resources exchange. The ability to integrate external and internal resources is usually built up over time through a trial and error process (Cohen and Levinthal, 1990). Therefore we should expect that the resource integration will be easier in firms with more experience in the management and development of similar assets (Cohen and Levinthal, 1989). 1 We have found some evidence apparently contradicting the Kock and Torkkeli (2008) conclusions. In fact, firms facing problems with great uncertainty tend to search widely across a variety of search channels. Companies with crowd sourcing initiatives, such as the IBM Innovation Jams, are examples of this wide search. However, all these firms that use crowd sourcing to fish new ideas, beyond the search phase tend to concentrate their efforts in very few ideas, confirming our argument that in high uncertainty environments openness has high costs. 9 Time also is crucial in developing a large base of complementary assets, which will be needed to successfully exploit external resources. Kock et al., (2007) suggest that large firms will be more likely to use external resources in their business models because the larger the asset base of the firm, the stronger the potential synergies for the integration of external resources. Finally, the company that incorporates external resources must be able to overcome its own organizational inertia. The integration of external resources often requires significant organizational changes. However, the high level of standardization of organizational routines can make the company reluctant to implement those changes and thus to integrate resources (Henderson and Clark, 1990). In the area of Open Innovation, organizational inertia produces syndrome "not invented here", by which an organization rejects external sources of knowledge simply because they were not developed internally (Katz and Allen, 1982). The firm Trek is a recent example of this syndrome. This bicycle manufacturer used to support only internal knowledge of the market in defining the requirements for its future products. However, shifting to the Open Innovation approach, part of the requirements of the new model Trek Lime were obtained directly from outside the company, precisely from occasional bicycle users through a series of focus groups. At the time of translating those user based requirements into the design, the firm needed some effort to convince the product development engineers of the benefits of the new design, as the engineers were very reluctant to abandon their traditional design criteria very focused on optimizing the aerodynamics and weight of the bicycles, to adopt the consumer proposed approach more focused on functionality and aesthetics of the bike. Some of the practices that may be considered to overcome organizational inertia by those companies with open business models are cyclic reorganizations (Tushman and Romanelli, 1985) or top-management turnover (Tripsas and Gavetti, 2000). The figure below summarizes our discussion of the buy side face of an open business model. Factors on the left of the figure cause models to be less open, such as when absorptive capacity is poor. Right hand side factors, by contrast, support more open business models, such as when absorptive capacity is high. 10 Figure 1. Determinants of Open Business Models in the Buying Side. OPENNESS BUYING SIDE + - Absortive Capacity - Internal Complementary Resources + Organizational Inertia + Outcome Uncertainty + Absortive Capacity + Internal Complementary Resources -Organizational Inertia - Outcome Uncertainty Using the Firm’s Resources in the Business Model of Other Companies: The Selling Side of an Open Business Model. There are a number of companies that have decided to exploit their resources and capabilities beyond its own business model. These companies create and capture value from sharing their resources and capabilities. Companies such as Bankinter bank or the Barrabés mountain gear online store have a deep understanding of the technology used in their business model. Opening the business model and sharing this asset with other banks in the case of Bankinter or other online shops in the case of Barrabés has created a new revenue source. Companies such as Philips are obtaining a significant flow of revenues from sharing their knowledge under patenting agreements. Besides technology knowledge, some telecom operators such as Telefonica, Telenor and BT are beginning to share market data with other companies. Due to the non rival nature of most intangible assets such as knowledge or brand (see for example Von Hippel and Von Krogh, 2003) it is easier to share them, yet some companies have also started to share the excess capacity of some of their tangible assets such as technological infrastructure, logistics assets or distribution channels as well. 11 Amazon shares with other firms its computing capacity through its Elastic Compute Cloud, Barrabés shares its logistics assets through the Rezepcion@ project, Ponoko shares its manufacturing platform, Cal Celdoni share its vineyard and its manufacturing capabilities, and telecom carriers such as Telefonica share the network. Rational behavior suggests that a firm is willing to share a resource with a third party when exploiting a resource in third party business models generates more value than constraining the resource to the firm’s own business model. This is fairly clear in the case of sharing the excess capacity of tangible assets. Sharing IT or logistics infrastructure with other firms increases their rate of utilization and the return on the investment on that infrastructure. A firm that shares a resource can also generate value through the creation or further exploitation of complementary and cospecialized resources (Teece, 1986). For instance, IBM shared with the open source community all of its Eclipse code, valued at $40 million (Stolinsky, 2004). Sharing this resource, IBM aimed at increasing the value created by the Rational product line, a suit of development tools that complemented the free Eclipse platform. Open telecom platforms are another example of a shared asset that increases the value of complementary resources. As discussed above, mobile telecommunications operators have decided to share its network with application developers. This decision may significantly boost the amount of services offered by the network, thus increasing the value of this resource. In turn, the higher value of the network can make potential distributors more interested in distributing the carrier services, thus increasing the distribution channel reach. This example shows that opening the business model can increase both the network value and the value of complementary resources such as the distribution channel. Indeed, opening the business model seems an interesting move for those carriers entering new markets or segments, or for those operators without a strong distribution channel. Companies must decide the extent to which share its resources with others. In this regard companies must take into account the breadth and depth of the open business model. Broadening the approach of Laursen and Salter (2006) we will consider that breadth refers to the number of actors with whom you share a resource, while the depth refers to the degrees of intensity and specificity of the relationship with those actors. Normally, an open business model requires a tradeoff between breadth, depth and control over the shared resources. In the extreme, an open business model would be 12 both broad and deep. Namely, the firm would share her resources with a large amount of partners and will actively manage all these partnerships so that she can capture the maximum amount of value from theme. However, keeping deep relationships involves high coordination costs. For this reason, it is more likely that the firms will deploy either a broad but not deep business model such as the application development communities in the telecom industry, where a large amount of partners share the resource with the firm, who tends to establish standard relationships with them and is not very involved in the partner’s daily management of the resource, or a narrow and deep business model as is the case with the cobranding strategy of El Bulli we will discuss in a later section, where the firm shares a resource with few partners, establishes specific relationships with each of them, and is actively involved in the management of the resource in the third party business model. In defining an open business model, companies must first consider the opportunity cost of sharing the asset. This opportunity cost is a function of the value of the asset and the uncertainty in the outcome of the exchange. Higher the value of the asset, higher the opportunity cost and less broad the open business model. This explains why large firms prefer to exchange secondary or non core assets (Kock et al., 2007) or why when the initial value of the platform is high, the sponsors of technology platforms prefer to take profits directly instead of relying in external innovation (Parker and Van Alstyne, 2008). In the case of cospecialized, highly specific and therefore very valuable assets such as brands, companies prefer to establish dyadic relationships, as the risk in the event of loss of control is very high. In the fashion industry brand is one of the most valuable resources. When in the 70s the industry suffered a major crisis, some companies like Gucci or Calvin Klein decided to license their brands to a large number of companies. In the short term, the strategy worked generating substantial license fees. However, over time the resource, the brand equity, was eroding, as many of the licensed products were of poor quality. The lack of control over the brand pushed Gucci near bankruptcy, and Calvin Klein was unable to find an investor interested in buying the company in 1999. Today both companies, as well as others such as Dior, Givenchy and Yves St. Laurent (Hill et al., 2001), are struggling to wriggle free of their licensing deals and to regain control of the brand. 13 The openness degrees of the business model also depend on the expected capability of the partners to create follow-on products using the shared resource. If the company believes that her partners are very capable of creating these follow-on products, the model will tend to be more open. In these cases, the companies share their resources with the aim of developing a large number of products and services. This was usually the case of platforms sponsors such as Reuters, Unience, AirSage, operators of telecommunications or Microsoft who decided to open their APIs. For example, Microsoft has granted access to the source code to boost up the development of the greatest number of open source applications that run on Windows rather than on Linux. In an optimal situation a company should have a high number of partners with capabilities to create significant value. However, in many cases the company does not know ex ante the capabilities of potential partners and does not know which of them will be able to use the resource in a more valuable way. That is why some companies prefer to define business models with a high number of partnerships. Indeed, the breadth of the business model reduces the uncertainty associated with the future follow-on products of the resource (Parker and Van Alstyne, 2008). Many companies with a broad open business model such as Reuters, Amazon, Telefonica, Microsoft or Salesforces.com in sharing some of their resources pursue Long Tail business models, trying to create a market with a large number of complementary products that individually generate a small amount of revenues. In order to increase the number of developers of complementary products, companies that own the resources may even decide to sacrifice some part of the value amount they intended to capture from the shared resource. For example, Indian mobile carriers prefer to sacrifice some of their revenue share to promote the development of a strong Value Added Services market, the lowest for those applications developed on new platforms for innovation, or Intel established subsidies to the community of producers of complementary products for its chips. Parker and Van Alstyne (2008) argue that the sponsor of a technology platform prefers to give developers pricing power, especially in the early stages of a technology where there is increased uncertainty about the outcome of the resource exchange. In these early 14 stages of a technology uncertainty will also related to deeper and more narrow open business models. As discussed above, in the early stages of their open business model both Google with Android or the telecom carriers preferred to open their APIs to a few selected application developers. Laursen and Salter (2006) confirm that in environments of radical innovation is more common for companies to establish close relations with a very small number of partners. In this sense, it seems that greater uncertainty can justify the management of resource portfolios using more hierarchical structures and business models less broad but deeper, as opposed to managing those resources using modular approaches which are closer to market mechanisms (Chesbrough and Kusunoki, 1999). Firms with Long Tail open business models share either resources that are not rival, such as knowledge or brands, either highly scalable resources, such IT infrastructure. We must take a closer look to the coordination costs. Strategic alliance theory suggests that the higher the mutual interdependence among partners, the greater the coordination costs (Gulati and Singh, 1988). Establishing highly interdependent relationships will difficult establishing open business models based on the Long Tail approach. Thompson (1967) argues that coordination costs are lower when the relationship is coordinated by standardization. Therefore, to mitigate the coordination costs associated with maintaining a high number of partners and with Long Tail open business models, companies must rely on standards that articulate the relationship with those partners. For instance, when the firm is sharing codified knowledge it is easier to establish Long Tail business models. For instance, several Long Tail open business models are based on open APIs platforms that are developed on standard programming languages. On the other hand, sharing knowledge which is not based on standards or resources such as brands whose management implies specific partner interaction requires a greater effort of coordination. In the fashion industry, Gucci or Calvin Klein tried to follow twenty years ago a Long Tail open business model consisting of sharing a resource such as the brand with a large number of partners. For example, Gucci had 13,000 licensed products. This large amount of partnerships made it very difficult or almost impossible to those two firms to effectively manage all these relationships. For months at a time, Calvin Klein did not attend design meetings with Warnaco, the licensee company for jeanswear brands who eventually diluted the CK trademark by selling it into discount channels. 15 The risk and the protection against opportunistic behavior should also be taken into account when defining the breadth of the open business model. The lower the level of trust in the potential partners and the higher the anticipated complexities in defining and enforcing property rights, the more hierarchical the relationship with the partners that will exploit the firm’s resources (Gulati and Singh, 1998). These factors are summarized in the Figure below, in which factors that lead to more controlled and narrow use of resources are on the left of the Figure (such as high mutual interdependence), while those that support greater sharing of resources with more parties are shown on the right (such as low mutual interdependence). Figure 2. Determinants of Open Business Models in the Selling Side. OPENNESS SELLING SIDE + + Mutual Interdependence - Partner Trust - Legal Protection + Internal Complementary Resources -External Complementary Resources - Scalability - Mutual Interdependence + Partner Trust + Legal Protection - Internal Complementary Resources + External Complementary Resources + Scalability According to the degrees of openness of the selling or the buying side of a business model we have found three different types of open business models: Partially Open Business Models (Buying-Side): In this category usually fall firms that lack the capabilities to develop a specific resource by themselves, so they embed third party resources in their business models. Within this category we can find firms such as NH Hoteles, Lavazza, Kraft or Procter and Gamble that are “customers” of companies with open business models in the Selling Side, firms that cooperate to create a collective and shared resource, such as 16 Trek, Giant or Raleigh creating common knowledge on the new Shimano’s Automated Coasting System or the consortium of 29 Costa Rican firms creating the Abril brand. Partially Open Business Models (Selling-Side): These are business models of firms that share some of their resources with another firm which embeds them into their business model. Firms following this business model try to maximize the rate of utilization of the shared resource to maximize utilization rate, such as the Brand licensees, reduce excess capacity, such as Amazon, obtain economies of scope for that resource such as Barrabés or Bankinter, or to maximize the production of complementary products which is normally the case of owners of platforms such as Unience, Apple, Facebook, SAP, Telecom Carriers, Google, Microsoft or Reuters. Fully Open Business Models: Those are firms that rely on both sides of an open business model. We find firms such as El Bulli, IBM, Amazon, Merck or Philips that relied on external resources to reinforce their business model and to experiment with new business models, while also creating new revenue flows from sharing the use of their resources to other firms. We also included within this category a new group of firms such as Innocentive or Conectainnova that are specializing in intermediating between firms with open business models in the buying side and firms with open business models in the selling side. Except for these intermediaries that already born open, there is a natural evolution from Closed Business Model to Partially Open Busines Models, and then to Fully Open Business Models. Merck or NH Hoteles are examples of firms that are evolving from Partially Open Business Models in the Buying Side to Fully Open Business Models incorporating the selling side of openness. More frequent are the cases of firms that are evolving from Partially Open Business Models Selling Side to Fully Open Business Models, especially when firms such as SAP, Telefónica, Google or Apple are platform sponsors and really expect a large crop of complementary partners from opening their business model. 17 Figure 3. A Taxonomy of Open Business Models. Partially Open Business Model (Buying Side) NH Hoteles, Lavazza, Kraft, Procter and Gamble, Trek,Abril Opennes Buying-Side Open Business Model Philips, IBM, Amazon, Merck, El Bulli, Conectainnova, Innocentive Opennes Selling-Side Closed Business Model Partially Open Business Model (Selling Side) Telecom Carriers (Telefónica, BT, Telenor,…), Google, Microsoft, Reuters, Shimano, Unience, SAP, Facebook, Gucci,Apple, Calvin Klein, Barrabés, Bankinter. El Bulli – An Illustration of an Open Business Model We have described a number of factors that affect the two faces of an open business model. In this section, we bring these points together through an examination of a unique organization, the world renown restaurant El Bulli. Based in Spain, El Bulli is widely considered one of the world’s best and most influential restaurants. Restaurant Magazine voted the restaurant #1 in the world four times, and the restaurant received its third Michelin star in 1997, and has held it ever since. Don’t rush to book a reservation however. The restaurant receives roughly 1 million requests for its 8,000 seatings each year (it closes for half the year each year for further research and development), and reservations are typically taken one day in October a year in advance. 18 The restaurant and its chef, Ferran Adria, are best known for its investigations into “molecular gastronomy”. By examining the micro-properties of specific foods, spices and ingredients, the restaurant develops unique recipes that provide radically new dining experiences. But this knowledge did not originate inside El Bulli. Illustrating the first face of an open business model, Adria was among the eager collaborators of a series of discoveries in molecular gastronomy by Hervé This, a French physical chemist. Furthermore, El Bulli was a participant in the European Union funded project called INICON, which intended to promote collaboration between scientists, chefs and restaurants. While El Bulli has made tremendous contributions to this movement, the core ideas initially emerged outside the organization, and were absorbed inside. Recently, the School of Engineering and Applied Sciences at Harvard agreed to bring scientific expertise and techniques in the formulation of foods, textures and structures to el Bulli. While its gastronomy is widely celebrated, El Bulli’s business model is largely unstudied. As noted above, the restaurant closes for six months every year, and reportedly loses money. However, the restaurant’s role in El Bulli’s model is one of an R&D laboratory, which is not expected to earn a profit by itself. Instead, the restaurant generates the knowledge needed for the profitable elements of El Bulli’s business model. Most of these elements illustrate the second face of an open business model, by allowing others to share El Bulli’s resources, including its brand and its knowledge. In 1999 the restaurant decided to share its knowledge with the food manufacturer Borges to design oils, sauces and snacks. Essentially this meant that Borges launches a series of co-branded items with El Bulli in the consumer marketplace. This added a new revenue stream for El Bulli. Other similar co-branding deals included collaborations with Kaiku (an award winning book of recipes), Lavazza (coffee), NH Hoteles (hospitality), Nestle (chocolate), Armand Bassi (tableware and kitchenware) and Diageo Group (whisky cocktails). In another area, a recent collaboration was launched between NH Hoteles and Iberia in 2007 to include Fast Good sandwiches in the menu onboard Iberia. This benefited El Bulli by positioning its brand into a new segment. 19 This last relationship between NH Hoteles and El Bulli deserves further comment. In order to reinforce the brand experience of their customers and to positioning in the market as an innovative organization, NH Hoteles decided to create the concept of “Fast Good”, a fast food restaurant delivering food of better quality, and Nhube, an initiative that seeks to combine in a single space the lounge, the restaurant and the café-bar of the hotels. To complete these projects NH Hoteles decided to rely on the capabilities and resources of the restaurant Bulli and Ferran Adrià. Through the cooperation with El Bulli, NH Hoteles was granted the access to resources such as the El Bulli’s expertise in creating brand experiences or the El Bulli brand itself that reinforced the positioning of NH Hoteles as an innovative brand. The cobranding path of El Bulli, based on a few and deep specific partnerships, aims at not making the same mistake and maintaining control over the brand. El Bulli restaurant prefers dyadic and deep relationships with their partners because one of the resources it shares is process technology which is more difficult to protect legally. The high opportunity cost explains why these firms prefer to follow narrower and deeper business models (Kock et al., 2007). In sum, El Bulli’s business model embodies both faces of an open business model. On the one hand, it actively engages with external research sources to learn about new breakthroughs in gastronomy. It closes for extended periods of time in order to identify and then absorb these new breakthroughs, which keeps El Bulli one step ahead of other restauranteurs who strive to copy its successes. And El Bulli commercializes its brand and its knowledge through a variety of closely managed relationships that have taken its brand into a number of areas beyond those of a typical restaurant. So while the restaurant itself does not make money, the company overall is profitable. Conclusion Companies operating in turbulent times need to become more agile and more open in developing and advancing their business models. As we have shown here, some firms have already or are in the process of developing open business models. In the past, those few firms with open business models were usually open in a specific function of their business model such as product development, internationalization or distribution, while 20 the rest of the business model remained close. Today, firms are in the process of redesigning all the aspects of their business models under the new open prism. While openness is not a panacea, and is not always better, the two faces of open business models provide a framework for companies to evaluate how best to innovate their own business models. On one hand, the buy side offers many new ideas for a company’s business model. An open approach may reduce the costs and the risk of experimenting with new business models, transforming fixed costs into variable costs. On the other hand, the sell side provides numerous possibilities for leveraging ones’ resources in others’ business models. The sell side creates new solutions to the always difficult task of maximizing the rate of utilization of the firm’s resources, and allows higher investments in resources that would not be justified by the returns from the firm’s own business model. El Bulli provides a stimulating illustration of both faces at work. We hope that it may serves as an inspiration for the business creativity of others who seek to create and capture value in more open and distinctive ways. References Acha, V. (2007) Open by Design: The Role of Design in Open Innovation. Department of Innovation, Universities and Skills. Ahuja, G., y Katila, R. 2004. “Where Do Resources Come From? The Role of Idiosyncratic Situations”. Strategic Management Journal, 25(8-9), p. 887-907. Baldwin, C. Y. y Clark, K.B. (2000). Design Rules. Volume 1: The Power of Modularity. MIT Press: Cambridge, Ma. Barney J. (1986) “Strategic factor markets: expectations, luck and business strategy”. Manage Science, 42, p.1231– 1241. Barney J. (1991) “Firm resources and sustained competitive advantage”. Journal of Management, 17, p. 99– 120. 21 Chesbrough H.W. y Kusunoki, K. (1999). “ The modularity trap: innovation, technology phase shifts, and the resulting limits of virtual organizations”. En Nonaka I. y Teece, D.J. (Eds.) Managing Industrial Knowledge—Creation, Transfer and Utilization, SAGE Publications, London, p. 202–230. Chesbrough, H.W. y Rosenbloom, R.S. (2002). “The role of the business model in capturing value from innovation: evidence from Xerox Corporation's technology spinoff companies”, Industrial and Corporate Change, Vol. 11 (3), 529-555. Chesbrough, H.W. (2003) Open Innovation: The New Imperative for Creating and Profiting from Technology, Boston, MA, Harvard Business School Press. Chesbrough, H.W. (2006) Open Business Models: How to Thrive in the New Innovation Landscape, Boston, MA, Harvard Business School Press. Chesbrough, H. y Appleyard, M.M. (2007), “Open Innovation and Strategy ”, California Management Review, 50(1), p. 57-76. Cohen, W.M. y Levinthal, D.A. (1989). “Innovation and Learning: Two faces of R&D”. Economics Journal, 99, p. 569 – 596. Cohen, W. M. y Levinthal, D. A. (1990). “Absorptive-Capacity - a New Perspective on Learning and Innovation”. Administrative Science Quarterly, 35, p. 128-152. Das, T.K. y Teng B. (2000). “A Resource Based Theory of Strategic Alliances”. Journal of Management, 26(1), pp. 31-61. Dyer, J.H. y Singh H. (1998).“The relational view: Cooperative strategy and sources of interorganizational competitive advantage”. Academy of Management Review, 23(4), p. 660-679. Falcone P. (2007) “Small and Medium Enterprise Clusters: Marketing and Communication Management” en MacGregor R.C. y Hodgkinson A.T. (Eds.) Small 22 Business Clustering Technologies, Application in Marketing Management, IT and Economics, Idea Group, Hershey, PA, US, p. 29-60. Gulati, R. y Singh, H. (1998). “Managing coordination costs and appropiation concerns in Strategic Alliances”. Administrative Science Quarterly, 43 (4), p.781-814. Helander, M; Lawrence, R.; Liu, Y.; Perlich, C.; Reddy C. y Rosset, S. (2007). “Looking for Great Ideas: Analyzing the Innovation Jam”. The 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 12-15 Agosto, San José (EEUU). Henderson, R.M. y Clark, K.B. (1990) “Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms”. Administrative Science Quarterly, 35, p.9-30. Hill, S., Lederer, C. y Lane, K. (2001). The Infinite Asset: Managing Brands to Create New Value. Harvard Business School Press, Boston, Ma. Katz R. y Allen, T.J. (1982). “Investigating the not-invented-here (NIH) syndrome: A look at the performance, tenure, and communication patterns of 50 R&D project groups”. R&D Management, 12, p. 7–19. Kock, C. J. y Torkkeli, M. (2008) “Open Innovation: A 'Swingers' Club' or 'Going Steady'?”. Available at SSRN: http://ssrn.com/abstract=1134489 Kock, C.; Torkkeli, M. y Salmi, P. (2007) “The limits of the "Open Innovation" paradigm, or: When does it pay to be open?”. Strategic Management Society 27th Annual Conference, San Diego, EE UU. Laursen, K. y Salter, A .(2006) “Open for innovation: the role of openness in explaining innovation performance among UK manufacturing firms”, Strategic Management Journal, 27, p. 131-150. 23 Lou K. y de Rond, M. (2006). “The not invented here myth”. Nature Reviews Drug Discovery, 5, p. 451-452. Nesse P.J. (2008). “Open service innovation in telecom industry – case study of partnership models enabling 3’rd party development of novel mobile services”, 12th International Conference ICIN, Bordeaux, France. Parker, G. y Van Alstyne, M. (2008). Innovation, Openness and Platform Control, (July 24, 2008). Available at SSRN: http://ssrn.com/abstract=1079712. Romer, P. M, (1986). “Increasing Returns and Long-run Growth,”. Journal of Political Economy, vol. 94(5), p. 1002-37 Roper, S.; Love, J.H. y Du J. (2007). “The Limits Of Open Innovation: Openness And (quasi-)markets In The Organization Of Innovation”. Aston Business School Working Paper, RP 0713. Rumelt R.P. (1984) “Theory, strategy and entrepreneurship”. En: Lamb R.B. (Ed.). Competitive strategic management. Englewood Cliffs, NJ: PrenticeHall, p. 556– 70. Sandulli, F.D. (2007). “CD music purchase behaviour of P2P users”. Technovation, 27 (6-7), p. 325-334. Shah, R.H. y Swaminathan, V. (2008). “Factors influencing partner selection in strategic alliances: The moderating role of alliance context”. Strategic Management Journal, 29, p. 471-494. Stolinsky, P. (2004). “The IBM Software Development Platform partner ecosystem”. The Rational Edge. July 2004. http://www.ibm.com/developerworks/rational/rationaledge/index.html Teece D.J. (1986). “Profiting from technological innovation”. Research Policy, 15(6), p. 285–305. 24 Teece D.J. (2007). “Explicating dynamic capabilities: the nature and microfoundations of sustainable enterprise performance”. Strategic Management Journal, 28(13), p.1319– 1330. Thomson, J.D. (1967). Organizations in action: Social Sciences Bases of Administration. McGraw Hill: New York. Tripsas, M. y Gavetti, G. (2000). “Capabilities, cognition and inertia: Evidence from digital imaging”. Strategic Management Journal, 21, p. 1147–1161. Tushman M.L. y Romanelli E. (1985). “Organizational evolution: a metamorphosis model of convergence and reorientation” en Cummings L.L. y Staw, B.B. (Eds.) Research in Organizational Behavior, Vol. 7, JAI Press, Greenwich, CT; p. 171–222. Von Hippel, E., y Von Krogh, G. (2003). “Open source software and the privatecollective innovation model: issues for organization science”. Organization Science, 14 (2), p. 209–233. Wernerfelt B. (1984). “A resource-based view of the firm”. Strategic Manage Journal, 5, p.171– 180. West, J. (2007). “Value Capture and Value Networks in Open Source Vendor Strategies”. Proceedings of the 40th Annual Hawaii International Conference on System Sciences, Hawaii, EE UU. Zahra, S. A. y George, G. (2002). “Absorptive Capacity: A Review, Reconceptualization, and Extension”. Academy of Management Review, 27(2), p. 185203. 25