Aligning Science-Based Partnerships: Attaining Jointly Beneficial Outcomes in Open Innovation Projects

Firms are increasingly using science-based partnerships to organize for open innovation. To attain outcomes like innovations and high-quality research, science-based partnerships usually consist of multiple short-term R&D projects in which partners work closely together. However, partners often find it difficult to achieve jointly beneficial outcomes. This study explores a science-based partnership and three of its R&D projects to gain multilevel insights into how partners through alignment practices can achieve jointly beneficial outcomes. We find that partner alignment happens through practices that are influenced by structured coordination at the partnership level and mainly unstructured coordination at the project level. Our findings contribute to the literature on open innovation and coordination mechanisms by providing a multilevel view of the dynamic process of partner alignment and showing how it influences outcomes in partnerships. Our findings provide insights into why some open innovation projects fail while other projects succeed, and they have important managerial implications related to how partners in R&D projects should align to attain outcomes.


I. INTRODUCTION
C OLLABORATIONS with external partners give firms access to external resources, reduce risk, and improve time-to-market when developing innovations [1], [2].Therefore, many firms have opened up their innovation processes [3], relying on external partners to improve their innovation performance [4] which is reflected in the growing number of interorganizational partnerships [5].Opening innovation processes center on purposeful knowledge flows across firms' and their external partners' organizational boundaries [6] during different phases of innovation development [4].Firms often use open innovation to find solutions to technical problems and to enhance their understanding of the technological possibilities or scientific knowledge to improve their processes.To this end, they engage in science-based partnerships [7].Firm partnerships with universities and research organizations (henceforth science-based partners) are beneficial for both firms and sciencebased partners for several reasons.For science-based partners, partnering with firms may help them disseminate their novel scientific knowledge and capture value from this knowledge via spinouts, licensing, and patenting [8].For firms, partnerships with science-based partners are a good way to test and refine technologies [7], [9].Like interorganizational partnerships in general, science-based partnerships between science-based partners and firms are growing in number [10].
However, prior studies have found mixed results on the success of open innovation and, more particularly, the success of science-based partnerships [11].Partners in such constellations are known to have different institutional logics and conflicting goals [12], which may impede collaboration and the attainment of outcomes [13].Recent literature reviews have also raised the point that open innovation failures have received little attention [e.g., 14], rendering our ability to understand these mixed results limited as well.A plausible reason for these mixed findings is that studies on open innovation and science-based partnerships have typically focused exclusively on the firm level [7], [15], disregarding what goes on at the project level.Much of the innovation activities and collaboration processes within science-based partnerships are organized in projects [7], [16], and as firms likely keep a portfolio of projects at any given time [17], firm-level results obtained via science-based partnerships may differ from the results at the project level [18], [19].Hence, to understand firm-level performance, a multilevel perspective that also accounts for how firms and science-based partners collaborate within such partnerships at the project level is needed [5], [19], [20].Furthermore, to ensure effective and successful science-based partnerships, prior studies have argued that the collaboration requires suitable management approaches [7].Thus, there is a need for more in-depth knowledge of how the collaboration process can be managed to achieve valuable outcomes [5], [21].
In this study, we draw on the coordination mechanisms literature [22], which explains how partners coordinate their behavior [23].With this, we garner in-depth knowledge on how firms and their science-based partners can align [24] and thereby manage the collaboration process to achieve valuable outcomes within projects and beyond [25].Hence, we ask the following research question: How does partner alignment at the partnership level and the project level influence jointly beneficial outcomes in science-based open innovation partnerships?
We address this research question through a single qualitative embedded case study of one science-based partnership and three of its R&D projects.This is a suitable setting for studying how partner alignment between firms and science-based partners influences the achievement of jointly beneficial outcomes, as it allows for investigating both the partnership level and multiple projects.
Our findings show that alignment practices happen through coordination activities at multiple levels and make three key contributions to the open innovation literature.First, our study contributes an in-depth understanding of how formality and informality can be combined during a collaboration at multiple levels [21].The findings show that coordination at the partnership level happens mainly formally and structured across various phases.The findings also show that at the project level, coordination is mainly unstructured and informal between the partners in the project group during the lifespan of the R&D project.Second, our findings resolve prior mixed results on the achievement of outcomes in science-based partnerships by explaining that partners engagement in alignment practices must happen at both the partnership and project levels to achieve jointly beneficial outcomes [19], [20].Third, our findings contribute with insights into how open innovation projects may fail due to the lack of specific boundary conditions, such as relational social capital [26] at the project level in science-based open innovation partnerships [14].
Furthermore, our study provides important managerial implications on how partners in science-based partnerships should be aligned at both the partnership and project levels.The rest of the article is structured as follows: Section II introduces the theoretical framework; Section III outlines the methodological approach; Section IV presents the findings; Section V provides the discussion; and Section VI presents the conclusion, implications, and limitations of this study.

II. OPEN INNOVATION
Open innovation can be understood as "a distributed innovation process based on purposively managed knowledge flows across organizational boundaries" [6, p. 17] and relates to how firms use external knowledge sources in their innovation processes [4].Purposeful knowledge flows across a firm's organizational boundaries can take three forms: inbound, outbound, and coupled knowledge flows [6].Inbound knowledge flows entail how firms acquire knowledge from external knowledge sources to enhance their innovation processes [27].Outbound knowledge flows mean how firms transfer their internal knowledge and assets externally so other organizations can utilize this knowledge [4].Coupled knowledge flows relate to how firms combine both inflows and outflows of knowledge with external partners so the partners can develop innovations together [28].There are several ways to harness these different types of knowledge flows, for instance through interorganizational partnerships [5].These are known to give firms access to new knowledge and external resources while simultaneously reducing risks when developing innovations [1], [2].However, benefiting from these partnerships is not straightforward.Prior studies have recognized that to attain benefits from open innovation, collaborations and knowledge flows requires management [6], [27].

A. Managing Open Innovation in Science-Based Partnerships
How to manage open innovation depends on who is collaborating with whom in interorganizational partnerships [5], and what they collaborate on [16].Open innovation may take place among market-based partners only, such as suppliers and customers, in which case it enables firms to attain information about market needs and to identify technical problems [7].However, open innovation may also take place with science-based partners, in which case it enables firms to access novel scientific knowledge as well as refine and test new technological solutions [9].Managing open innovation in science-based partnerships is different from, and perhaps even more demanding than, in market-based partnerships.This is partly because knowledge sharing in these settings requires in-house expertise and absorptive capacity on part of the firms, more so than with other forms of open innovation.Therefore, firms typically opt for science-based open innovation only when they have such absorptive capacity [17].
Further complicating the management of open innovation with science-based partners is the conflicting nature of firm partners' and science-based partners' goals and corresponding logics [12].Science-based partners pursue objectives, such as advancing science, applying their knowledge to solve socioeconomic problems, and/or seeking value capture from their knowledge to fund new research, e.g., [29] and [30].The logic underlying and guiding their activities in R&D is also centered around the desire for academic freedom [31].Firms on the other hand, seek to gain new knowhow that they can apply immediately or in the near future to refine and test new technologies, solve pressing problems, e.g., [32] and [33], and eventually attain pecuniary benefits [4].The logic underlying and guiding firms' R&D activities is ultimately rooted in the desire to commercialize the derivatives of their R&D and innovation processes [12].
Open innovation with science-based partners thus requires tailored governance modes to ensure successful collaboration.Prior studies are not conclusive when it comes to what these modes of governance should be.On one hand, empirical evidence suggests a need to establish formal governance modes (e.g., contractual agreements), to ensure partners remain committed throughout a partnership [16].When these contracts are in place, partners can closely interact and share knowledge at the project level [34].On the other hand, strictly formal governance modes at the project level may hamper the collaboration process, if meetings and report on progress comes in the way of actual progress and productivity [7], [35].Thus, tailoring governance modes suited for specific collaborations and specific projects Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.
seems especially important when aiming to achieve jointly beneficial outcomes, in collaborations where partners often aspire to attain different outcomes.

B. Managing Open Innovation in Science-Based Partnerships
How to manage open innovation depends on who is collaborating with whom in interorganizational partnerships [5], and what they collaborate on [16].Open innovation may take place among market-based partners only, such as suppliers and customers, in which case it enables firms to attain information about market needs and to identify technical problems [7].However, open innovation may also take place with science-based partners, in which case it enables firms to access novel scientific knowledge as well as refine and test new technological solutions [9].Managing open innovation in science-based partnerships is different from, and perhaps even more demanding than, in market-based partnerships.This is partly because knowledge sharing in these settings requires in-house expertise and absorptive capacity on part of the firms, more so than with other forms of open innovation.Therefore, firms typically opt for science-based open innovation only when they have such absorptive capacity [17].
Further complicating the management of open innovation with science-based partners is the conflicting nature of firm partners' and science-based partners' goals and corresponding logics [12].Science-based partners pursue objectives, such as advancing science, applying their knowledge to solve socioeconomic problems, and/or seeking value capture from their knowledge to fund new research, e.g., [29] and [30].The logic underlying and guiding their activities in R&D is also centered around the desire for academic freedom [31].Firms on the other hand, seek to gain new knowhow that they can apply immediately or in the near future to refine and test new technologies, solve pressing problems, e.g., [32] and [33], and eventually attain pecuniary benefits [4].The logic underlying and guiding firms' R&D activities is ultimately rooted in the desire to commercialize the derivatives of their R&D and innovation processes [12].
Open innovation with science-based partners thus requires tailored governance modes to ensure successful collaboration.Prior studies are not conclusive when it comes to what these modes of governance should be.On one hand, empirical evidence suggests a need to establish formal governance modes (e.g., contractual agreements), to ensure partners remain committed throughout a partnership [16].When these contracts are in place, partners can closely interact and share knowledge at the project level [34].On the other hand, strictly formal governance modes at the project level may hamper the collaboration process, if meetings and report on progress comes in the way of actual progress and productivity [7], [35].Thus, tailoring governance modes suited for specific collaborations and specific projects seems especially important when aiming to achieve jointly beneficial outcomes, in collaborations where partners often aspire to attain different outcomes.

C. Aligning Partners: Coordination Mechanisms at Different Levels
Considering the pros and cons of formal governance of open innovation with science-based partners, we propose that such partnerships should be coordinated using more than just formal governance modes, instead using a variety of coordination mechanisms simultaneously [15], [24].Coordination mechanisms are the activities and tools used to manage uncertainty in collaborative activities [36] and mark a well-established concept within the management and organizational literature [37].Coordination mechanisms have been studied in intra-and interorganizational contexts [23], [38] as well as in the specific context of open innovation and R&D projects [e.g., 15] In the context of interorganizational partnerships, coordination mechanisms can be defined as "the deliberate and orderly alignment or adjustments of partners' actions to achieve jointly determined goals" [23, p. 12].This definition seems fitting in the context of science-based open innovation partnerships and R&D projects as the firm and science-based partners must adjust and align their actions to achieve valuable outcomes [15].According to prior studies, there are two types of coordination mechanisms: structured and unstructured activities [22].Structured coordination activities are formal, predetermined, and established by a centralized management strategy [36], [39], whereas unstructured coordination activities are informal, ad hoc, and often rely on a decentralized management strategy [37], [40].
In science-based partnerships, structured coordination mechanisms are often used when establishing partnerships, and are related to activities, such as establishing contracts, overall goals, and progress plans [41], which in turn routinizes the collaboration processes.Structured coordination may also be useful when establishing partnerships with dissimilar partners, and especially when the partners have weak relational social capital (prior collaborative experience) [42], because structured coordination establishes the partnerships foundations and can facilitate for information sharing and collaboration between the partners [23].Unstructured coordination activities are more likely to be used when the partners aim toward knowledge creation and innovation development [15].In particular, unstructured coordination activities relate to unplanned meetings [43], ad hoc resource allocation [44], and information generation and knowledge sharing [22].These types of unstructured activities are needed especially in complex projects which involve uncertainty [45].Thus, dealing with projects that are complex and uncertain often requires flexibility and informality, implying that for sciencebased partnerships a lack of unstructured coordination may tip the scale at the expense of the project's success.
Unstructured coordination in open innovation with sciencebased partners seems all the more important because it enables them to mutually adjust [37], [45].This room for informal communication between partners [15] and for unscripted adjustment of their actions toward each other [46] is important considering the conflicting goals and logics of these partners [12].Conflicting goals and logics may threaten partnership performance and, in the worst case, lead to the dissolution of a partnership [13].To manage these, the partners need to handle unforeseen demands [47] that could arise over time and require both parties to partake and engage in unplanned activities to overcome these challenges, something that often cannot be resolved through structured coordination activities, such as contract development [47].
Exploring alignment practices through structured and unstructured coordination mechanisms at both the partnership and

Note:
project levels is valuable as it contributes to understanding about how partners can achieve mutually beneficial outcomes, such as scientific research, innovations, and technology development all at the same time [15], [24].Furthermore, it can provide more indepth understanding of how project partners can combine both formality and informality throughout open innovation processes at various levels [21].

A. Research Design, Context, and Case Selection
We use a qualitative research design since the research question calls for in-depth insights into the open innovation process within a science-based partnership and its projects [48].More specifically, we use a single embedded case design [49], [50] to contribute to theory development on how partner alignment occurs during collaborations and affect the attainment of outcomes.
The case is a science-based partnership and three of its R&D projects in Norway.The partnership is a part of the research program "Centers for Environment-friendly Energy Research," established to develop knowledge and value creation on renewable energy and reduction of greenhouse gasses.The program started in 2009, with eight different partnerships.These partnerships had a duration of eight years and included several universities, research organizations, and firm partners.The partnership in this study, is a part of the second round of "Centers for Environment-friendly Energy Research" which started in 2017.It has approximately 40 partners-including firms and science-based partners.The partnership has multiple project areas, some focusing on basic and fundamental research, while others concentrate on innovation and manager, and firm partners.At the project level, the informants include researchers and firm representatives that were working on the projects.technology development.The projects in this study focus on the latter, where the aim is to develop research but also contribute directly to firms' innovativeness.
The selection of units (partnership and projects) in this study is based on a combination of theoretical sampling and maximum variation sampling [50], selected for their potential relevance to contribute to the open innovation and coordination literature on partner alignment at the partnership and project levels (see Table I for overview) [15], [24].
At the partnership level, the informants include researchers from universities and research organizations that are project area managers, an administrative employee, the partnership.The firms are from two different industries-the process industry and the food industry-involved in projects with different characteristics, processes, and outcomes (see Table II).
The projects had different technology complexity, implementation constraints, and timeframes (Beta and Delta were complex not only in terms of technology, but also the available time to complete these projects and implement the solutions).The variation in industry, project focus, and outcome achievement enabled us to compare the projects from the point of view of our research question [49].

B. Data Collection
The primary data consists of 38 semi-structured interviews with firm representatives and science-based partners.
The interviews were mainly conducted face to face, but some interviews in distant parts of Norway were conducted over the phone.The interviews lasted about 60 min, and they covered topics on the partnership, the project collaboration, and the achievement of outcomes.All the interviews were recorded and transcribed.Additionally, we collected documents, such as the first partnership draft, annual work plans, project memos, and results.We also participated and observed annual meetings and workshops (see Table III).
To increase our understanding of how the partnership and projects were conducted, we supplemented the primary data with ten additional interviews between 2017 and 2019 with firm Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

TABLE II PROJECT CHARACTERISTICS AND ATTRIBUTES
Note: partners who were not directly linked to the three projects studied.This additional information enabled us to increase internal validity and corroborate our findings from the primary data [50].
To ensure the anonymity, we use codes to refer to specific informants.For the partnership, "PAM" refers to project area managers, "AE" refers to the administrative employee, and "PM" refers to the partnership manager, and "FP" refers to firm partners at the partnership level.For the projects, "F" refers to a firm representative and "S" to refer to a science-based partner.In addition, the letters "A," "B," and "D" highlight which project the informant is part of.Documents, such as annual work plans are referred to as AWP and include the year that it relates to.

C. Data Analysis
The data analysis process was inspired by a cross-case comparison analysis [49] and process analysis [51], and consisted of three steps.In the first step, we developed an overview of all the data, by reading and re-reading the interviews, documents, and field notes.We created within-case summaries to reconstruct the partnership and project events chronologically.From the pointof-view of our research question, we used these summaries to get a comprehensive understanding of the collaboration process and to distinguish events and behaviors at the partnership level from those at the project level, and to see the chain of events and relationships between these levels.In the second step, we used the theoretical framework to identify and uncover the coordination activities the partners undertook in the partnership and projects.When we had identified the coordination activities (e.g., "scheduled partnership meetings initiated by the partnership management" and "developing partnership documents"), we explored how partners used these coordination activities (e.g., "establishing the main goals of the partnership").By carefully analyzing the data and studying the coordination activities and outcomes from the coordination activities, we clustered the activities into partner alignment practices (e.g., we qualified "developing the partnership overall goals" as "structured coordination").When we had done this for the partnership level and the three projects, we did a cross case comparison of the three projects [49], which revealed similarities and differences in the partners alignment practices.In the third step, we traced practices and coordination activities over time, focusing especially on patterns in the processes.Doing so, we identified three phases at the partnership level, and three phases at the project level.More importantly, this step in our analyses allowed us to pinpoint 1) in which phase specific alignment practices were used, and 2) the interrelations between the partnership and project level.Finally, we compared our coordination findings and alignment practices with the outcomes these projects achieved and pinpointed the boundary conditions which influenced the success of the projects.

IV. FINDINGS
The findings are presented in two parts.First, we present how the partners were aligned through structured coordination at the partnership level over time.Second, we present our results on the project level, and show the variation in project outcomes and coordination in the three R&D projects.The findings document how project partners with high level of relational social capital were able to achieve partner alignment during the project, and in turn produce outcomes valuable for both, whereas partners without relational social capital did not manage to align during the project process, and in turn were unable to achieve valuable outcomes for either of the partners.

A. Partner Alignment At the Partnership Level
This section details how science-based partners and firms aligned themselves in the partnership.These findings document how partners enable partnership alignment through structured coordination in the following three phases.
2) Transition from establishment phase to operational phase.
3) Operational phase.Tables IV-VI detail and illustrate these results for each phase, showing examples of coordination activities related to scheduled meetings and workshops and the development of partnership documents.
Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

TABLE IV PHASE 1-OVERVIEW OF THE ALIGNMENT PRACTICES AND STRUCTURED COORDINATION ACTIVITIES
The establishment phase started in 2015, when the sciencebased partners and firm partners began developing a partnership draft in order to obtain funds.The partners engaged in two alignment practices in this phase.The first included establishing the main goals of the partnership.The partnership draft with aim and overall goals was first created by the science-based partners before they invited firm partners to meetings, where the main goals were presented and discussed.Then the draft was revised according to the feedback and aligned the partners towards the partnership's common goals, as the partnership manager (PM) explained: "I experienced the establishment phase as very good and that we matched what was important for the researchers and the firm partners-that the partnership as a whole can contribute to something big both for Norway and for the Norwegian firms." The second alignment practice in the establishment phase involved establishing overall working plans.The science-based partners organized meetings and workshops for all partners, to agree on what the various project areas should include: "There are a lot of discussions and meetings about what the various project areas should have as topics" (PAM2).The meetings aimed to establish activities that were of interest for all the partners, as PAM2 explained: "We have 19 project groups … the goal is that all firm partners will find at least one project group relevant."Based on these discussions, science-based partners made the overall working plans and the first annual working plan for the partnership.
When the partnership got accepted by the Research Council Norway in 2016, the partners began working on establishing the partnership constructs.This second phase marked the transition from the establishment phase to the operational phase.The partners engaged in three specific alignment practices (see Table V).First, the partnership focused on establishing and signing the partnership contract, which was done through multiple meetings, as PM explained: "I have been involved in multiple meetings with the firm partners and participated in the establishment phase, where I had the responsibility to develop the partnership contract."The contractual agreement aligned the partners through established routines related to the partnership budget and the distribution of funds to the different project areas, as well as routines that dealt with intellectual property rights agreements and publishing rights: It is a complicated budget: you have in-kind financial funds from the Research Council, from the firm partners, and from the research partners, and everything is supposed to be distributed to every activity [in the partnership] over all eight years and across all the partners (PAM2).
The second alignment practice was establishing the partnership procedures, i.e., how the partnership could ensure progress and organize the day-to-day operations.This practice was based on a set of regular meetings with the managers: We have established multiple management teams.The main management team consists of the partnership manager, a person who has the coordination responsibility, another one who is responsible for the research and one that is responsible for education.We meet every week where we discuss what happens [in the partnership], and what we need to follow up on … We Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

TABLE VI PHASE 3-OVERVIEW OF THE ALIGNMENT PRACTICES AND STRUCTURED COORDINATION
also have another management team, which includes those I have mentioned, but also the project area managers.our control group, to make that we have an overview what is happening throughout the partnership.The board of directors is the next step in the partnership.They meet four times a year, two physical meetings and two over the phone (PM).
Additionally, the partnership established reporting tools to be able to track progress, e.g., a digital system where the project area managers could report deliverables.
These structures aligned and monitored the achievement of goals.
The third alignment practice related to establishing multiple meeting arenas and included these in the annual working plans.In doing so, they aimed to develop relational links between the partners and to encourage them to discuss opportunities and ideas, both at partnership and project level, e.g., there were gatherings with all the partners, workshops with specific partners, and one-to-one meetings between project managers and firm partners.PAM5 explained: "The whole partnership is planning to meet twice a year, but smaller groups are meeting several times during the year."The firm partners in Alpha, Beta, and Delta used these meeting arenas to express their expectations and project interests.For example, a firm representative in Alpha (AF1) explained how they participated in one of these annual partnership meetings: "I had a presentation at the annual meeting, [explaining] that the science-based partners are working on projects that are not our home ground.Because it is not our technologies which they work on … so it is not so easy [for us] to find these mutual projects [they can work on]" When the partnership was fully operational, the third phase started.Here, the partners also engaged in three alignment practices (see Table VI).First, during 2018, the partnership engaged in alignment practices related to adjustment of the partnership goals to ensure that these continued to represent the partners' needs as PAM3 explained: The firms have changed their focus a bit, but again, it is a couple years since we actually wrote the partnership draft [and the overall goals of the partnership].One example I can think of is that we wrote that we wanted to use a specific gas to reduce [emissions], but the participating firms were not interested in this.
The partnership managers at various levels established a formal series of meetings and workshops for this purpose: We had a meeting with the different project area managers and project managers, where we focused on the project-level goals.We changed [the main goals] a bit, nothing radical, but specified them a bit (PAM3).
The second alignment practice in the operational phase, related to developing working plans, and ensured the partners had a unified path to follow in the partnership.These plans were discussed at the annual partnership meeting in the fall and then sent out for comments and suggestions to all partners.For instance, PAM3 told: "We have a couple of meetings during each year related to planning and the development of progress plans for the years ahead.. .. It is important that the partnership manages to get the maximum of the available resources by being coordinated".In contrast to the plans established in 2016, the new working plans focused on one year, included specific descriptions of planned activities, and an overview of the deliverables produced during 2018.
The third alignment practice was registration of partnership progress, to track developments and achievements in the partnership.For example, the tools registered PhD-students' progress plans, journal publications, conference papers, and memos related to innovation projects.The partnership also registered any deviations (PM): "It is important that we can report deviations, and that we have as much overview as possible, when we are reporting to the board about the progress."

B. Partner Alignment At the Project Level
Comparing three R&D projects within the partnership, our findings show a specific set of alignment practices and coordination approaches used by the partners, ensuring the achievement of mutually beneficial outcomes.Additionally, our findings demonstrate that specific coordination approaches and alignment practices fails when the partners have low levels of relational social capital.Hence, while both formal and informal coordination approaches were important to achieve mutually beneficial outcomes, social capital seems to be conditional for such mechanisms to work.
Next, we summarize the project outcomes and the partners levels of social capital, before we examine the project partners coordination approaches and alignment practices during the following three project phases.
1) Aligning the project idea.
3) Developing the project.1) Project Outcomes and Relational Social Capital: Our findings show that the projects yielded various results.The Alpha and Beta projects successfully managed to develop projects that had beneficial outcomes for both partners, while the Delta project failed to develop outcomes that were beneficial for either partner.We summarize these project outcomes and offer our assessment in Table VII.
In Alpha, the project group developed a report related to a proof-of-concept, proving how the firm could preheat materials before putting them into the furnace, thereby making the firm's production process more efficient.In Beta, the project group developed a technology assessment report which the firm partners used to apply for additional funding before implementing one of the heat pumps that was assessed into their new factory.
Regarding science-oriented outcomes: in Alpha, the junior researcher was able to develop a report and a master thesis, and the science-based partners developed a research article.In Beta we also identified a report and one research article produced by the science-based partners.In this case, the science-based partners in Beta had also wanted to secure their participation in implementing the heat pumps and evaluating the system.Yet the firm partner moved on to implement the pumps without involving the science-based partners (BS1): "The emails stopped coming, we contacted them and after a while we got the message."However, while the science-based partners were not able to fulfill their objectives in the project, they were able to do an evaluation of the implemented heat pump system in a different project some years later.
In Delta, none of the partners ever got what they aspired to.Even though the science-based partners developed a novel heat pump concept for the firm partner's new factory, the firm partners did not apply the concept.The firm partners ended the collaboration and went with another project altogether, which resulted in a deviation report (DS1): "If the firm partners change their focus or reject the project, the partnership still has their own bureaucracy that expects a nice memo, report, simulation, or something like that [to register as an outcome].This [process] became a mismatch, so I said, "No, this doesn't work, so this has to be a deviation."It is not a negative thing, but we cannot finish the project [alone]." When considering the outcomes of the projects, we also looked at relational social capital of the partners, in prior collaborative experience.There was considerable relational social capital at the partnership level, as several partners in each of the projects had been in some prior collaborations before.Yet relational social capital was not as strong at the project level for all projects.Relational social capital only existed among the individual partners in Alpha and Beta, who had prior collaborative experience and maintained relationships also outside of the partnerships' scope.The project group in Delta had no prior collaborative experience and had not developed a collaborative relation before entering the project.As such, while the projects Alpha and Beta had strong relational social capital, Delta lacked this social capital at the level of the project.
In the following, we explore in detail how the projects used coordination approaches and alignment practices to achieve valuable outcomes.The ability to use coordination and alignment in specific ways to the benefit of project outcomes appears to be conditional to the level of social capital that exists among project partners.
2) Phase 1-Aligning the Project Idea: When examining how project partners in Alpha, Beta, and Delta coordinated at the project level to align and achieve valuable outcomes from the partnership, we find that all the project groups coordinated through both structured and unstructured coordination in phase 1-Aligning project idea (see Table VIII).
Initiating the project idea: All the project partners coordinated in unstructured ways when initiating the project idea.This was done mainly by getting in contact with already known partners directly.For example, in Alpha and Beta, the partners initiated contact with those who they already had established relations with, to discuss the project ideas.In Delta, the firm partners also relied on their priorly known contacts to initiate the project idea.However, during this stage, the firm partners ended up getting in contact with another science-based partner, who they had no prior relations with.
Establishing the project: The science-based partners in all the projects reported general descriptions of projects to the partnership level by including them into the partnership's annual work plans.For example, the science-based partners in Beta introduced general project ideas that the firm partners could collaborate on in AWP 2016-2017: "Main objectives for the following period: Initial studies of several industry projects relevant for the [firms' industry]."The reporting of projects to the partnership, established the projects for the partners in Beta and Delta.The science-based partners in Alpha, however, contacted the firm partners and proposed the project idea to them directly.This way, they could verify that all partners subscribed to and agreed on the project idea, which they did indeed (AS1): "We [the partners] agreed to follow up [the idea] and see what we can do." Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

Note:
Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

TABLE VIII PHASE 1-ALIGNING THE PROJECT IDEA THROUGH STRUCTURED AND UNSTRUCTURED COORDINATION ACTIVITIES
3) Phase 2-Developing Project Boundaries: During phase 2, all the projects used mainly unstructured coordination and focused on developing project boundaries (see Table IX).These unstructured coordination activities rested on direct contact between the project partners and project partners' engagement in meetings.
Developing project outlines: All the project partners used unstructured coordination to develop the project outlines.The project partners in Alpha and Beta engaged in discussions related to timeframes and project outlines.For example, the partners in Alpha outlined their project in several project meetings.During these meetings, the partners in Alpha discussed their own role and their boundaries in the project, as AS1 explained: We can contribute on the general and fundamental tasks, but when it comes to projections, how the technology should look, 1 meter or 1.7 meters, and if it needs this or that type of metal, which has the same quality but different price, then this is not our area anymore.
They also decided to establish a new project, as soon as the project was finished and results attained, the AS1 explained: "I will send some junior researchers to the firm next summer, which is related to our new project."The new project was registered in the annual work plans as an activity in the project area for the next year.
The partners in Beta, did not establish meetings, but instead developed the project outlines using emails: "[The communication] was mainly through emails" (BS1).In this communication, Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

TABLE IX PHASE 2-DEVELOPING PROJECT BOUNDARIES TROUGH STRUCTURED AND UNSTRUCTURED COORDINATION
the firm partners informed the science-based partners about their timelines for the project among other things: "The firm said [to us]: "Since they [the firm] decided internally on the factory, things needed to happen fast, and then, we have to deliver fast" (BS1).
However, the partners in Delta struggled to specify and align their ideas on what the project and its boundaries were supposed to be: "There never were any real specifications related to what the firm actually wanted" (DS1).
Specifying knowledge and technology focus: When Alpha and Beta had established the project outline, they worked on specifying the knowledge and technology focus.In Alpha, the partners established meetings to discuss technological possibilities, as told by the junior researcher: "We had a couple meetings with [firm representatives] about what we wanted to look at, what the firm thinks is a good idea to explore, and what is actually possible to realize."In these meetings, the partners came up with two project suggestions, as AF1 explained: "During these discussions, we developed two hypotheses, one on preheating one of [the firm's main production material] and one on preheating the secondary [production material]." The partners in Beta specified the knowledge and technology focus using emails, and this helped to merge their interests and needs.For example, BF1 told: "We discussed back and forth in relation to the practical feasibility and some of the researchers' project [suggestions]."The partners in Delta struggled to specify the technology and knowledge focus.Despite persisting in having meetings to specify a focus, the project never succeeded to align in this respect (DS1): "There were always some suggestions during meetings, we can do this, or we can do that, but we never got the project defined, regarding what they want to do, and what they do not want to do."Hence, developing project boundaries and outline, and specify knowledge and technology focus was essential activities in Phase 2. However, succeeding with these activities seems to rely on prior relations and collaboration experience.
4) Phase 3 -Developing the Project: During phase 3, we identified divergent coordination practices in the three projects.Whereas the projects with strong relational social capital continued their coordination, to develop a project that could benefit both partners, the project with weak relational social capital stopped coordinating altogether.In fact, the firm partners in this project chose to readjust and terminate the collaboration (see Table X).
Facilitating the project group: The firm partners in projects with strong relational social capital facilitated the project group in several ways.For instance, in both projects firms allocated their human resources to help with data gathering and project discussions: "The firm employees have a lot of knowledge that is not necessarily published, that you cannot find in books, but that we can use when we write our reports and analyze our data" (AS1).In Alpha, the firm partner even built testing equipment for the science-based partners, so they could do tests in the firm's factory.By contrast, the firm partners in Delta did not facilitate for the project.Instead of facilitating the project group, their activities set out to try to readjust the project.
Readjustment of project: At one point, the project groups in Alpha and Delta began to readjust the project.In Alpha, the project group had to slightly readjust the project due to the lack of useful results from the first tests (AS1): "We had a concept that went on preheating some material, with a [specific] technology ….But it did not give us the results we expected, so we tried something new."In Delta, it was not the results that gave rise to readjustments, rather, the firm partners began to reconsider the project (DF1): "Everything in the beginning was an overall concept, which we thought we would go for, but then it changed."In Beta, the partners did not (need to) readjust the project and proceeded to work on the content.
Producing the content: Both Alpha and Beta began working on producing the content the groups agreed upon, with the science-based partners taking the main lead on the work.
In Delta, the science-based partners began working on content which was not agreed upon.
The Beta project group continued to work on the original idea they had agreed upon during Phase 2, with the sciencebased partners taking the lead in doing the work (BS1): "We did not have a very big discussion with the firm partner, it was expected that we would be innovative and a bit forwardleaning.It was the science-based partners that came up with the technological concepts, based on prior research."Similar to Beta, the science-based partners in Alpha were in the lead during this process, as the junior researcher explained: "I did not have so much contact with others beside the science-based partners."However, the junior researcher also felt he could contact the firm partners or schedule meetings whenever he wanted to.
Although the partners in Delta struggled to specify the technology and knowledge focus of the project, the science-based partners eventually began to work on producing a concept that the firm partner could use when building their new factory, thus following up on the original plan (DS1): "The initial idea for the new factory was supposed to be built as a standalone factory."Their solution was presented to and discussed with firm partners in a meeting: "We presented a vision that was very cutting-edge, based on a prior project the firm had internally, with combined energy system [and] energy storage … it was a good concept" (DS1).The science-based partners reported the concept to the partnership as well as to the firm partners.However, whether and how the concept was evaluated by the firm never became clear to the science-based partners (DS1): "We developed a report to the partnership, that was on a couple of pages, and we said that it could be used this way and that way, but we didn't get the firm partner to use the developed concept." V. DISCUSSION Our findings illuminate how structured and unstructured coordination activities are used at the partnership and project levels.The findings offer a twofold contribution: 1) describing the process and multilevel perspective on coordination in science-based open innovation, that is scarce in extant literature [19], [20], [35] which unravels this chain of events mostly at a single level, and 2) developing theoretical insights into the boundary conditions under which this process enables open innovation projects with science-based partners to attain the expected jointly beneficial outcomes in terms of high-quality research and innovation [19], [20].We visualize the process of coordination and the interplay between the partnership and project level of open innovation with science-based partners in Fig. 1.
Partnerships are established via structured and formal coordination at the partnership level, by establishing goals and working plans [41].Alignment at the partnership level in the establishment phase enables the partners to initiate and suggest projects [34].This in turn initiates coordination at the project level, which we find to be mainly unstructured and informal in nature (e.g., when engaging in aligning the project idea, developing the project boundaries, and developing the project phases), while structured coordination is used by the project group only when reporting to the partnership level.However, alignment at the partnership level does not stop at the establishment phase either.The process of alignment runs via structured coordination through all phases-also the phase related to transition from establishment phase to operational phase and the operational phase.More specifically, our findings show that the partners engage in alignment practices related to establishing partnership procedures, contracts, and meeting arenas (during phase 2) and in adjustment of partnership goals, developing working plans, and registration of progress [19] (in phase 3).How the partners align during the collaboration affects the success (or failure) of projects in attaining benefits [24], and our study is thereby able to illuminate the path toward such benefits.Yet our findings also allow us to discuss under what conditions this particular process of alignment might be expected to lead to the eventual attainment of mutually beneficial outcomes.In the following, we proceed to discuss the two main findings.

A. Mapping the Process: Partner Alignment Through Coordination At Different Levels
We find that alignment at the partnership level happens through structured coordination [22].Structured coordination, such as partaking in meetings established by centralized management, and developing partnership documents that include predetermined strategies and progress plans [36], [39], enables the partners to align.Establishing and engaging in these types of activities and practices at the partnership level is not unprecedented [16].Thus, our study echoes prior literature by suggesting that at the partnership level, firms and sciencebased partners engage in various alignment practices through structured coordination mechanisms [15].At the project level, however, our findings highlight that partners align mainly by using another type of coordination mechanism [15]: unstructured coordination.Here, we observe activities, such as initiating and participating in meetings established by the project group and contacting each other ad hoc and directly [46].These activities enable mutual alignment through the informal adjustment of behavior, knowledge sharing, and information generation [37], [45].
In contrast with prior studies, we find that unstructured coordination activities are used to fulfill functions that are usually the domain of structured coordination.Examples of activities that we identified as unstructured but are normally structured include aligning the project idea and establishing the project boundaries [41].We find that these activities mainly occurred informally and unstructured at the project level.Hence, our findings shed new light on the nature of some coordination activities and their use [22], as they suggest that the use of unstructured and informal coordination mechanisms may align partners on aspects that were previously understood to be coordinated through formality.
Potential explanations for this alternate pattern are 1) the nature of the partnership we examined and 2) the multilevel perspective we attained.First, the coordination literature is mostly rooted in firms and firm collaborations [51], [52].Instead, our study focuses on collaborations between firms and science-based partners, which means that the partners involved are institutionally different and have different ways of operating [12].These characteristics impact how the partners experience different coordination activities (e.g., science-based partners are often frustrated by very formalized processes) [7], [14].Second, the coordination framework is often applied at only one level of analysis, which explains why prior studies often find that partners align through both sets of coordination mechanisms [35].By doing a multilevel study, we suggest that the different coordination mechanisms are at play at different levels of a collaboration and that at the project level, partners mainly need to align on aspects that are often handled formally but do so in an informal and unstructured way.
There is a minor variation in coordination at the project level, when comparing the two successful projects.While the two projects (Alpha and Beta) used multiple alignment practices mainly based on unstructured coordination activities throughout three phases, the Beta project stopped engaging in alignment practices altogether.The firm partner in Beta abandoned the project during the project development phase, went on to implement the heat-pump without the science-based partners.This is possibly related to the different complexity of the two projects.That is, the Beta project was more complex than the Alpha, with not only novel technology and complex implementation process, but also much a shorter timeframe for completion.

B. Relational Social capital: A Boundary Condition for Successful Coordination
Beyond a descriptive account of the multilevel process by which open innovation with science-based partners unfolds and is coordinated, we are able to pinpoint a boundary condition for attaining project success when following this process.
Particularly, unstructured activities at the project level eventually seem decisive in whether projects succeed or fail to attain Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.
mutually beneficial outcomes, and this is related to the extent of relational social capital that exists between project level partners at the outset.Relational social capital involves the advantages that are derived from having had prior contacts with each other [53], [54], [55], such as trust and willingness to transfer knowledge.In general, this is found to drive partner selection [26], and can be leveraged to ease coordination of collaboration, by reducing opportunism and monitoring costs [56].In this study, to be precise, while all projects used relatively similar coordination mechanisms throughout the phases, the only case (Delta) where relational social capital between project partners was lacking at the outset resulted in project failure.Hence, the unstructured and informal activities used to coordinate at the project level were unable to align partners in such ways that they were able to produce any relevant outcome at all.
These findings regarding the role of previous collaborative experience offer yet a more precise perspective on earlier studies that considered relational social capital's mitigating role in collaborative tensions between science-based partners and firms [42], [54] going well beyond its role in the selection of partners and social capital's development over time.That is, while it was already clear that relational social capital could mitigate collaborative tensions and that social capital may grow over the course of a collaboration, we can now also 1) propose that unstructured and informal coordination activities will likely be dysfunctional in the absence of relational social capital, but work to the benefit of attaining outcomes in partnerships when it is present at the project level, and 2) show at what level of science-based open innovation initiatives relational social capital is of most importance.

VI. CONCLUSION AND IMPLICATIONS
By studying partner alignment practices at multiple levels within one science-based open innovation partnership and three R&D projects, the aim of this study was to contribute with in-depth insights into how partner alignment in open innovation partnerships with science-based partners may help partners achieve jointly beneficial outcomes and why it fails to do so in some projects [15], [21].In general, coordination must happen between science-based and firm partners due to their conflicting goals and institutional logics in science-based partnerships [24].We propose that the use of different coordination mechanisms at multiple levels over time is necessary to attain mutually beneficial outcomes.More specifically, formality of a collaboration is often established at the partnership level through structured coordination activities, while informality is often present at the project level through unstructured coordination activities.Hence, different kinds of coordination activities are at play simultaneously at different levels to achieve partner alignment, and the interplay of these activities and levels affects the attainment of jointly beneficial outcomes in projects and the evolution of the partnership at the partnership level.Relational social capital is a boundary condition to coordinate successfully by alternating between structured coordination at the level of the partnership and unstructured coordination at the level of projects within the partnership.These findings are relevant for open innovation research in at least three ways.First, by studying partner alignment at multiple levels, this study contributes to the open innovation literature by providing a deeper understanding of how the partners within a partnership can combine both formality and informality at multiple levels of a collaboration [21].Second, our findings put in perspective the mixed earlier findings obtained via firm-level analyses related to the achievement of outcomes in science-based partnerships [11].We investigated science-based partnerships at multiple levels.Thus, our study contributes to explaining these diverse results and offers insights into how partner alignment happens at both the partnership and project levels [19], [20].Moreover, the process view of coordination over time unraveled what happens within partnerships and their projects.Third, this study offers important insights into boundary conditions associated with failure in open innovation projects [14].Specifically, although the general mix of structured coordination at the partnership level with unstructured coordination at the project level should produce outcomes for all partners in science-based open innovation, it is necessary for partners have considerable relational social capital at the outset in order to align successfully by using unstructured activities in projects.
Furthermore, this study holds valuable lessons for the coordination literature.To date, the coordination literature has focused on how various modes of coordination contribute to partner alignment and thereby help partners attain valuable outcomes [15], [57].However, the present study highlights that successful coordination at one level is contingent on the use of other coordination mechanisms at another level in a partnership.Hence, we show how the various modes of coordination influence each other and the outcomes of partnerships.This marks a contribution to the coordination literature [22] in terms of highlighting the interrelated processes of science-based partnership coordination and the interplay between different coordination modes used at multiple levels.

A. Managerial Implications
Overall, coordinating science-based open innovation for success is a continuous effort (during different phases) that requires different coordination mechanisms (structured and unstructured) and needs to happen at different levels.Particularly, managers should leave room in science-based partnerships for projects to coordinate using unstructured, informal, and ad hoc activities at the project level, even though coordination at the partnerships works well, it is structured in nature.There is a catch here though, as the relational social capital that may exist at the partnership level and thus drive partnering choices, may not exist between the individuals who are to actually collaborate in scope of specific projects.Hence, managers should seek to establish project teams with some relational social capital, as to prevent coordination issues at the project level hindering both projects and partnerships in attaining the desired outcomes.

B. Limitations and Future Research
Case study research often has the limitation of lack of generalizability.While we did not intend on developing generalizable Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.
findings but rather aimed to contribute to theory building via an in-depth exploration of how partner alignment in open innovation with science-based partners influences the attainment of valuable outcomes, our findings are limited to the context in which they are set.When we then consider that open innovation in science can take numerous shapes and forms (i.e., ranging from academic startups and spinoffs to individual-level collaborative ties and even consulting assignments [e.g.Furthermore, while our study documents every step of the process of partner alignment during collaborations, and at multiple levels, it also raises new questions.For example, our findings indicate that prior collaborative experience (relational social capital) can tip the scale in favor of attaining outcomes while aligning cf. the process in Fig. 1.This raises the question of whether there are yet other boundary conditions that operate in similar ways.One of these boundary conditions that needs to be further investigated is project complexity, as our findings hint that even partners with prior collaborative experience might experience alignment struggles when projects they work on get increasingly complex (i.e., when these involve novel technology, implementation constrains and short time frames).Project complexity has been shown before to play a role in open innovation [17], [58], and we encourage future studies to explore the project complexity, and other potential boundary conditions in aligning partners at the project level.
Finally, we suggest that the use of a fuzzy-set qualitative comparative analysis method could shed light on further and more precise configurations of structured and unstructured coordination mechanisms that are combined at different levels of partnerships while simultaneously accounting for some of the aforementioned contingencies [59].
Aligning Science-Based Partnerships: Attaining Jointly Beneficial Outcomes in Open Innovation Projects Irina Isaeva , Ward Ooms , and Jens Petter Johansen Abstract-Firms are increasingly using science-based partnerships to organize for open innovation.To attain outcomes like innovations and high-quality research, science-based partnerships usually consist of multiple short-term R&D projects in which partners work closely together.However, partners often find it difficult to achieve jointly beneficial outcomes.This study explores a science-based partnership and three of its R&D projects to gain multilevel insights into how partners through alignment practices can achieve jointly beneficial outcomes.We find that partner alignment happens through practices that are influenced by structured coordination at the partnership level and mainly unstructured coordination at the project level.Our findings contribute to the literature on open innovation and coordination mechanisms by providing a multilevel view of the dynamic process of partner alignment and showing how it influences outcomes in partnerships.Our findings provide insights into why some open innovation projects fail while other projects succeed, and they have important managerial implications related to how partners in R&D projects should align to attain outcomes.Index Terms-Coordination mechanisms, open innovation, partner alignment, R&D projects.
, 8]), we see it fit to recommend that future research should try to replicate our findings across different forms of science-based open innovation.Studying different types of science-based open innovation may allow researchers to further break down relevant contingencies and identify further boundary conditions [e.g., 14].

TABLE I PROJECT
PARTNER CHARACTERISTICS AND GOALS IN PROJECT AND RELATIONS

TABLE V PHASE 2 -
OVERVIEW OF THE ALIGNMENT PRACTICES AND STRUCTURED COORDINATION ACTIVITIES

TABLE X PHASE 3 -
DEVELOPING THE PROJECT, OVERVIEW OF ALIGNMENT PRACTICES AND COORDINATION ACTIVITIESAuthorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.