Open Innovation Strategy in Indonesia ICT Industries

The research objective is establishing the relationship between business model, open innovation, and a firm’s performance in a context-based corporate accelerator program held by Indonesia ICT industries. The crucial relationships between startup and industry collaboration that were identified in the literature study are: (1) co-creation; (2) the need for IP-sharing; (3) reducing R&D costs; (4) the need for risk sharing; and (5) the primary driver of companies’ sustainability growth. The result shows that “co-creation” represented by business models becomes the most reliable driver in the issue of the importance of collaboration, and it leads to an independent factor that should be firstly developed and continuously improved. The IP-sharing and risk-sharing factors are both reliable drivers but medium dependent; thus, the relation is in a linkage zone, which means that these two factors are unstable or sensitive for both sides. The startups and industries must have a proper contractual agreement once they decide to share the IP and risk. The integrity factor will be necessary for mutual benefit. Reducing R&D costs and the main driver of the company’s sustainability growth are the dependent variables. or outcomes of the collaboration between startups and leading industries.

The research objective is establishing the relationship between business model, open innovation, and a firm's performance in a context-based corporate accelerator program held by Indonesia ICT industries. The crucial relationships between startup and industry collaboration that were identified in the literature study are: (1) co-creation; (2) the need for IP-sharing; (3) reducing R&D costs; (4) the need for risk sharing; and (5) the primary driver of companies' sustainability growth. The result shows that "co-creation" represented by business models becomes the most reliable driver in the issue of the importance of collaboration, and it leads to an independent factor that should be firstly developed and continuously improved. The IP-sharing and risk-sharing factors are both reliable drivers but medium dependent; thus, the relation is in a linkage zone, which means that these two factors are unstable or sensitive for both sides. The startups and industries must have a proper contractual agreement once they decide to share the IP and risk. The integrity factor will be necessary for mutual benefit. Reducing R&D costs and the main driver of the company's sustainability growth are the dependent variables or outcomes of the collaboration between startups and leading industries.
Industries realize that innovation is a complex and multi-factorial challenge, and it depends on the industrial environment to obtain more potent competitive advantages (Sivam et al., 2019).
In the ICT industry, technological turbulence is high, and the pressure to be innovative, unique, and first to market becomes stronger (Pile, 2018); thus, it would be harder for large companies to keep pace with the rapid of a nascent company or startup velocity (Kohler, 2016). ICT industries have been facing a very turbulent environment, and they should be more adaptive and innovative. Otherwise, they will be getting substantial risks of decline (Wikhamn & Styhre, 2017). The advancement of digital economy in ICT industries has transformed the concept of growth crossover in nations and firms both involving input and output. It has been confronting an ambiguous between input increases and output decreases. Excessive increase in input could result in decreased productivity. To solve this dilemma, firms have to use the power of soft innovation resources that lead to neo open innovation (OI) concept in the digital economy (Tou et al., 2019).
There is a paradigm shift from closed innovation to OI applied for several companies' collaboration (Moschner & Herstatt, 2017), either between small and big companies or companies with research or other institutions. Further, nascent company innovation activities are more distributed, multidisciplinary, cross-border, cross-institutional, and inter-temporal processes (Kratzer et al., 2017).
In other words, the emergence of OI should be based on principles of integrated multidisciplinary collaboration, co-created shared value, cultivated innovation ecosystems, unleashed exponential technologies, and focused on innovation adoption (Curley, 2015). The other issue that originates from OI is crowdsourcing, which becomes a tool to integrate users into the innovation process.
Crowdsourcing is the practice of engaging a crowd for a specific task, and they work for solving a project with community-based co-creation. The problem is found that how to find the right crowd to minimize the uncertainty result (Koivisto, 2012). Co-creation with complementary partners that have particular skills and capabilities will be effective as the starting point of startup-industry collaboration (Aquilani et al., 2017

Open Innovation (OI)
OI has been known since the early 2000s; this model was utilized by various organizations to develop strategic partnerships and to create "win-win" scenarios. Globalization became the key driver of this new concept of innovation, it is the logical way to maintain a company's competitive advantage (Abulrub & Lee, 2012;Ozkan, 2015). If appropriately implemented, the OI model can generate better products and services (Pile, 2018). The OI term was first popularized by Chesbrough (2004). It was a theory that explains how organizations can be more efficient if they utilize external input to innovate.
OI has encouraged organizations to acquire intellectual property and values beyond their internal limitation. What makes OI distinguishable is the challenge faced to revamp the organizations' traditional R&D model that has a closed innovation system (Pile, 2018).
Due to the challenge mentioned above, OI has become a paradigm that forces companies to seek external ideas as much as internal ones and as a means to market development by adopting cutting-edge technology. With this definition, collaboration with external partners results in three OI strategies. The first one is the "outside-in" approach, which hons the company's knowledge based on innovation from external sources. The second is the "inside-out" approach, where an internal knowledge source exploits external knowledge. The last process is a merge of both the "outside-in" and "inside-out" approaches (Moschner & Herstatt, 2017). On one hand, by funding a startup company, the well-established industry will gain insights into new technological developments and emerging markets. On the other hand, startups will gain funding and access to administrative resources in exchange for the ideas they generated (Moschner & Herstatt, 2017 (Moretti & Biancardi, 2020). The relationship between OI practices and the firm's innovation performance is measured by some variables with inbound OI and outbound OI as independent variables and the firm's innovation performance as the dependent variable (Rangus & Drnovšek, 2013).
Other research also found how OI affected the firm and innovation performance of the firm. Reliable performance could be measured by sales growth, market share, profitability, financial indicators, customer performance, and turnover, while new products, R&D, intellectual property, and turnover are indicators of innovation performance (Lopes & de Carvalho, 2018  On the other hand, a startup will increase funding and access to administration resources in exchange for the ideas they generated (Moschner & Herstatt, 2017). Moreover, the turbulent environment in industries will be solved by introducing OI, which involves knowledge flow across the firm's boundaries as a complex co-creation process. Technological capability deals with a capacity that serves its technical function through the use of "state-of-the-art" technologies to benefit from external technology resources and to influence the process of external resource acquisition and exploitation. In contrast, market information management capability refers to the firm's ability to manage knowledge obtained from customers and competitors' activities.

OI and BM
Since it becomes necessary to produce, innovations are generally conceived in unrelated fields across the company's boundaries. Afterward, the company's management needs to be assertive in seeking external sources. One common way is to utilize emerging entrepreneurs from both formal education institutions such as universities and informal institutions such as incubators and co-working space (Kohler, 2016). A big company that collaborates with startups should firstly identify critical organizational practices to achieve a sustainability-oriented innovation as its growth driver (Kennedy et al., 2017). Nowadays, startups have started to initiate innovations to replace existing technology with a new BM. Therefore, the leading industry has to prepare its R&D division to attract startups by offering various kinds of BMs that are possible to generate disruptive innovations (de S. Fabrício Jr. et al., 2015). As a consequence, big firms need to reorganize their current BMs and organizational structures to succeed the collaborating with startups.
The R&D projects that do not fit with their current BMs might be commercialized elsewhere (Durst & Ståhle, 2013).
In a previous study, it was mentioned that the concept of a BM was not included in the definition of OI, although they were closely related. In OI, external and internal ideas are combined in a system that will be used in a BM. A BM, whether from an internal or external idea, generates values and defines an internal mechanism to determine the value itself (Vanhaverbeke & Chesbrough, 2014).
The new development of OI is mainly in the area of BM innovation and the area of shifting the BM from products to services. The hardest challenge is how to link the front end of OI to the back end of the BM that must bring these inputs to the market (Chesbrough, 2017). Furthermore, this model will raise the funds needed for research and shorten the product cycle. Hence, they will no longer be able to solely depend on internal knowledge. Most big companies are now also relying on external sources (Saebi & Foss, 2015). The hardest challenge is how to link the front end of OI to the back-end business that must bring these inputs to the market. Developing a BM innovation in an organization will underpin to identify the useful knowledge inflow for innovation and knowledge outflow or knowledge released to the outside (Chesbrough, 2017).

IBM-
The main idea of the OI model focuses on interactive processes through which knowledge and technologies flow across firm boundaries without great effort. The underlying assumption is that invention and innovation do not require the same place to develop. Internet use is ubiquitous, and it causes the ease of global knowledge and technology to create collaboration rapidly. The use of crowd-funding will reduce fixed costs for R&D due to its risk-sharing cooperation between partners (Inauen & Schenker-Wicki, 2012 (Kirschbaum, 2005). Furthermore, Chesbrough (2017)  properties (Rangus et al., 2016).
Another study revealed internal and external motives that affect larger companies in creating a CA program and how it was implemented in a new partnership with a startup firm. The research shows that the idea of creating a CA was mainly proposed by the company's CEO (Moschner & Herstatt, 2017). This is due to the fact that CEOs mostly understand the importance of OI as an instrument to gather external opportunities to ensure the company's long-term abilities to innovate. The role of dynamic innovation capabilities and OI activities should be a central focus in order to produce a breakthrough innovation, in this case, OI proposed to be a moderating variable (Cheng & Chen, 2013). CAs, as the rapid business incubation, moves startups from ideas to commercialization. By doing this, big firms actually drive the survival and growth of their business enterprises (Jackson & Richter, 2017).
CA's program is meant to support newly-established companies in forms of infrastructure, mentoring, training, and networking. It has been shown that the most effective model for a company to adopt startups technology through CA is by involving start-up-related parties in routine meetings and workshops for the duration of three months. According to Kohler (2016), there are four dimensions in designing a CA program as a link between a company and startups as follows: proposition (programs offered), process (how the program is run), people (parties involved), and place (where the program is located). According to another study, there are eight dimensions of CA's focus configuration, which are the following: (1) locus of opportunity (internal vs external); (2) strategic logic (exploration vs exploitation); (3) industry focus (tight vs broad); (4) equity involvement (yes vs no); (5) venture stage (early vs later); (6) external partner (yes vs no); (7) connection to parent (integrated vs independent); and (8)

Interpretive Structural Modelling (ISM)
ISM is a qualitative analysis tool for studying and analyzing the complicated relationship between interdependent variables to transform the entangled system into visible, well-defined models with graphical representation. The device is based on best practice and expert knowledge, examining each element pair to identify their direct relationships recorded in an interaction matrix.
Thus, ISM modelers could map the hierarchy structure for an entangled system efficiently and effectively through matrix transformation and decomposition (Li et al., 2019). In other words, ISM is one of the effective methodologies for determining relationships among specific factors showing a problem overview.
Researchers have been using this approach to find relationships among variables referred to as the problem. There are some advantages to using this methodology, such as the following: (1)  Generally, ISM starts with variable identification related to the problem issues. The second stage of the ISM approach is building a structural self-interaction matrix (SSIM) from a pair-wise comparison of variables, building a contextual relation such as "leads to," "depends on," "increases," "decreases," etc. Four symbols are used to define the relation between two sub-factors (i and j). The V symbol indicates that factor i leads to factor j, the A symbol that factor j leads to factor i, the X symbol indicates that factor i leads to factor j and factor j leads to factor i, and O means there is no relationship between factor i and factor j. The third stage is converting SSIM into a reachability matrix The autonomous quadrant consists of components with weak dependence and weak driver power.
The property of those components is relatively less connected to the system. The other clusters consist of components with low driver power and high dependence, high driver power, and high dependence, high driver power, and low dependence (Jain & Banwet, 2013). Components with high driver power and high dependence are called "linkages." These factors are unstable, and any action on them creates a significant impact on the goal development process. They also have feedback effects on themselves (Chidambaranathan et al., 2009). ISM is mostly used at a high level of abstraction either for long-range planning or a more concrete level (Figure 1). It has become a problem solver for process design, career planning, strategic planning, engineering problems, financial decision making, human resources, competitive analysis, and electronic commerce (Chidambaranathan et al., 2009).

Hypotheses of Start-up and Industry Collaboration
Since large corporations in developing countries, such as Indonesia, cannot afford high internal R&D expenses and also learn from global trends, they started to establish startup ecosystem fostering their business growth. BM and OI strategy from both startup and corporation will be the key success of constructing a CA. The OI type, whether inbound or outbound, would vary due to the maturity level of the existing internal R&D of the corporation (Inauen & Schenker-Wicki, 2012;Rangus & Drnovšek, 2013).
A previous study also found that value proposition is the main component of BM to pivot the strategic partnership between startup and corporation (Vanhaverbeke & Chesbrough, 2014). It was also found that innovation performance could be measured from general indicators such as profitability, growth, market share, and sales (Moretti & Biancardi, 2020). The hypotheses are as follows: H1: OI strategy will be suitable for startupcorporation partnership in ICT industries

Data Collection
The three experts are interviewed one by one to verify the variables from the literature study, and the SSIM matrix is cross-checked. If there is a difference in any sub-factor relation, then the modus will be the final SSIM matrix. The experts have at least 15 years of experience, and the companies' profiles are described in Table 2.
The following is the SSIM matrix (see Table 3     The second stage is transforming the SSIM matrix into the initial RM (see Table 4).  Table 5.
The factors for which the reachability and intersection sets are the same occupy the top-level position. Then, the same step is repeated to obtain the factor for the next level. The step is repeated until the level of each factor is obtained, as seen in the last column of Table 5. The final stage is drawing an ISM digraph model based on the factor level from Table 5 Figure 4.

RESULTS AND DISCUSSION
The driving power and dependence power matrix in Figure 3 shows The ISM result supports a previous study. West and Bogers (2014) suggested a co-creation between firms and external partners as communities and value networks for sourcing external innovation.
Co-creation is a tool to expand the firm's innovation and value creation. An example of an OI model that has been successfully implemented is the Apple company with its signature product, iPod.
By utilizing the structure of the OI model, Apple manages to focus on its' dynamic capability by  where OI will lead to CA as the startup-industry collaboration. The culture and integrity factors will be necessary for both to succeed, for they generate mutual trust and benefit.
The two last factors, reducing R&D cost and the main driver of the company's sustainability growth, are the dependent variables or outcomes of the collaboration between startups and industries.
These factors have a strong dependence and less driving power; thus, they lead to the collaboration performance measured by profitability, the company's growth, market share, and total sales (H3). The OI is part of BM from its co-creation process, and BM indicates the collaboration performance in revenue streams. The conceptual model in Figure 2 could be simplified into the relationship between independent and dependent variables in the context-based CA, as described in Figure 5.
The result in Figure 5 is supported by previous studies. Companies that adopt OI must manage both technical and market uncertainties through a co-creation journey so that they can achieve the successful commercialization of new technology.
Furthermore, technological progress and innovation are believed to be business success factors in developed economies, and innovation is an enabler to obtaining sustained growth for many companies (Sivam et al., 2019;Teplov et al., 2017). Most problems are coming from the poor capability and understanding of the latest technology adoption, such as how technology might be applied by customers and benefits from the customer's point of view. As a result, both false positives and false negatives of measurement errors are unavoidable (Chesbrough, 2004). Therefore, OI could be the best strategy choice in which both companies will share IP and risk to shorten the time to market and reduce the overall innovation cost.   (Oltra et al., 2018).
A BM and OI-based strategy will ally the two forces of startups and large corporations, as shown by Figure 5. Hence, large corporations become a kind of "business catalysts" and keep looking for more initiatives to improve the innovation system. Vanhaverbeke (2013) found that OI must be embedded in the corporate strategy to understand the real value of OI initiatives. One of OI implementations could be defined as selling or buying IP of other companies to succeed in new products or services development (Rangus et al., 2016). The active collaboration between different firms includes IP sharing, while a cocreation process will create values by collecting all stakeholder's contributions (Pile, 2018).

MANAGERIAL IMPLICATION
The previous study agreed that co-creation is an independent factor driving OI as a good firm strategy. The notion of co-creation with complementary partners arises in the ICT sectors.
This idea stresses diverse types of cooperation with external stakeholders from various industries, each with its own set of skills and competencies.
Co-creation means that organizations build their knowledge bases through external ideas, knowledge, and resources. It highlights the critical role of interconnected innovation networks, the mode of customer integration (e.g., crowdsourcing), and the intermediation activity of third parties that facilitate and support interactions and collaboration between heterogeneous participants.
However, co-creation is not easy since human resource policies must be integrated into OI processes. It needs those huge organizations to update their capabilities for collecting information, expertise and innovative ideas that are not part of their core competencies. Companies in the same industry that receive the same resources may not combine them in the same way, so adding to the uniqueness of the combination. The department of human resources is essential to organizational intelligence operations because it prepares all employees by incorporating external information into the organizational learning capabilities.
Innovation culture and innovation openness are two essential factors that impact the co-creation process. The most challenging aspect of building and cultivating an innovation culture is retraining the organization's mindset to mobilize teams to swiftly provide breakthrough products and services to the market. Shortly, ICT firms must adopt a new culture that fosters an open innovation mentality.

CONCLUSION
Since the ICT industr y has been in a global turbulence, the company's innovation could not stand alone. CA is one of the collaboration types, which relates to the startup and its leading industry.
Collaborating with startups is more beneficial for big firms or incumbents. The most important benefit for big firms is the flexibility and openness of startups to generate new opportunities for disruptive innovations. Big firms will also obtain gap resources and assets cheaply to keep their innovation engine running. CA benefits for startups are solving a lack of resources, lack of legitimacy within the market, funding constraints, and getting a more competitive business environment.
Data obtained from three Indonesia ICT companies support all hypotheses, which conclude that the OI strategy is suitable for startup-industry collaboration with proper alignment in BM. The OI strategy also generates a positive relationship for both companies. The hypotheses, linked to previous studies, have been transformed into the following five factors: co-creation, IP sharing, reducing R&D cost, risk sharing, and the main driver of the company's sustainable growth. All five factors are judged by three experts using ISM method.
The result shows that "co-creation" represented by BM becomes the most reliable driver in the issue of the importance of collaboration, and it leads to an independent factor that should be firstly developed and continuously improved. The IP-sharing and risk-sharing factors are both in the linkage quadrant and become sensitive factors. Therefore, IP management should be the key issue to succeed in the OI collaborations due to the existence of information asymmetry, defined as hidden information and hidden characteristics between companies. Eventually, startups and industries must have a proper contractual agreement once they decide to share the IP and risk either with good corporate governance or an adequate legal understanding for both sides. The integrity factor will be necessary for both to generate mutual benefit.
Reducing R&D costs and the main driver of the company's sustainability growth are the dependent variables or outcomes of the collaboration between startup and leading industries. Both factors show the positive relationship between OI strategy and the company's performance.

Limitation And Future Research
The data comes from three big ICT companies and three top managements, which might not represent the industry cluster accurately. However, one of the research objects is the biggest ICT company, which is a state-owned enterprise. Future research should involve startups to get their perspectives on how an appropriate co-creation can be reached optimally.
It might also develop a dynamic model to simulate the firm's performance related to OI activities.
It might be a linkage between OI and dynamic capabilities. The next research should broaden the model involving those capabilities. The effectiveness of OI activity depends on the firm's capacity, especially the dynamic ability, which is the dynamic resource-based view of the firm, exploring the synthesis of the RBV, and the dynamic capability in creating a competitive advantage. Therefore, it should be a proposed model describing the linkage between different types of dynamic capabilities and various forms of innovation, leading to value creation and industry leadership (Parthasarathy et al., 2011).
From the dynamic capability perspective, implementing OI will require a sensing tool development to outsource the appropriate technology, and a seizing tool to integrate all skills (Bogers et al., 2019). Realigning the organization to incorporate external knowledge will be part of the dynamic capability, together with developing a collaborative culture and adjusting the mix of internally and externally developed technologies.
The ISM is a qualitative tool with a basic binary concept. The main limitation of ISM is the relationship among variables that depends upon the expert's knowledge and experience. For further research, this tool could be enriched with machine learning, such as open mining or sentiment analysis, to obtain more accurate and better results with more extensive data, either secondary or primary data.