Leveraging Social Media in New Product Development: Organisational Learning Processes, Mechanisms and Evidence from China

Purpose – The main purpose of this paper is to investigate how social media can provide important platforms to facilitate organisational learning and innovation in new product development (NPD) process. Design/methodology/approach – Using a multiple case-study approach, this study assesses qualitative data collected via 56 interviews from 13 world-leading Chinese companies in the high-technology industry. Findings – The study identified three distinct types of organisational learning mechanisms for firms to extract potential innovation inherent in social media. It further determined various organisational enablers that facilitate the connections between these mechanisms and NPD


Introduction
Successful new product development (NPD) requires effective exploration and exploitation of different kinds of knowledge from various sources (Stevens and Dimitriadis, 2004;Nguyen et al., 2015). However, for many companies, this process is not straightforward, given the considerable levels of complexity with which they are confronted, both in terms of technical feasibility and market competitiveness (Roberts and Candi, 2014;Zhu et al., 2017;Ramanathan et al., 2017). This is especially true in NPD, where customers increasingly demand speed delivery, better quality and lower price (Mount and Martinez, 2014;Neeley and Leonardi, 2018). As a consequence, the understanding of how companies can effectively manage knowledge from different sources and develop new products to meet market needs is essential.
Use of social media to identify, collaborate and interact with external stakeholders has gained popularity over the past decades (Culnan et al., 2010;Lam et al., 2016;Neeley and Leonardi, 2018). According to Barnes and Pavao (2017), more than 98% of Fortune 500 companies have implemented a range of social media applications (e.g. LinkedIn, Twitter, Facebook and YouTube) for internal purposes or to interact with customers, suppliers or partners. The improved media richness and outreach provided by these applications support many-to-many interactions and contribute to effective NPD (Sethi et al., 2003;Mount and Martinez, 2014;Huang et al., 2017). Nevertheless, while there is a general understanding of the relationship between social media applications and NPD performance (Roberts and Candi, 2014;Roberts et al., 2016), current research offers little insight into how these processes actually work and how they can be organised effectively throughout the entire NPD process (Sethi et al., 2003;Franke and Piller, 2011;Roberts and Candi, 2014;Felix et al., 2017).
Organisational learning theory suggests that different approaches to knowledge management exist (Cook and Brown, 1999;Neeley and Leonardi, 2018), but these have not been studied in the context of NPD. NPD can be seen as an organisational learning process to exploit and acquire business opportunities and new knowledge (Atuahene-Gima and Murray, 2007); thus, knowledge is a source of sustainable competitive advantage (Nguyen et al., 2015). Bharati et al. (2015) point out that the use of social media for better knowledge management can significantly improve product effectiveness, and He et al. (2017) encourage studies to pay particular attention to the micro-levels of knowledge management by integrating such processes into social media initiatives. Therefore, it is imperative that the use of social media 3 be regarded as an organisational learning and creativity process for knowledge management (Kane and Alavi, 2007). This article, therefore, seeks to investigate how companies develop new products through organisational learning and knowledge management. More specifically, it intends to study how social media can facilitate innovation throughout the entire NPD process.
This study investigates the research question using a case-study approach. The findings offer qualitative evidence of usable mechanisms to leverage social media in enhancing NPD performance from 13 world-leading Chinese companies. Moreover, the findings also reveal a previously unidentified learning process, entailing the cost-effective and time-efficient pursuit of a cyclic process of organisational learning through exploration and exploitation. The findings show that while social media is initially instrumental in facilitating information search and knowledge acquisition, the success of product ideation still heavily depends on firms' capacity to actively seek distinctive or qualitative knowledge. Additionally, the ability to acquire and identify external resources and knowledge through social media is crucial for the innovativeness and success of product R&D. Likewise, social media can act as important platforms for reflective learning and knowledge transformation during the product testing and launch process.
This research extends organisational learning and knowledge management theories by demonstrating that social media not only intensifies firms' volume of information and knowledge, but also helps to leverage different forms of learning activities throughout the NPD process. Particularly, the study responds to recent calls for a more integrated perspective on the use of social media in innovation and knowledge management (Lam et al., 2016;Ramanathan et al., 2017;Cheng et al., 2018). It suggests that NPD can be organised in a resource-effective way through various social media mechanisms and knowledge learning strategies. Besides, the identification of how social media acts as the leverage platform for NPD adds to the organisational learning and knowledge management literature. This study is also one of the first to illustrate how organisational learning and knowledge management can be translated into innovation mechanisms via social media. As this approach combines exploration with efficient exploitation of existing knowledge, it also reconciles the apparent juxtaposition between 'knowledge searching' and 'knowledge in-action'-an important transformation from product ideation to successful commercialisation (Daft et al., 1988;Nag and Gioia, 2012).
While social media has traditionally been conceived as implying the random collection of  (Franke and Piller, 2004;Sigala, 2012;Mount and Martinez, 2014), these findings reveal how social media provides the capabilities for enhancing learning through the three mechanisms of NPD: dynamic searching for product ideation, collective learning for product R&D and knowledge in-action for product testing and launch.

Theoretical Background
To study the use of social media in NPD, this paper relies on two main literature streams: social media in NPD and organisational learning. The literature on the use of social media in NPD formed the initial basis for this study, while the relevance of organisational learning processes for knowledge management emerged from the case-study analysis. For clarity, both are discussed here.
The use of social media for product ideation and design. In the literature on product development, the most-recognised benefit of using social media is that of providing value and knowledge to organisations (Aral and Walker, 2011;Cooper, 2016;Neeley and Leonardi, 2018). In fact, the use of social media for product ideation and design is relatively well developed in the NPD literature (Lüthje and Herstatt, 2004;Mount and Martinez, 2014). While many studies have noted that social media plays a valuable role in product ideation and design to develop innovative products, others have criticised using social media to involve customers in product ideation and design, arguing that this can result in imitative and unimaginative products (Roberts et al., 2016). It is now recognised that the use of social media differs with the levels of technology readiness and alignment of product development strategies to current customer requirements (Culnan et al., 2010;Chiu et al., 2012;Roberts et al., 2016). Research has demonstrated that successful companies tend to implement different social media applications for their product ideation and design (Franke and Piller, 2004;Aral and Walker, 2011;Zhu et al., 2017). For example, Lenovo Social Champions and Starbucks Ideas are  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   o  n  a  l  J  o  u  r  n  a  l  o  f  O  p  e  r  a  t  i  o  n  s  a  n  d  P  r  o  d  u  c  t  i  o  n  M  a  n  a  g  e   6   popular social media platforms on which users can not only share their comments, but also create new content, with each other or the company's NPD teams.
The use of social media for product R&D. Social media plays an essential role in product R&D.
It supports activities such as the prioritisation and design of product prototypes, the validation of product features, the establishment of development-process metrics and the identification of product-interface requirements (Sigala, 2012;Piller et al., 2011;Cooper, 2016). This use of social media is consistent with open forms of NPD (West et al., 2014), virtual communities and online platforms (Du et al., 2016). Also, it is highly pertinent in the era of 'big data', in which emphasis is placed on sharing and extracting knowledge and perceptions from external stakeholders (Manyika et al., 2013;Zhan et al., 2016). By engaging in such interaction, the company's knowledge is enhanced, and the information generated may be embedded in product R&D (Hoyer et al., 2010;Roberts et al., 2016). This can lead to new products that are more valued by consumers and closer to their requirements than products created by companies in isolation (Kim and Wilemon, 2002;Cooper, 2016). For example, Polyvore is a communitypowered social commerce platform owned by SSENSE 1 that focuses on collaborative NPD with social media communities. The public send ideas to Polyvore, who then creates prototypes, which are subsequently voted on by the members of different social media communities. This process allows the rapid development of new products that have strong market attractiveness.
The use of social media for product testing and launch. As product testing and launch is the final and very costly phase of the process of NPD (Roberts and Candi, 2014), companies that are proficient in establishing social media strategies can affect consumer perceptions and therefore increase the likelihood of new product acceptance and successful launch (Du et al., 2016;Roberts et al., 2016). Research has identified the remarkably productive use of social media for product testing and launch (Ramanathan et al., 2017;Tan and Zhan, 2017). Roberts et al. (2016) note that by involving a variety of consumers through social media in product testing, companies can acquire in-depth understanding of how their new product fits with different types of consumers. In addition, using social media in product testing allows companies to identify product defects early, and reduce costly rework and redesign (Felix et al., 2017). For example, many companies in the high-tech industry (e.g., IBM, Tencent and 1 SSENSE is a pioneering international fashion platform. It integrates its programme with a variety of social media channels and utilises social media data and reporting capabilities to produce a more data-driven product.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   o  n  a  l  J  o  u  r  n  a  l  o  f  O  p  e  r  a  t  i  o  n  s  a  n  d  P  r  o  d  u  c  t  i  o  n  M  a  n  a  g  e   7 Lenovo) have applied social media to connect with its end-users in their beta product testing, a process that allows companies to minimise investment in internal product testing procedures (Chiu et al., 2012). During a product's launch, social media can also offer practical and innovative ways of interacting with a wide range of consumers, facilitating communication in a way that traditional methods cannot provide and allowing companies to become close to their target markets (Hagel and Brown, 2011;Cooper, 2016).

Social Media in Organisational Learning
In general, NPD is one part of the organisational learning process through which companies try to solve problems and achieve competitive advantages (Stevens and Dimitriadis, 2004). Rojo et al. (2018) point out that organisational learning is a dynamic process of generating new knowledge and applying to where it is required, leading to the generation of new knowledge for later use and integration. Therefore, social media can be regarded as high levels of mediarich modalities for collaboration, offering companies a powerful means of knowledge acquisition and integration that can be implemented in organisational learning (Nguyen et al., 2015). According to Hemsley and Mason (2013), managers leverage social media to search for new knowledge, and synthesise multiple sources of knowledge into the body of existing knowledge through integration, which then guides action. This is in line with Daft et al.'s (1988) arguments that companies differ systematically from each other in their patterns of searching, integration and action, pointing to the need to develop an in-depth understanding of relationships among knowledge acquisition, integration and action.
Social media for knowledge acquisition. A related but distinct stream of studies stresses the significance of social media as a key knowledge search/acquisition behaviour in NPD (Koskinen and Vanharanta, 2002;Robert et al., 2008;Hemsley and Mason, 2013). This paper defines knowledge acquisition from social media as the capability to search for critical and insightful knowledge emerging from social media necessary to a company's NPD projects (Robert et al., 2008). Studies illustrate that managers of successful companies apply social media more widely for knowledge acquisition in NPD (Chiu et al., 2012;Barnes and Pavao, 2017), and these managers possess 'information flexibility', adopting different social media initiatives given the perceived uncertainty of external environments (Roberts and Candi, 2014;Zhu et al., 2017). Leonardi (2015) argues the need to expand the conception of social media beyond the narrow view of formalised information-gathering activities. Thus, it is critical to Social media for knowledge integration. As a vital part of knowledge management, knowledge integration refers to the process of synthesising multiple specialised streams of knowledge into situation-specific systemic knowledge (Kane and Alavi, 2007). Robert et al. (2008) point out that the essential source of competitive advantage comes from knowledge integration rather than knowledge itself. According to McKinsey (2012), companies are incorporating social media inside their operations to enhance communication and R&D collaboration. While research has offered exploratory insights into social media's impact on knowledge management (Leonardi, 2015;Hemsley and Mason, 2013), the influence of social media on knowledge integration has been overlooked. Most importantly, studies emphasise the significance of investigating how social media can be leveraged for knowledge integration in NPD.
Social media for knowledge-in-action. So far, empirical study on how social media generates competitive advantages have mainly emphasised the quality and quantity of knowledge created by companies (Koskinen and Vanharanta, 2002;Bharati et al., 2015;He et al., 2017). This 'possession' perspective of knowledge has been further enhanced by the perception that knowledge is an organisational asset in use rather than something located in dispersed units in a company (Nag et al., 2007). Based on early studies, researchers have identified a practical perspective on knowledge wherein it is regarded as embedded and integrated into the practice of a company's daily work (Pentland, 1992;Nag et al., 2007;Nag and Gioia, 2012). In this way, social media can be used as a kind of knowledge application to enable companies to communicate information with specific experts or broadcast information to the entire company.
Accordingly, this research builds on calls for more empirical case studies that focus not only on the body of knowledge that a company acquires, integrates and transfers, but on methods via which companies can dynamically conduct to support innovation (Cook and Brown, 1999).
In essence, the literature review provided insights into the use of social media in NPD and organisational learning. These are separate streams of research, and this is perhaps the main reason why, to date, no study has provided holistic insights into how managers can leverage social media for organisational learning as well as facilitating innovation throughout the entire NPD process. These streams provide a starting point for the data collection and analysis of this study.

The Context of the Study
Chinese companies were chosen as the research population, for three reasons. Firstly, according to Internet World Stats (2018), China has the world's largest internet user base-772 million people, more than double the 312 million users in the US-and the world's most dynamic environment for social media. Additionally, China's online users spend over 40% of their time on social media, and this percentage continues to increase rapidly (Chiu et al., 2012). As a result, Chinese companies are operating in an increasingly demanding consumer market environment that is catalysing the use of social media in product development (Woetzel et al., 2015), as the country attempts to meet its 'innovation imperative'. Secondly, several Chinese companies (e.g., Xiaomi, and Tencent) have been aggressively experimenting with social media intensive, novel innovation models that have demonstrably facilitated and achieved great benefits in their NPD activities (Williamson and Yin, 2014). In fact, the country's social media activities in this area have been so impressive that McKinsey & Company (Woetzel et al., 2015), the global consulting firm, has specifically called for other countries to take note of and learn from the Chinese model of social product development. Thirdly, most research into NPD has focused on Western economies and companies (Roberts and Candi, 2014). Because of the size and rapid growth rate of its economy, China has emerged as the world's largest social media market, and this market differs from those in Western countries (Chiu et al., 2012). For example, the social media sector in China is very local and fragmented (e.g. there is no Facebook, Twitter or YouTube). Each social media platform has its own local players, for example, Sina Blog (a Chinese version of Twitter), Tencent WeChat (the Chinese WhatsApp), Youku (the Chinese equivalent of YouTube), but with more users and improved features. Therefore, the specific nuances of how social media intensive NPD occurs within the Chinese context are extremely relevant on both a practical and theoretical level for those attempting to engage Chinese consumers. Nonetheless, examination of these nuances has been largely overlooked in the literature. Thus, this study helps to fill this knowledge gap.

Research Design and Data Sources
This research identified the case-study approach as the most suitable methodology because of the research objective and the context-specific nature of the phenomenon being explored  lasting several years or less (Brexendorf et al., 2015), were targeted first. Most companies in fast-cycle industries are high-tech or electronics companies, and important drivers of innovation and economic growth (Woetzel et al., 2015). It was expected that these companies would be more socially active and proficient at using social media in NPD. The second criterion was the presence of social media engagement and applications to facilitate NPD. The third criterion was evidence of projects focused on different NPD phases (because this research examines the entire NPD process). In particular, companies' archival information and reports, including business reports and news articles, were evaluated to determine whether they were committed to and proactive in social media and NPD.
Starting from July 2017, we adopted a retrospective approach to get familiar with the companies identified (Miller et al., 1997). Pilot tests were conducted with senior managers from the Chinese offices and headquarters of these companies to evaluate their eligibility based on the sampling method. During this process, eight companies were excluded, because either their social media initiatives were too limited (e.g., mainly focused on marketing) or they were unable to grant access to confidential information. Ultimately, 13 companies were identified to participate in this study. All companies extensively used social media to re-engineer their R&D and NPD, but differed in their NPD directions; namely, product ideation and design, R&D, and product testing and launch.

Data Collection and Analysis
It is worth noting that all 13 chosen companies are very large corporations operating globally or nationally. The research studied specific innovation projects within the case companies.
According to Tidd and Bessant (2018), conducting innovation research at the project level allows researchers to reflect on their own experience during the whole innovation process and become a part of the project. The study's semi-structured interview protocol began with general questions and then focused particularly on the mechanisms of social media use in NPD and expertise in developing these mechanisms for competitive advantage. Data was collected between March 2018 and September 2019. In total, 56 in-depth semi-structured interviews were conducted with executives from 13 case companies. Almost all interviewees were senior managers performing important roles in the NPD activities, but other employees identified by senior managers as playing key roles in successfully using social media in NPD were also interviewed. Table 2 summarises the data collection methodology, including visiting operations sites of all companies, participating in internal workshops and meetings, gathering company newsletters and annual reports, and observing the teams using social media in NPD.
According to Eisenhardt (1989), onsite observations can offer insights into everyday working activities because they shift the focus to the direct study of contextualised actions. Meanwhile, archival data were used primarily to generate a deeper understanding of the strategic impacts of social media on NPD and to verify empirical findings to provide strong support for the results developed. Following Miles and Huberman's (1984) suggestions for analysing data from multiple cases, data analysis was conducted in two steps. First, each case was analysed in detail using withincase analysis. Adopting the theme analysis and categorisation techniques, each interview was coded separately according to categories derived from the interviewees' quotations (i.e. firstorder categories). Second, the cases were compared through cross-site analysis to identify emergent categories from different cases. Specifically, each interview was checked many times to identify differences and similarities among interviewees. The constant-comparison method was adopted across multiple interviewees to determine concept patterns (Strauss and Corbin, 1994). Similar codes were identified and allocated to the same first-order categories. To reach theoretical saturation (Strauss and Corbin, 1994), interviewing and data analysis was continued until no further distinct codes or shared patterns among interviewees emerged. Concurrently with the establishment of the first-order categories, relationships among the categories that could result in the formulation of second-order themes-theoretically distinct concepts induced by the researcher and developed at a more abstract level-were identified. The formulated second-order themes were then integrated into aggregate mechanisms, which led to developing a theoretical framework that integrated the different ideas captured from the data.

Findings and Propositions: Insights from NPD projects illustrating three different types of mechanisms
The analysis indicates that successful Chinese companies operating under similar exogenous conditions display distinct variations in how they used social media in NPD. The emergent theoretical model comprises various mechanisms under the entire process of the three core NPD phases: dynamic scanning for ideation, collective learning for R&D and a fast-launchand-improve cycle for testing and launch. Figure 1 presents the data structure, including the first-order categories (categories derived from the interviewees' quotations) and second-order themes (theoretically distinct concepts induced by the researcher) that led to the development of the aggregate mechanisms underlying the three NPD phases. These identified data structures represent the content of senior managers' interpretive frameworks on how social media can be leveraged in NPD. A summary of cross-case synthesis is shown in Appendices A, B and C.
According to Eisenhardt (1989) and Nag et al. (2007), a developing theoretical framework based on grounded theory approaches such as constant comparison and data reduction must specify not only constituent concepts, but also connections among the constituent concepts to explain or describe a phenomenon. Therefore, following the approaches of systematic combining, this paper identifies which aspects of social media usage pertain to a specific context and can be generalised theoretically (Hullova et al., 2019). These are displayed in the form of six propositions, which lead to directions for future study.

Dynamic Scanning for Ideation
The findings indicate a distinct ability of the cases to generate excellent product ideas and designs addressing seemingly untapped areas of the existing market. In particular, social media was associated with differences in both the character and amount of searching focused on acquiring knowledge in product ideation. Consistent with existing literature, the analysis confirmed the prevalence and significance of the 'intensity' with which searching was implemented (Tóth et al., 2019). A second mode of searching also appeared from the analysisthe extent to which managers apply social media resourcefully and proactively search for information. Thus, the findings expand the basic searching concept into two distinct forms: searching intensity and searching proactiveness.  These findings suggest that managers have different searching modes to leverage social media for knowledge acquisition in the product ideation phase. However, they need to apply appropriate searching modes in line with the heterogeneous nature of NPD projects. The cases show that NPD projects with a high degree of social media searching activities tend to adopt a customer-focused organisational culture. According to Cooper (2016), customer focus refers to a business philosophy that positions customer value and opportunities at the centre of every company decision. Social media introduced novel customer-focused applications that enable companies to interact with customers and other stakeholders during the NPD process.
Typically, it is highlighted as a critical marketing strategy. However, the findings indicate that a customer-focused culture is perceived by operations, product R&D and data analytics as essential to their primary roles. As stated by the technical director of Case H, 'we implemented a customer-focused culture within our department. We have strived to integrate our social media searching activities to develop a unified infrastructure permeating all networks, products and businesses'. Besides, integrating a customer-focused culture to the searching activities views the customer as the unit of analysis. It can help to identify the appropriate questions to ask and support companies to apply relevant cases, which transform sophisticated social media data into exploratory behaviours, prescriptive models, and fact-based decisions.
This leads to the following proposition: Proposition 1. The degree of social media searching intensity for product ideation is higher when the company has a higher level of diffused customer-focused culture.  The findings indicate that information acquisition cannot be adequately conducted by analysing searching intensity only. The proactiveness of searching was another distinctive form of leveraging social media for product ideation. Notably, the cases reveal that a data-driven culture plays an essential role in improving the effectiveness of social media search proactiveness for product ideation. According to Davenport (2012), a data-driven culture possesses data as a critical resource for generating decisions and insights in every department of the company. This cultural suggests that an experimental process relying on hypothesis testing and data integration to generate new insights can enhance social media searching performance and overall performance of the project. For example, the vice general manager of Case A commented: 'people can always find an explanation for anything but without knowing whether it is right or not. I would like to cultivate a cultural shift across the company through encouraging the conviction that decisions should be made by data analysis, while depending less on experience and intuition'. More specifically, when managers considered that their companies have a data-driven culture, they will be more frequent and motivated to use social media in searching for knowledge activities. Also, these managers tend to dedicate more effort to ensure that they stayed on the cutting edge of knowledge development in their areas. Thus, we propose: Proposition 2. The degree of social media searching proactiveness for product ideation is higher when the company has a higher level of diffused data-driven culture.

Collective Learning for R&D
The analysis identified an important ability of the case companies to effectively leverage social media to identify emerging trends and generate new insights, which were sent back to internal production processes. This mechanism is termed here collective learning for R&D, and refers to the knowledge synergies that emerge from collaborative NPD on social media. This is consistent with the theory of open innovation, which states that access to a wide range of capabilities, knowledge and skills enables individuals to integrate different solutions into coherent and innovative ideas (West et al., 2014).
The first subtheme is external accessibility, representing whether company knowledge from external environments can be easily obtained. In this regard, many companies (e.g. cases B, D and G) had adopted external digital agencies to deal with their social media data and knowledge integration process. For instance, a critical issue encountered by Company D was an internal constraint on harvesting values from the sheer volume of user-generated content on different social media platforms. The adoption of crowdsourcing intermediaries to manage the social space was therefore essential to ensure that emerging trends and new product insights were formalised and integrated to facilitate R&D processes. Additionally, external accessibility was supported through community development, which is one of the approaches used to enable user co-development, and refers to the process through which companies leverage customer interests and desires by creating communities of like-minded individuals. Particularly, all the cases were found to cultivate a diverse range of officially recognised social media platforms and online communities for customers with different interests. Content on social media is in real time, with both consumers and employees of the companies interacting and contributing to product R&D. When asked about the development of the company's social media community, the CIO of Case G stated the following: 'The key to cultivating a successful social media community is being active. We need to post updates and news frequently to give customers a sense of community, so they feel they are an important part of the NPD team.' The findings show that external information accessibility is enhanced by significant support from senior management. This is always reflected in a powerful manager who champions the implementation of social media-enabled NPD projects and overcomes challenges that might generate in the company (Troilo et al., 2017). The manager is always at the executive level who is keen to implement the project based on the previous successful experience of social media projects, or has distinct knowledge about novel social media opportunities. In our study, most Compared to the investments on the integration with external digital agencies, some cases received support from senior management with the development of new functional teams focused on establishing commitment toward social media-enabled innovation. For instance, Case E developed a "social media innovation group", which communicates to the CIO but has no direct connections with other departments. The group is responsible for cultivating social media communities, incorporating customers' knowledge to identify novel innovation opportunities, and communicating them to relevant R&D teams. It is also responsible for disseminating the latest progress to the rest of the company by weekly reports, intending to improve internal communication of social media-enabled NPD projects. The previous head of the R&D department noted: 'More than just socialising, it is vital to transform social media into useful information, representing customers, suppliers, analysts and partners towards social media -enabled innovation.' This leads to the following proposition: Proposition 3. The external accessibility of social media for product R&D is higher when the level of social media-enabled innovation is higher.
Project novelty refers to the degree of a project's importance and innovativeness, as perceived by the project managers, in terms of its influences on the success of the company. Among the cases, managers differed in their ascriptions of the knowledge domain critical to their companies. Some reported that achieving product novelty was important to success. Others also appreciated innovativeness but suggested it was more essential to achieve operational process control and product cost efficiency. For instance, the head of innovation of Company B highlighted the significance of project novelty: 'We are receiving an increasing number of design and engineering specifications from social media. Thus, our product engineers are developing some very novel functional features to satisfy our customers. That's the reason why it is critical that we work on the cutting-edge of technology.' This finding suggested that it was the utilisation of social media activities that led to the exploration and development of new product lines. Nevertheless, the senior IT project manager of nearby Company G argued for the importance of efficiency as an approach to controlling costs and uncertainties. The case analysis demonstrates that Companies G and K are using social media through a 'panentertainment strategy'-the development of multilevel products on the basis of a company's existing intellectual property (e.g. animation, information, customers, patents, fiction and Although highly novel projects are critical to business success, they involve a high level of uncertainty and thus need an exploration-focused approach (e.g. radical innovation). With a low level of novelty, the NPD teams act mostly with their existing knowledge acquired from years of experience in attempting to improve the efficiency of their production and operations processes (e.g. incremental innovation). The findings suggest that customer education becomes an effective way of improving the impact of both high and low novelty projects on the use of social media for product R&D. Customer education refers to an organisation's role to provide their customers with the skills, abilities and knowledge required to use their products in a more informed way (Lee et al.,203). The interviews show that most of the case companies announced customer education-oriented messages regularly to make customers more familiar and comfortable with their new products. As in this statement from the head of IT department of Case D: 'In today's big data era, customers are becoming increasingly aware of the power of information and their role in product R&D. Thus, it requires companies to be able to convey their knowledge back to consumers in a manner that they can directly recognise and obtain values from'. The findings imply that customer education shifts from as an individual level capability to a firm level capability that can strengthen the impact of social media in product R&D, in terms of improved project novelty, cost-effectiveness and customer experience. Thus, we propose: Proposition 4. The relationship between project novelty and the success of social media-enabled product R&D is moderated by the level of customer education implemented by the company.

Fast-launch-and-improve Cycle for Testing and Launch
The data analysis identified the fast-launch-and-improve cycle as another essential mechanism, which illustrates how social media was actually applied to create competitive advantage. This

Employing what you learnt is what makes you successful. The use of social media as a testand-learn toolkit is what really causes the difference. Many people don't know this, and it could
be the key to the whole business.' Here, 'knowledge-in-action' is used as the label for this pattern as the case companies consistently mentioned the terms 'intelligent' and 'appropriate' application of social media in daily operations. Two distinct types of knowledge-in-action were identified in the findings: information processing and information transformation.
Information processing concerns actions that confer a capability to leverage social media to support specific operational activities and offer digital services in NPD. The focus is on developing digital solutions to address local challenges through creativity and ingenuity. In some companies, this pattern of knowledge-in-action was regarded as the capability to 'tweak' social media. For example, Companies C, F and J established product improvement units within their IT departments to process valuable inputs from different sources and convey the information to relevant NPD teams quickly. The unit serves the role of business analyst, with one of its fundamental objectives being to process useful knowledge from social media and provide valuable feedback as inputs to facilitate the process of product modification and launch. This reactive improvement approach allows R&D teams to relaunch the product quickly and trigger further innovation. In addition, comparison of the interviews revealed a distinct approach showing that these companies follow a capabilities-driven strategy. That is, rather than developing a 'fully-fledged' product before launch, they centralise the new product's core capabilities and routinely launch new products quickly, and then pay close attention to process the feedback that they receive from partners and customers. This pattern is exemplified by the following comment from the VP technology of Case F: 'It [social media for information processing] has proven to be especially effective in market segments that are fluid and fast growing, that are populated with many open-minded consumers and first-time

buyers. It allows companies to spend less time and money that would typically be required to launch new products and identify its customers' requirements.'
The findings provide evidence that by establishing a centralised analytics team within the IT department can improve the information processing of social media and facilitate problemsolving. The interviews suggest that most case companies take this initial action when they choose to achieve a social media transformation. It enables companies and NPD managers to It was suggested that the establishment of a centralised analytics team within the IT department can improve the information processing of social media. Nonetheless, the findings illustrate that placing data analysts in every department that needs them, companies can improve the information transformation of social media that drive new insights. The rationale here is to disseminate a data-driven culture, and to provide every department with analytical thinking and skills that can be used to enhance data-driven decision making. For example, when discussing innovation, the R&D senior manager of Case M commented: 'data analysts are as a virus that needs spread to the whole organisation.' The interviews validate the significance of this approach among the case companies. It provides different examples regarding how data analysts in different departments tested prescriptive analytical models irrelevant to specific company requirements, but instead generated from a probe-and-learn process. This process is related to a greater level of re-evaluation to create novel, enhanced insights into understanding issues in principle. It offers the company with great opportunities such as the development of next-generation products based on cutting-edge technologies from new experience and learnings during the NPD. Also, it supplies innovation approaches with new ideas that differ from the formalised NPD process of the company. In this way, the knowledge was transferred from the project level to the company level. This result aligns with Koskinen and Vanharanta (2002), Cooper (2016) and Hullova et al. (2019), who identified that companies can depend on the learnings from a specific project in the advancement of future innovations. This leads to the following proposition: Proposition 6. The information transformation of social media for product testing and launch is stronger when analytical thinking and skills are highly decentralized across various departments.

Conclusion
Many researchers have pointed out social media could play various roles for company to engage consumers and to improve NPD (Sethi et al., 2003;Decker et al., 2005;Huang et al.,  ). However, the mechanisms through which social media can be leveraged for organisational learning, facilitating innovation and generating new ideas throughout the entire NPD process remain underexplored in the literature. Most prior literature has focused on social media applications in the product ideation and design phases, and ignored social media's critical approaches during product R&D and testing and launch (Kim and Wilemon, 2002;Sigala, 2012;Zhu et al., 2017). By assimilating the themes and categories presented in Figure   1, this research synthesises the six propositions and summarises the major concepts and their connections in a general form (see Figure 2) that works as a blueprint for NPD project managers to effectively leverage social media for organisational learning as well as facilitating innovation throughout the entire NPD process. In essence, the framework identifies three interdependent essential mechanisms and relevant organisational enablers that facilitate the connections between these mechanisms and NPD performance. Although these mechanisms are not in It is vital to develop a more systematic approach to leveraging social media in NPD throughout the innovation phases.
Page 24 of 37 International Journal of Operations and Production Management   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  From an organisational learning point of view, the use of social media has a positive impact on effectiveness and innovativeness across all phases. In terms of product ideation, by enabling stakeholders to contribute and collaborate, companies could seek dual processes of innovation exploitation and exploration. Particularly, stakeholders were able to provide incremental improvements as well as radical new ideas to support ideation without too many internal constraints. In this way, social media offered an environment for organisational ambidexterity during the process. During product R&D, the tendency for ambidexterity decreased as the main resources and development activities were allocated towards either radical or incremental innovation. As products moved from the mid-phase to formal launch and testing, the findings suggest that the application of social media had a positive influence on incremental innovation and the exploitation learning process.

Implications for Research and Practice
This research contributes to innovation and knowledge management literature by identifying three mechanisms by which companies can leverage social media to facilitate NPD.
Particularly, the research findings complement existing literature on using social media for knowledge acquisition, proposing that it is not just how intensely managers' notice and search, but how they manage to obtain better information than their competitors (He et al., 2017;Cheng et al., 2018). In addition, this study identifies collective learning for R&D as a mechanism for using social media to facilitate NPD. It appears to be one of the key factors of dynamic capabilities determined by Nag et al. (2007), which implies company's capability to implement internal learning and combine it with external sources of knowledge. Also, it complements prior literature on identity shifts to encourage companies to transfer from product-centric and to user-centric innovation in NPD (Hoyer et al., 2010;Cooper, 2016;Neeley and Leonardi, 2018). For the identified mechanism of fast-launch-and-improve cycle, the idea of knowledgein-action and its two inherent patterns can be used to fill in the gaps not only from the microlevel of study regarding different kinds of innovations, such as radical vs. incremental (Ritala and Hurmelinna-Laukkanen, 2013), but also from an macro or organisational level, to support the notion of double-and single-loop learning (Argyris, 1976). While prior research notes that social media offers negligible marginal costs, providing great potential for NPD (Mangold and Faulds, 2009;Mount and Martinez, 2014), this research extends the idea of this advantage by providing further case evidence of a critical mechanism of fast-launch-and-improve cycle. practices, as these differ among companies and rely on the fundamental beliefs and activities of project leaders. Although the introduction of the mechanisms identified in this study may impose considerable stress, any difficulties incurred will be more than compensated by the growth of the customer base and improvements in the effectiveness, innovativeness and market performance of the NPD process.

Limitations and Future Research
This research has some limitations. In particular, both the core strength and weakness of this research rests on the fact that all case companies selected were Chinese. While this enabled the choosing of the most representative cases to provide an in-depth understanding of how social media is used and implemented for NPD and organisational learning across all phases of the innovation funnel, it also limited generalisability. However, this approach was followed in light of the fact that no rich empirical evidence exists in the innovation and knowledge management literature, despite its importance for theory and practice. Moreover, while the focus on a single country/industry could be regarded as a limitation, it also enables a richer and deeper understanding of the elements of competitive heterogeneity with various contextual conditions remaining more or less consistent, which would be challenging in a multi-country/industry investigation. Further, the cases collectively provide coverage of different high-tech companies. Although diversity exists in the high-tech industry, the key mechanisms of leveraging social media in NPD are shared among all industries, given that social media is a relatively 'low-content' technology and can be generalised across different contexts (Adner and Kapoor, 2010;Chiu et al., 2012). Thus, the research findings can be somewhat extended, and advance policy and practice beyond the high-tech industry and Chinese context. Future empirical research could be developed in different countries and industries with different levels research on leveraging social media in NPD sheds light on innovation and knowledge management research. This is still an emerging topic in operations and production management, and it is hoped that the identified mechanisms will offer a new way to integrate the diversity of research on digital product development and social media to stimulate both practice and research.  Research Policy, Vol. 43 No. 5, pp. 805-811. Williamson, P.J. and Yin, E. (2014), "Accelerated innovation: the new challenge from China", MIT Sloan Management Review,Vol. 55 No. 4, Woetzel, J., Chen, Y., Manyika, J., Roth, E., Seong, J. and Lee, J. (2015),  i o n a l J o u r n a l o f O p e r a t i o n s a n d P r o d u c t i o n M a n a g 32 Identify and apply external source to support such as social media service providers, digital agencies and data intermediaries.
Social media offered an environment for organisational ambidexterity during the ideation process.
Pay attention to the use of social media and spend much effort to monitor various social media platforms all the time.
Listen, discuss and observe daily activities and try to think about how to improve the current situation.
I Use social media to generate novel new product ideas that could be further developed.
Integrating social media content with existing database and systems.
Higher quality in terms of ideas generated.
Keep searching for new information; interact with lead innovators regularly.
Gaining and evaluating the subtle things that are prevalent on social media.
L New campaign to identify new product features and improve brand awareness.
Use social media as an informal source of information to provide a summary for new product options.
Facilitated time-tomarket and reduced cost for ideation activities were documented.
Created a special production procedure to connect research with market information.
Offer freedom and resources to encourage innovators; apply social media to understand the market. Adapt external intermediaries to overcome the time burdens of analysing and processing.
Reported reduction of project risk and uncertainty.
Explore different sources for innovative production; combine external information to support own R&D activities.
Prefer low to medium uncertainty projects; focused on enhancing current production efficiency.
K Implement social media to improve an existing electronic tablet product features.
Use social media technical analytics to determine and improve insights generated in R&D activities.
Documented a reduction of more than 50% R&D research cost.
Exploit existing research capabilities; don't find value or input from external sources.
Put particular emphasis on feature improvements and cost reduction of existing processes and products. Appendix C: Summary of cross-case findings: Fast-launch-and-improve for product testing and launch Knowledge-in-action

Cases
Aim of the project Implementation Impacts Information processing Information transformation C Apply social media to support product testing and launch for developing a new application.
Adopt add-in functions to engage with various crosssection of customer in exploitation initiatives.
Reported 60% growth in product involvement on social media; the total cost of production was reduced by 30% through using social media.
Always think about clever ways to solve a social media task; try to make it effective and less costly to maximise the benefits.
Utilise networks and develop absorptive capacity to generate insightful views in product testing and launch.
F Adopt social media to enhance market performance and facilitate new innovations.
Use social media technological features to achieve 'viral' effects of new products.
Improved product market performance by 8% in target segment; increased potential product advocates.
If an employee comes up with a better way of solving a task, implement that new method in other NPD projects.
Harvest external good ideas, discuss and share these across NPD teams to facilitate problem solving.
J Use social media to support and accelerate the launch of new products.
Establish an internal social media platform to connect employees to communicate, coordinate and collaborate.
Documented facilitated exploitation and incremental development during product testing and launch through using social media.
Keep asking questions about how to do it effectively, how to use social media in a better way, and what you can get away with. Implement a reflective learning process and apply the outcomes to a bigger picture to work with a particular situation.
M Implement social media to identify the main and conduct various tests faster.
Analyse social media to identify customers' preferences, and enable the company to develop the right products.
Reported higher return on investment compared with traditional approaches.
Focus on using social media to solve problems through small steps rather than giant leaps.
Have an open innovation belief of 'innovation happens elsewhere' and add the science to enhance current understandings. We were pleased to have an opportunity to revise our manuscript IJOPM-04-2019-0318.R2, titled "Leveraging Social Media in New Product Development: Organisaitonal Learning Processes, Mechanisms, and Evidence from China." This paper fills a gap in the literature by investigating the mechanisms on how social media can provide important platforms for organisational learning to facilitate innovation and generate new ideas throughout the entire new product development (NPD) process.
In revising the paper, we have carefully considered your comments and suggestions. Please find enclosed the revised version of the manuscript IJOPM-04-2019-0318.R2. We have made extensive modifications to the paper and the main changes are:  Improving the findings section by making the paragraphs more consistent, and framing the propositions to make it more specific and testable.  Revising literature section (i.e., Table 1 and Section 2.1) to make the roles of social media in NPD clearer.  Improving the Abstract to show the research findings in a straightforward manner.  Improving the methodology section by highlighting the time when the research was conducted.
After providing a brief overview of the ways in which the paper was revised, we then respond to each comment and suggestion in a point-by-point fashion and outline the modifications that have been made. We are very grateful for the opportunity to revise what we consider to be a worthwhile contribution to the emerging literature in this area, and welcome further feedback on route to making the paper suitable for publication in the International Journal of Operations & Production Management.

Yours sincerely,
The Authors The paper has undergone substantial revision and rewriting to incorporate the comments and suggestions of the editors. The comments were very helpful overall, and we appreciate the very insightful and constructive feedback in response to our original submission.

Comments from the Guest Editors:
It's great to see the authors had put significant efforts to revise the paper and to address the concerns raised in the previous round of review. Overall, now the paper is easier to follow and more understandable. Still, there are a few concerns as listed below, but they should be addressable. I am looking forward to reading the revised paper.
Editors#GENERAL OVERVIEW. We are glad to know that the editors recognise the values of this study, and would like to thank the editors for the positive comments.
1. While the whole paper becomes more readable and understandable in general, I still encounter difficulty in understanding those six propositions. After reading the propositions many times, I am still unsure what messages they are trying to deliver. I also read those paragraphs prior to the propositions carefully and cannot find clear connections between those paragraphs and the corresponding propositions. Could the authors pay more attention to this issue? Is it possible to make the propositions more specific and understandable rather than just putting different terms and jargons together? Also, need to make better connections between those paragraphs and the corresponding propositions to enable readers to make sense of why such propositions are proposed in this research.
Editors#1.We would like to thank the editors for pointing this out. In the revision, we have re-visited our data to identify different organisational enablers that can facilitate the connections between the mechanisms and NPD performance. For example, after determining different cases prefer different searching modes (i.e., searching intensity and search proactiveness) to use social media for knowledge acquisition in the product ideation phase, we further explored the reasons and identified several organisational enablers that facilitate the emergence and effectiveness of different searching modes. More specifically, we identified that NPD projects with a high degree of social media searching activities were identified as always relying on a customer-focused organisational culture, while NPD projects with a high degree of social media searching proactivness were recognised as based on a diffused data-driven culture. In this way, the findings section is more consistent and propositions are more specific and testable. For the changes, please refer to section 4, on pages 13-23.
2. It's good to see a more commonly-agreed definition of social media is used in the revised paper, but I am still confused about how social media can be viewed as "co-developer" or "enduser" (Table 1). Social media, as defined in the revised paper, is just "a group of Internet-based applications" such as Facebook and Twitter. It is unclear how such an application can be a "co- developer" or "end-user". It should be the user of social media, rather than the social media per se, viewed as "co-developer" or "end-user". Moreover, I am also not connived by the matches between social media categories and NPD phases as indicated in Table 1. For example, social media as resources can be used for product R&D and also product testing and launch, not just for product ideation and design. A simple solution: drop those social media categories across Table 1 and section 2.1 to avoid unnecessary confusion and too strong arguments.
Editors#2. We agree with the editors on this point. We have followed the suggestions and removed the social media categories from Table 1. Therefore, rather than identifying the key roles of social media, the revised version now focused on the use of social media in different NPD phases. Meanwhile, we have checked Section 2.1 thoroughly and changed the terms regarding the roles of social media (i.e., as resource, as co-developer, and as end-user) to avoid unnecessary confusion. For the changes, please refer to Section 2.1, on pages 4-7.
3. In the abstract, the findings section is very brief. I am not sure whether readers can understand what the authors find in this research after reading this section. Could the authors add one or two sentences to spell out this research's findings?
Editors#3. Noted. In the revision, we have re-written the findings in the abstract. Particularly, the study identified three distinct types of mechanisms through organisational learning that extract the innovation potential inherent in social media. It further determined various organisational enablers that facilitate the connections between these mechanisms and NPD performance. Please refer to the abstract on page 1.
4. When was this research conducted? Could the authors make it clear, at least in the methodology section?
Editors#4. We appreciate the comments provided by the editors. In the revision, we have pointed out the time when the research was conducted. Particularly, starting from July 2017, a retrospective approach was applied to become familiar with the companies identified (Miller et al., 1997). Then, the main data collection took place between March 2018 and September 2019. Please refer to the changes in Section 3.2 (paragraph 3, on page 10) and Section 3.3 (paragraph 1, on page 11).