To outcomes and beyond_ Discursively managing legitimacy struggles in outcome business models

Outcome business models (OBMs) guarantee and deliver economic and operational results for customers. The risk transfer from customer to provider enables the emergence of new value drivers, such as mutual learning. However, if the outcome-based service (OBS) customer learns the operational capabilities, based on which they are willing to rely on the OBS provider to achieve outcomes, how then does an OBS provider justify its role as a legitimate partner in the future? To answer this question, we conducted an in-depth single-case study and performed a critical discourse analysis with an OBS provider delivering outcomes. We identify causes for legitimacy struggles (lack of intentional and competence trust) in an OBM and subsequent discursive legitimation strategies used to defend legitimacy: 1) trustification, 2) rationalization, 3) authorization, and 4) normalization. For managers, we elaborate certain OBM problematics causing legitimacy struggles and offer discursive resources that can be mobilized to recreate legitimacy.


Introduction
To avoid the commoditization trap, technology companies have turned to servitization and outcome business models (OBMs) as a path to competitive advantage (Baines & Lightfoot, 2014;Story, Raddats, Burton, Zolkiewski, & Baines, 2017;Visnjic, Wiengarten, & Neely, 2016). Through outcome-based contracting, the customers pay for the outcomes achieved (Ng, Maul & Yip, 2009). Probably the most renowned examples are Rolls-Royce's 'Power by the Hour' and 'Total Care' offerings, where operating hours of jet engines are being sold instead of the actual products (Ng, Parry, Smith, Maull, & Briscoe, 2012). To deliver the outcomes agreed to in the contracts, the service providers change their business model to closely resemble and align with their customer's business model (Visnjic, Neely, & Jovanovic, 2018). The given OBM change has been presented as a process consisting of three phases: define, design and deliver (Sjödin, Parida, Jovanovic, & Visnjic, 2020). An essential feature of OBMs is that the service provider takes greater accountability of the operational and business risks of the customer (Hou & Neely, 2018). The greater accountability not only exposes the provider to increased risks but also allows them to create value in new ways (Visnjic, Jovanovic, Neely, & Engwall, 2017). For instance, the outcome provider may pursue marginal gains through internal optimization (Böhm, Backhaus, Eggert, & perspective on mitigating relational risks in OBMs. Namely, trust can be regarded as a viable factor for mitigating the relational risks (Klein Woolthuis, Hillebrand, & Nooteboom, 2005;Nooteboom, 1996). The definition of trust is twofold. First, competence trust is built on technical, cognitive, organizational and communicative competences of a partner (Nooteboom, 2002), and second, intentional trust refers to trust in the intentions of the relationship counterparty (Rousseau, Sitkin, Burt, & Camerer, 1998), i.e., that they will refrain from opportunism (Klein Woolthuis et al., 2005). Calls for further research on relational elements of OBMs have been made (Kleemann & Essig, 2013;Ng et al., 2013;Sjödin et al., 2020). Therefore, we set out to study a well-established OBM relationship between power generation solution providers and independent power producers (customers) as our unit of analysis. In this OBM, the provider promises not only to deliver the facility turnkey but also to guarantee its output (kWh) for a certain period of time under the risk of liquidated damages terms. We used 31 managerial interviews from the OBS provider firm as our primary data and annual reports of both the provider and five example customer firms as secondary data. In the process view of OBMs (Sjödin et al., 2020), we focused on the delivery phase, where the contract is already signed and the focus is on how the two parties collaborate and how well the provider is responding to evolving needs of the customer (Sjödin et al., 2020, p. 173). Discourses assist in creating legitimacy for change (Phillips, Lawrence, & Hardy, 2004;Suddaby & Greenwood, 2005). For example, managers use storytelling to build trust and to create a common sense of meaning among storytellers and listeners (Küpers, Mantere, & Statler, 2013). Thus, the current study applies discursive methodology to answer how OBS providers manage legitimacy struggles using discursive legitimation strategies (Rojo & van Dijk, 1997;Van Leeuwen & Wodak, 1999) to avoid OBS partnership's dissolution (Sjödin et al., 2020).
Studies suggest that trust is an integral element of legitimacy (Jackson & Gau, 2016;Tyler, 2001). Thus, we opted to look at legitimation through the lenses of intentional trust and competence trust (Nooteboom, 2002). Considering OBM development's legitimacy from the perspective of trust is not exclusive of formal control (i.e., contract theory; Hensher & Stanley, 2008). Research indicates a complementary and dynamic nature between the two (Klein Woolthuis et al., 2005). We found that in our OBS case, the service provider's managers use four discursive legitimation strategies to address legitimacy struggles and to extend the partnership with their customer: rationalization, authorization, normalization (Vaara, Tienari, & Laurila, 2006;Van Leeuwen, 2007;Van Leeuwen & Wodak, 1999) and trustification. The discovery of the new discursive legitimation strategy, namely, trustification, makes an important theoretical contribution to both OBM literature and discursive legitimation strategies. First, if the customer remains in doubt that the provider's intentions are opportunistic (e.g., to develop operational capabilities to sell to competitors), the strong formal control inherent to OBS makes it relatively simple to dissolve agreements. Thus, trustification as a discursive legitimation strategy can be applied to increase intentional trust (Nooteboom, 2002) among OBS customers. Second, we show that the other side of trust that is based on competence actually unfolds in the three other discursive legitimation strategies (rationalization, authorization and normalization). This finding not only adds to the content validity of the given preexisting discursive legitimation strategies but also amplifies the importance of the discovery trustification as a new discursive strategy.

Organizational legitimacy
Legitimacy is defined as social acceptability of activities and actors (Suchman, 1995;Tost, 2011). The legitimation theories have roots in sociology (Habermas, 1975). Research on discursive legitimation in management and organizations has substantially built on the seminal grammar work of Van Leeuwen (Van Leeuwen, 1996, 2007 and currently arouses considerable scholarly interest (Luyckx & Janssens, 2020;Suddaby, Bitektine, & Haack, 2017). Common for many studies concerning discursive legitimation is the unit of analysis that is of interest to wide audiences. For instance, prior studies have focused on social construction of legitimacy (and illegitimacy) of widespread publicly debated issues, such as the Eurozone crisis (Vaara, 2014), corporate social responsibility (Joutsenvirta & Vaara, 2015;Siltaoja, 2009) and mergers and acquisitions (Demers, Giroux, & Chreim, 2003;Vaara et al., 2006;Vaara & Monin, 2010). These studies have added much to our understanding of how organizations apply discursive strategies to render debated activities legitimate for an audience consisting of multiple stakeholders with varying interests (Glozer, Caruana, & Hibbert, 2018). However, fewer studies have focused on construction of legitimacy in closer relationships, such as those between an industrial service provider and its customer.

The role of legitimacy in outcome business models
Studies identify OBM as an ideal and typical form of a servitized business model, separating servitization business models as 1) productoriented service provider, 2) industrializer, 3) customized integrated solution provider, 4) platform provider, and 5) outcome provider (Kohtamäki, Parida, Oghazi, Gebauer, & Baines, 2019;Kowalkowski, Windahl, Kindström, & Gebauer, 2015). In OBMs, customers hire service providers to deliver outcomes, rather than the separate service activities and resources resulting in the outcomes (Ng et al., 2013). For example, downtimes of advanced technical systems such as manufacturing systems may lead to serious losses (Öner, Kiesmüller, & Van Houtum, 2015), which is why OEMs have turned to after-sales services to ensure operational continuity (Kim et al., 2007). In an advanced form, the service providers take over the operations leading to the outcome (Baines & Lightfoot, 2014). The arrangement does not come without barriers (Sjödin et al., 2017) but also allows new value drivers to emerge (Visnjic et al., 2017). However, to seize the opportunities enabled by the better aligned business interests (Selviaridis & Van der Valk, 2019) and mutual learning of the provider and the customer, renegotiating value creation and capturing logic becomes inevitable (Sjödin et al., 2020).
Commonly, the contract delivery phases in OBSs are relatively long (Kleemann & Essig, 2013). Given the long-termism and the close collaboration OBM usually entails, the partners in the relationship have to develop relational capital (Huikkola, Ylimäki, & Kohtamäki, 2013) required for further development of customer-oriented services, which are especially important for service providers in mature industries (Visnjic et al., 2016). Therefore, the circumstances for renegotiating outcome-based contractual terms can be argued to differ from the ones under which the initial contract was crafted. Studies highlight the importance of the qualities of social acceptance and reasonability, which the actors and their activities must have when implementing the business model (Suchman, 1995;Tost, 2011). To create and improve these qualities, organizations apply discursive legitimation strategies.

Theory development
Rationalization is a discursive legitimation strategy that is based on references to the usefulness or function of distinct practices (Vaara et al., 2006;Van Leeuwen & Wodak, 1999). As OBSs are often found in complex engineering systems (Hypko, Tilebein, & Gleich, 2010a, 2010b, driven by lifecycle logic (Kleemann & Essig, 2013), it is likely that upgrades and technological advancements will emerge over the course of the outcome delivery (Erkoyuncu et al., 2013;Öner et al., 2015, p. 868) Therefore, outcome-based service providers must rationalize subsequent the system alterations to the customer, such as whether to upgrade components preventively before they fail or correctively after they fail (Öner et al., 2015). The customer relies on the provider's competence to achieve the desired outcomes (Nooteboom, 2002). This capability-driven power allows the provider to legitimate the search for rents further in the value system and even beyond its current borders (Hao & Feng, 2018). To create acceptability and reasonability for new activities, the provider may apply rationalization as a discursive legitimation strategy by making references to discursive resources (Luyckx & Janssens, 2016) that rationalize the efforts. These references may include, for instance, reliability improvements (Patra et al., 2019) and economic benefits (Böhm et al., 2016) as instruments.
Authorization, on the other hand, is a discursive legitimation strategy that is based on references to authorities (Van Leeuwen, 2007;Van Leeuwen & Wodak, 1999). In OBMs, the risk of not achieving an outcome is not only dependent on the provider's activities (Hou & Neely, 2018). Rather, the success of the outcome delivery depends on the provider's capability to orchestrate multiple value-adding facets, such as suppliers (Caldwell & Howard, 2013;Li, Zhang, & Fine, 2013); and financers (Li, Qiu, & Wang, 2014). Thus, the justification of the evolving alignment of value creation and capturing can derive from the capabilities outside the provider's dyadic relationship with its customer (Hao & Feng, 2018). Discursive authorization attempts to create a position for an organization as a credible voice among other voices (Glozer et al., 2018). For instance, as the OBS provider is dependent on third parties, such as component suppliers, its authority over these parties depends on their ability to "keep the suppliers on their toes" (Li et al., 2013, p. 280). The given types of network capabilities (Kohtamäki, Partanen, & Möller, 2013) form a valuable reference category because they are founded on competences such as the ability to plan and coordinate activities to achieve outcomes (Nooteboom, 2002, p. 50). In conclusion, authorization as a discursive legitimation strategy builds on references to value-adding authorities to whom the provider has connections and influence.
Furthermore, normalization is a discursive legitimation strategy that is based on references to normality (Vaara & Tienari, 2008) via various discursive tactics. A manager adopting normalization as a discursive legitimation strategy should not only make exemplar references retrospectively but also in a prescriptive manner, i.e., providing examples of how should the things be (Vaara et al., 2006). Normalization can also be considered as a component of both rationalization and authorization strategies (Van Leeuwen & Wodak, 1999). Take benchmarking for instance: many OBMs owe a great deal to the ICT industry in terms of well-established practice of service level agreements (Alamri, Abbasi, Minas, & Zeephongsekul, 2018;Jain, Hasija, & Popescu, 2013;Sumo, Van der Valk, Van Weele, & Bode, 2016). By referring to cross-industry benchmarks, the service provider offers not only examples but also the rationales and authorities behind them. Thus, the discursive legitimation strategy of normalization also emerges in parallel and dynamically with other discursive legitimation strategies.
Additionally, we do not consider narrativization (Vaara et al., 2006) or mythopoesis (Van Leeuwen & Wodak, 1999) as separate legitimation strategies because we consider storytelling and narratives to be more instrumental by nature. For instance, managers may use stories as a technique to demonstrate the rationality or irrationality of a certain procedure. Furthermore, normalization strategy leans almost entirely upon a narrative method because referring to examples predominantly includes a story or a narrative to begin with. Lastly, we find moralization (Vaara et al., 2006) a bit problematic as a discursive legitimation strategy in closer business relationships. This is because when competing with the infamous "nothing personal, just business" norm (Stein, 2001), referencing moral values may prove ineffective.

The gap in the discursive legitimation theory
Moralization is a discursive legitimation strategy that indeed is based on references to higher order distinct values (Rojo & van Dijk, 1997;Van Leeuwen & Wodak, 1999). Thus, moralization seems more appropriate in cases where the audience of discursive legitimation is broader. For instance, in global corporate restructuring, the change at hand affects not only the companies involved but also national economies and labor force (Luyckx & Janssens, 2020). Thus, moralization as a discursive legitimation strategy used in global restructuring refers to higher order distinct values, such as national interests (Erkama & Vaara, 2010). In contrast, we argue that in closer relations such as between OBS customer and provider, conceptualization of a discursive legitimation strategy (Vaara et al., 2006) that aims to enhance relational trust (Sjödin et al., 2020) is required. This is because studies have found that OBSs are more likely to be successful when the relationship between the customer and the provider is characterized with qualities such as trust and regular exchange (Schaefers, Ruffer, & Böhm, 2020;Sjödin et al., 2020;Randall, Gravier & Prybutok, 2011) and relational assets such as complementarity and congruent expectations (Durugbo & Erkoyuncu, 2016;Ng et al., 2013;Schaefers et al., 2020;Visnjic et al., 2017).
In economics, moral hazard refers to a dilemma when an opportunistic hidden action results from a transaction (Arrow, 1985;Holmström, 1979). For instance, the customers of insurance companies may have a tendency to take more risks compared to uninsured consumers (Guesnerie, 1989). In the context of OBMs, opportunism (Williamson, 1975) can occur, for example, when the OBS provider develops new operational capabilities by operating on behalf of the customer (Visnjic et al., 2017) and then offers the given capabilities to competitors of the customer (Sjödin et al., 2017). In another instance provided by Sjödin et al. (2020), an OBS customer began to prioritize remote location deliveries that entailed higher delivery costs for the provider but not corresponding revenues, upon which the gain sharing in the contract was largely based (2020, p. 175). This example not only emphasizes that the risk of opportunism may also emerge from the customer's side but also that hidden information or knowledge (Cabrales & Charness, 2011;Li et al., 2013) plays a major role in OBMs.
The key consideration of the above examples is that OBMs by design must be revisable because in the long term, the emergence of unpredictable contingencies is likely (Sjödin et al., 2020). Importantly, the successful revision of value creation and capturing alignment often depend on whether a trustful relationship exists between the customer and the provider (Luo, 2002;Sjödin et al., 2020). This is in alignment with prior research indicating a positive link between trust and organizational learning (Nonaka & Takeuchi, 1995) and innovation management (Lester & Piore, 2004). To distinguish between moral values and trust, the latter can be generally defined as the trustor's positive expectations of valued behavior of the trustee (Jackson & Gau, 2016;Lumineau, 2017;Tyler, 2006). Moral values, in contrast, are standards of good and bad (Van Leeuwen, 2007) that are often derived from individuals' social surroundings, such as society or religion (Vaara et al., 2006;Van Leeuwen & Wodak, 1999).
Scholars have distinguished between intentional trust and competence trust (Klein Woolthuis et al., 2005;Nooteboom, 2002). Competence trust, refers to the trustor's belief in the technical, cognitive, organizational and communicative competences of the partner (Klein Woolthuis et al., 2005, p. 814). The aforementioned features are likely to be found in the rationalizing, authorizing and normalizing discourses of the legitimating actor. Intentional trust, in contrast, means the trustor's trusting belief in the good intentions of the partner, and especially their refrainment from opportunism, even if short-term opportunities or incentives would emerge (Chiles & McMackin, 1996;Nooteboom, 1996). The notion of opportunism is particularly important for OBS providers seeking legitimacy because customer lock-in is usually higher than in traditional time-and material-based services (Visnjic et al., 2017). Thus, an opportunity to leverage a partner's greater dependence to appropriate a larger share of value may exist (Nooteboom, 2002, p. 51). Other forms of active opportunism include lying and stealing as means to seek interest from the relationship to the detriment of the partner (Klein Woolthuis et al., 2005;Williamson, 1975). Conclusively, a theoretical gap exists in terms of the legitimation discourse used by firms engaged in close partnerships in effort to legitimate their intentions towards their partners.

Research setting and data
The research on OBS shows that their most prominent applications are in B2B service business settings (Essig, Glas, Selviaridis, & Roehrich, 2016;Ng et al., 2013). Thus, this study is based on a broader set of data gathered for research around servitization of industrial solution sellers. The case company of our choice (henceforth Alpha) is an OEM and power solution company with operating revenue of over 6 billion USD. Over the past 10 years, service business has accounted for 41.74%, on average, of the firm's net sales, with a yearly increase of 1.82% (average absolute value 0.74%). We chose to focus on the delivery phase (Sjödin et al., 2020) of OBS. In this regard, Alpha is an interesting case company because even their earliest available annual report online (from 1999) discussed the OBS arrangements with independent power producer customers -a customer segment that originally emerged in the seventies (Eberhard, Gratwick, & Kariuki, 2018). Furthermore, the CEO of Alpha has listed developing more robust partnerships and subsequently innovating solutions and services for increased value creation as the firm's major strategic aims. Thus, Alpha as a case company provides not only evidence of OBS activities but also prospective insights into OBM development. The case evidence constituted of 31 interviews with managers involved with service offerings (titles ranging from project managers to general managers and vice presidents) and summed to 1806 min of recorded data (455 pages of transcribed interview data, Times New Roman 12, single space, 2.54 margins all around).

Research approach
We adopted an abductive single-case study design (Dubois & Gadde, 2002) because we wanted to elicit valuable prescriptive sociological knowledge of the OBM development, which currently lacks scholarship. The microlevel investigation of legitimation discourse in a complex system allowed us to take a closer look into what individuals actually experience in organizations (Boje, 2006). To explore the causes of effects (Sayer, 1992) in OBMs through an intensive research strategy, we first needed to account for the context and conditions in which the causal mechanism under scrutiny, namely, discursive legitimation, operates. To holistically describe the setting where the discursive legitimation strategies are operationalized, we gathered data of the OBS relationship value system. For instance, we asked the respondents to describe the service and solution business models of Alpha and the value created through them. The method provided a clear illustration of the business logic behind the OBMs applied by Alpha but also brought forth the problematics involved. The depiction of the context, however, is hardly self-sufficient to construct a theory.
However, the emerged problematics laid the cornerstones for our critical analysis of the legitimation discourse. We continued the data collection by asking how the managers see value should ideally be created in order for the given problematics to be avoided. Therefore, our approach simulated Alpha respondents' legitimation towards their customers, thus differentiating our methodology from the prior studies on legitimation that often consider legitimation discourse retrospectively, through media texts for instance (Vaara & Tienari, 2008). The experiences with OBMs were easily retrieved and expounded upon by the respondents currently experiencing the phenomenon (Gehman et al., 2018).
To cross-validate both the contextual information and the perspectives on discursive legitimation, we also analyzed 2008-2018 annual reports of Alpha, with a specific emphasis on strategy, service business and energy market sections (the number of pages analyzed amounted to 191). To further validate the context from the customers' perspective, we analyzed the five most recent annual reports of five independent power producer (IPP) companies that are among the largest in the world (Asia/Pacific Rim: 2; Americas: 2; and EMEA: 1) according to Standard and Poor's Global Platts 250 ranking list 2018. The sections analyzed for Asia/Pacific Rim and Americas companies were 'Management discussion and analysis,' whereas for the EMEA company following a different report structure, the sections analyzed were 'Business report' and 'Risk report.' The page count of these sections was 874. The combined extent of the secondary data of the research amounted to 1065 pages of corporate reports.

Data analysis
The coding and analysis of the data was carried out following the Gioia methodology (Gioia, Corley, & Hamilton, 2013), utilizing the qualitative data analysis computer software NVivo by QSR International. We started by identifying common words and phrases connected to OBSs to reframe the scope of discourse to be analyzed. Such terms included but were not limited to words such as 'outcome,' 'performance,' 'guarantee,' 'risks' and 'responsibility.' The more technical terms included phrases such as 'per kilowatt hour,' 'investment,' 'return on investment,' 'profit' and 'independent power producers.' Once assured of the contextual fit of the discourse analyzed, we started to look for discourse concerning the problematics of OBMs. Subsequently, what Alpha's respondents regarded key issues challenging Alpha's legitimacy as an OBS provider, formed the first-order concepts (left-hand side) in the Fig. 1. Additionally, not only the causes for legitimacy struggles (Luyckx & Janssens, 2016;Vaara & Tienari, 2008) were discussed, but we were also able to detect discourses that represented managers' suggested responses to the issues (i.e., discursive legitimation). These discursive practices formed the first-order concepts of discursive legitimation (right-hand side). We continued the coding by grouping the first-order concepts of the detected issues into second-order problematics (left-hand side), and the first-order discursive practices into second-order discursive resources (right-hand side). Due to the emerging trust-related topics in the discourses, we reflected the findings in the light of organizational trust theories in business partnerships (Nooteboom, 1996(Nooteboom, , 2002. The use of the additional theoretical lens served as a data triangulation practice that not only strengthened our standpoint (Bryman, 2006) but also aided in bridging the related concepts of trust and legitimacy (Jackson & Gau, 2016;Tyler, 2001). The aggregate dimensions we present in the middle of the data structure Fig. 1. are a result of systematic combining and analysis of the findings. The OBM problematics form the causes of legitimacy struggles, to which the discursive legitimation strategies constituting of different discursive resources aim to answer.

Situating the legitimacy struggles and discursive legitimation strategies
When a utility firm wishes to sell more energy, buying a new facility is not its only option. Instead, the firm can establish a power purchase agreement (PPA) with an investor who invests in power generation. These investors are called independent power producers. The given investment firms usually outsource everything from the turnkey to the operation and maintenance of the facility. The hired service provider (such as Alpha) may also assist the customer in the initial stages of the venture with issues such as funding, environmental permits and establishing PPAs with the utilities to whom the power will be sold. In the contract, power is billed by kWh produced. Errors in production are compensated for the OBS customer by the provider through liquidated damages clauses. To illustrate the arrangement, we used our secondary data (annual reports) to map the actors related to this OBM. In the resulting ecosystem figure (Fig. 2), we indicate the relationship of interest for the current study by a solid line between the OBS provider (Alpha) and the OBS customer (independent power producer). In addition, we elaborate the output being sold (dashed arrow) and point out a closely interrelated business relationship (dotted line), in which the produced output is sold by the OBS customer, in turn, to its customer (utility company). Henceforth, we refer to respondents by A (as in Alpha) + number (respondent identifier).

Managing the lack of intentional trust
Paradoxically, the high contract specificity endemic to OBS, can make relational opportunism contractually feasible. This is because, in our case, the customer remains as the owner of the facility operated by the OBS provider. Thus, nothing restricts the customer to monitor and learn the valuable operative skills sold by the provider in the first place (see Table 1 in Appendix). Therefore, Alpha respondents emphasized the more informal control's importance in ensuring continuation of the partnership: A11: "…the customer-relationship should be good so that the customer practically operates as our partner and that there is a well-established trust in the collaboration and else." In order to convince the customer, that their intention is not just to capitalize on the high share of their equipment on the installation, but to build a trustworthy partnership, we found that Alpha uses a discursive legitimation strategy of trustification. Trustification builds on references to benevolence, fairness and honesty as values that guide the provider's operations. Alpha's respondents emphasized how the demonstration of benevolence often starts already in the design phase of the initial contract: A10: "…and the operations and maintenance agreement can be, for instance, 5 or 10 years, but we intend to renew it to cover all of the socalled lifecycle." Thus, from the outset, the OBS provider strives to demonstrate its refrainment from opportunism by acknowledging that outcomes provided are situated by nature and that their intention is renegotiate the terms actively during the entire lifecycle of the facility. Additionally, we found that Alpha's offerings also involve contractual bidirectional adjustments on shorter timeframes (e.g. annual escalations). Because the customer (being an investment firm) may lack knowhow of technical specs and developments, Alpha could seize the opportunity to gain advantage based on the customer's dependency. Instead, the firm offers customers a transparent process description of the outcome delivery. During the delivery, this approach allows the customer to transparently monitor the relationship with regard to its fairness, resulting in dialogue of provider deservedness: A27: "And they don't have the knowhow. That's when Alpha offers a package to deliver the turnkey and operations for 10, 15, 20 years with certain pricing policies, of course including annual escalations and

reviews, etc. But that's when investors (customers) are satisfied."
Lastly, honesty is a crucial part of intentional trust because despite communicating intentions and allowing evaluations of fairness, the trustee may not be truthful about them (Nooteboom, 2002). In other words, the OBS provider could be dishonest about its intentions or withhold information needed to evaluate fairness. In both instances, legitimation discourses impeding participation such as technologization, mystification and disciplining (Mantere & Vaara, 2008), could be employed by the OBS provider. However, in contrast, we found that Alpha endorses discursive practices that intend to involve the OBS customer in the development instead of restricting its participation. These practices included honest discussion of opportunism and serving by challenging the OBS customer's needs to renew the business model (see Table 2 for more quotes on legitimation discourses).
A28: "Before we practically served their needs, based on what they asked us to do… The current situation, on the other hand, is that we see together what works and, based on that, measure the performance. Then, based on these measures, when we reach certain performance targets, it will benefit us both. So, the contracting model is quite different from what it was before."

Managing the lack of competence trust
We found that to maintain its legitimacy as a competent provider (competence trust) in the face of evolving customer needs, the OBS provider applies discursive legitimation strategies of rationalization, authorization and normalization. First, the rationalization strategy is employed because it offers the customer a rationale of how the provider's capabilities are useful in pursuit of sustained economic gains. The given strategy is required to align the different, path-dependent strategic goals and the converging knowledge bases of the two partners (see, Appendix 1, Table 1). For example, an Alpha respondent emphasized how the OBS customer, as an investment firm, easily values transaction costs over technical specifications:

A2: "In power plants you talk with investors, people who invest money in something that is going to give them a return on the investment, and they look at money, money out and money in."
To address this problem, Alpha communicated rationales that tie together synergies between economic capabilities and operational capabilities, thus creating legitimacy for their means to realize the interlinked opportunities. We found that the rationalization strategy of the OBS provider is based on references to forecasts, technology economies and R&C of the provider. Forecasts as a source for rationality referencing are quite valuable for Alpha. This is due to the systemic complexity of not only the energy systems themselves but also the market dynamics in the energy sector. For instance, the five-minute settlement (electricity spot price settled every five minutes, used, e.g., in the US) was brought forth by an Alpha respondent:

A16: "If you think about that, the prices in the market change every five minutes; we have then taken all the prices for the last five years. That's 104,000 prices per year. And then, there are the service prices, like frequency control or these kinds of grid support services, so their prices. So that's well over half a million input cells counting in Excel. And that is then optimized every five minutes; that's what should be done with facility."
Importantly, in energy markets as described above, making a profit with a power plant does not merely depend on the ability to predict market prices. An equally important feature is whether the facility can flexibly address price peaks technically. For instance, Alpha respondents highlighted how certain traditional generation units, such as gas turbines, are somewhat suboptimal for these types of environments because they need to cool down before restarting. Conclusively, the second discursive resource used in the rationalization strategy of the OBS provider is technology economies, as well summarized by an Alpha respondent below: A16: "…these are the kind of models that turn engineering techniques to economics." Lastly, as technologies develop, operational capabilities based on the given technologies are prone to evolve as well. Thus, Alpha's respondents underscored in their discourses that keeping them involved in the business model is rational because their resources and capabilities develop together with technologies due to their manufacturing heritage. Therefore, the rationalization strategy also reflects the initial problem of the customer learning the operational capabilities during the outcome production.
A29: "The models and configurations, they need people who know the components. And the operations. That's not even necessarily the same person." Secondly, to assure the customer of their collaborative competencies, Alpha applied authorization as a discursive legitimation strategy. The authorization strategy employed by Alpha leans on references to value-adding authorities and aims to render Alpha a credible actor among these other actors. Gaining this authority-based legitimacy in the eyes of the customer is crucial for OBS providers because systeminternal internal performance (e.g., average units produced vs. capacity; Grando & Turco, 2005) can only be improved to a certain point (for more quotations, see Appendix 1. Table 1). Thus, to innovate further value creation opportunities, collaboration and gain-sharing with external authorities becomes inevitable: A15: "We just must be able to share knowledge, we must share this business model, we must share the digitalization with other industry players. Because no one of us alone can cover the entire value chain." We found that the typical authorities Alpha's discursive legitimation strategy refers to are financial, operational and technological authorities. Financial authorities such as banks or insurance companies have a major influence on the OBS customer's ventures in our case study. In this regard, OBS providers such as Alpha enjoy a special importance because they often have a proven history with financing institutions. Thus, Alpha's legitimation discourse underlined that many lenders favor Alpha to be involved as an operator. Operational authorities, such as grids or governments, form another party to which Alpha refers to in their legitimation discourse. For instance, the authorization discourse frequently speculated on possibilities the third-party collaborations could enable, such as joint predictions of system interruptions, conducted together with grid representatives (see, Appendix 2. Table 2). Lastly, in addition to financial and operational authorities, technology remains an indispensable authorization source for Alpha. For example, an Alpha respondent emphasized that the firm's technology adds to its authority among other authorities because it bridges together operational authorities' interests (e.g., renewable energy targets of governments) and financial authorities' interests (e.g., the stability of production, and hence revenues, in the intermittent renewable energy production):

A31: "And then it's perhaps also a societal element… our technology, Alpha's engines power plants; in many ways, they enable increases in renewables quite radically."
Third, not only rationales and authority are required to convince the customer of the provider competence, but also examples illustrating both sources of legitimacy in practice. Where prior studies have identified retrospective and prospective examples as main discursive resources for the discursive legitimation strategy of normalization (Vaara et al., 2006), we found that firms may also refer to more intangible, ideal examples that work as frameworks for constructing the prospective examples. Retrospective referencing often builds on problematics of the prior activities, elaborating the kind of setbacks normally expectable with those activities. For instance, a respondent brought forward an example where a facility was operated by someone other than Alpha, which led to problems caused by a lack of technical knowledge:

A16: "And I remember when I had a chat with the gentleman running that system… It turned out that they had been running it only a quarter of the operating hours agreed in the original tender… And that fellow said that they had no clue, nor did they realize, what kind of features this facility actually had."
Ideal examples, in contrast, focused on ideals that add value to the current business model, such as how Alpha regards solutions ideally as something that yields cash conversion (see Table 2 in the Appendix 2). Lastly, prospective examples present in Alpha's legitimation discourses referred to developments in progress that are likely to add value to the OBM. Importantly, these prospects leaned on not only Alpha's own resources but also those of other network actors, for example, how to integrate components such as other portfolio data into Alpha's equipment: A29: "What is ahead of us is… that the devices communicate with each other and we can gather commensurate data of the operations."

Discussion
In relations that involve high uncertainty and complexity, specific investments, and intensive knowledge transfer, one can expect explicit clauses to safeguard against multiple forms of opportunism (Klein Woolthuis et al., 2005). As all of the listed features characterize OBMs, high degree of contractual control functions are inherent to our case as well. Therefore, additions or alterations to the outcome production process require substantial efforts to renegotiate contracts (Sjödin et al., 2020). However, we found that not only technologies and customer requirements evolve during the service delivery (Erkoyuncu et al., 2013;Sjödin et al., 2020), but also the relationship between the customer and the provider is likely to develop because of the cooperation during outcome delivery. Nooteboom (2002) refers to this as "routinization," meaning that a relationship has been satisfactory for a while, and partners do not actively seek ways to gain advantage from it, while Morgan and Hunt (1994) argue that commitment and trust work in favor of relationship marketing because "it encourages marketers to resist attractive short-term alternatives in favor of the expected long-term benefits of staying with existing partners" (p. 22).
Organizations may apply discursive legitimation strategies to create, maintain and defend legitimacy (Suchman, 1995). For instance, to legitimate its existence, the auditing profession has faced a challenge of convincing users that accountants can be trusted (Holm & Zaman, 2012) in the wake of the global financial crisis and corporate scandals. Thus, trust is an integral part of gaining legitimacy (Jackson & Gau, 2016). The high contract specificity characterizing OBS seems to discourage relationship building, which may lead to opportunistic intentions and behaviors, causing legitimacy struggles between the customer and the OBS provider. To gain customer's trust in their intentions (Nooteboom, 2002), we found that OBS providers use a discursive legitimation strategy of trustification. Trustification builds on references to benevolence (Klein Woolthuis et al., 2005), fairness (Sjödin et al., 2020, p. 175) and honesty (Nooteboom, 2002) as values that guide the provider's operations. Thus, the use of trustification aims to answer the problem of conflict between contracts and trust (Malhotra & Murnighan, 2002), as strong contract control functions can decrease "goodwill-based trust" and thus decrease the net likelihood of continued collaboration (Malhotra & Lumineau, 2011). After all, trust has no meaning if benevolence or goodwill are not present (Klein Woolthuis et al., 2005;Williamson, 1993).
In addition to trust in the intentions of the partner, the broader definition of trust includes the perception of competence of the partner to perform as expected (Malhotra & Lumineau, 2011;Nooteboom, 1996). We found that to maintain its legitimacy as a competent provider in the face of evolving customer needs, the OBS provider applies discursive legitimation strategies of rationalization, authorization and normalization. The triggering legitimacy struggle for rationalization is the lack of competence trust in the means of the partner to help the customer in achieving their long-term strategic goals. The crux of the rationalization strategy of the OBS provider is to persuade the customer to be "customers who want us to do it with them" (Baines & Lightfoot, 2014, p. 4) for rational reasons. Authorization, on the other hand, aims to create, maintain and reinforce the OBS provider's legitimacy as a value adding authority worth being associated with in the industry's many networks. Therefore, it aims to help in the legitimacy struggle of lacking competence trust in the provider's ability to collaborate. As "exemplars form an important element of organizational culture, in guidance of individual action" (Nooteboom, 2002, p. 51), we found that OBM legitimacy struggles may also be caused by the lack of concreteness, especially concerning the future of the partnership. In other words, the customer requires examples of the developments envisioned by the OBS provider. We found that the corresponding discursive legitimation strategy used by the OBS provider in this legitimacy struggle is normalization, which uses examples to support legitimacy.
Importantly, neither the legitimacy struggles, nor the discursive legitimation struggles manifest without reciprocal influence. The lack of intentional trust may impede the creation of competence trust, for example, if the customer does not trust that the provider is honest. How can one trust say the examples if there is doubt whether the normalizing actor is telling the truth? The interplay works in the other direction as well. Should there be doubts concerning the means of the partner (competence trust), their benevolent orientation will have less relevance. Although distinguishable, also the discursive legitimation strategies have dynamic influence over one another. For example, an actor using the trustification strategy may demonstrate their benevolence by referring to "satisfied customers" reference-cases with other customers. The given approach uses not only argumentation demonstrating the refrainment from opportunism, but also the elements of authorization and normalization. In summary, we present our findings in Fig. 3. We display the detected problematics that have potential to create legitimacy struggles (Habermas, 1975) during outcome delivery and the discursive legitimation strategies and their respective discursive resources to manage the struggles. Importantly, although the identified discursive legitimation strategies can help defend and maintain legitimacy, it is important to remember that legitimacy can also be built through means other than discursive (Luyckx & Janssens, 2020).

Theoretical contribution
OBS offerings are not a new business model (e.g., Rolls-Royce introduced the 'Power-by-the-Hour' trademark already in 1962; Rolls-Royce.com), and relatively much is known about the formal control exercised through contracts in many industries (Huang et al., 2019;Kim et al., 2007;Patra et al., 2019;Qin et al., 2020;Selviaridis & Van der Valk, 2019). Discourse, narratives and social interactions, however, have received limited interest not only in industrial B2B cocreation research (Kohtamäki & Rajala, 2016) but also in studies concerning OBMs (Hou & Neely, 2018;Kleemann & Essig, 2013). Our study intended to extend the previous research on discursive legitimation in OBMs. Our scope of interest, namely, discourse, was motivated by strategy as practice (SAP) theory. Practices are patterns of human activities that transpire over single individuals and even organizations (Vaara & Whittington, 2012) and can be divided into sayings and doings (Luoto, Brax, & Kohtamäki, 2017). Focusing on the former, we regard discourses as linguistically mediated representations of the world (Fairclough, 2003;Vaara et al., 2006;Vaara & Tienari, 2011) that can be instrumentally deployed to gain advantage (Rojo & van Dijk, 1997), such as social acceptance for an action or an actor (Van Leeuwen & Wodak, 1999). The given process of discursive legitimation aims to respond to legitimacy struggles (Habermas, 1975;Kostova & Zaheer, 1999) that deal with power relations of the actors involved (Vaara & Tienari, 2008, p. 986).
Focusing on the delivery phase (Sjödin et al., 2020) of a well-established OBM, we found that managers in an energy technology company providing outcome-based services mobilize four discursive legitimation strategies: trustification, rationalization, authorization and normalization. We did not consider narrativization (Vaara et al., 2006) or mythopoesis (Van Leeuwen & Wodak, 1999) as separate strategies because we regard them as tools for the other strategies. Moralization (Vaara et al., 2006), on the other hand, is a legitimation strategy more suitable in cases of broader legitimation audiences. The discovery of a new discursive legitimation strategy of trustification makes an especially important contribution to the legitimation research in marketing management. This is because to the best of our knowledge, the role of trust as an endemic part of legitimacy judgements has not been discussed in the management legitimacy literature, despite its ingrained presence in legal and societal disciplines (see, e.g., Jackson & Gau, 2016;Tyler, 2001). To distinguish between the two concepts, trust is willingness to rely on others based on positive and confident expectations concerning their intentions and behavior (Klein Woolthuis et al., 2005;Nooteboom, 2002;Rousseau et al., 1998), whereas legitimacy is a quality of social acceptability (Suchman, 1995;Tost, 2011) and possessing rightful power (Jackson & Gau, 2016). We not only found that the process of building intentional trust (Klein Woolthuis et al., 2005) is mirrored in the trustification legitimation discourse of an OBS provider but also that elements of competence trust (trust in, e.g., the means, collaborative and communicative abilities of a partner) are clearly present in the formerly identified discursive legitimation strategies of rationalization, authorization and normalization. This finding adds to the content of the preexisting strategies and amplifies the importance of the discovery of trustification.

Managerial implications
This study was designed bearing in mind not only the theoretic contributions but also the implications for managers developing OBMs. Prior studies have greatly advanced our knowledge about formal control in OBMs. However, rather than seeing trust and contracts as substitutes (see, e.g., Lyons & Mehta, 1997), this study adds to the knowledge of partnership development rather than just effective contract design and enforcement. We posit that the creation of trust (both intentional and competence trust) helps OBS providers to defend, maintain and recreate legitimacy as a value-adding factor in the outcome-based service system. We found that this is especially important in the case of OBMs because of the high contract specificity inherent for the given business models (Qin et al., 2020) that can make relational opportunism especially appealing and contractually feasible. We offer managers not only four empirically found problematics (lacking trust in the intentions, means, collaboration and communicative abilities of the partner) that can cause legitimacy struggles (Kostova & Zaheer, 1999) during the outcome-based service delivery but also four discursive legitimation strategies (with their respective discursive resources) to defend, maintain and regain legitimacy as an OBS provider.

Limitations and future research
Like all research, this study has limitations. As the study represents an idiographic, single-case study, it seeks not to make nomothetic contributions. Nevertheless, we argue that our contextualized explanation of the causes of effects better captures the dynamics of the phenomenon. After all, legitimacy is a spatiotemporal phenomenon (Vaara et al., 2006), the discursive side of which should be studied at the levels of textual elements, production and institutional context (Fairclough, 2003). The given aspect also posits that different discursive legitimation strategies may be found in service relationships other than OBS, warranting more research. Additionally, because we chose to take a discursive perspective, further studies could investigate practices other than discursive (see, e.g., Omizzolo Lazzarotto, Borchardt, Pereira, & Almeida, 2014) by which the outcome providers build trust and legitimacy among their customers. Moreover, as we chose to investigate a well-established OBS relationship and the outcome delivery phase specifically -although not fully disregarding the path that had led to this point -future studies could place emphasis on how OBS providers create legitimacy in the define and design phases (Sjödin et al., 2020).

Funding
This research did receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Table 1 Legitimacy struggles in OBM.

Appendix 1
Further examples of quotations contributing to first-order concepts Problematics Legitimacy struggles A7: "Selling O&M agreements was… A lot of contract negotiations." A24: "Because the fundamental problem with these has always been how to measure, how to come to an agreement about the measures and how indisputable are the measures. So it's far from simple." A27: "…the investor (customer) knows that they will be able to cut costs when they, at some point, start operating themselves. They basically want to gain access to the core."

Formal control overrun relationship building
Lack of intentional trust A27: "…and of course, it's their property. You cannot restrict them there, they can monitor… after which our prospects of continuing the operation agreement are fairly limited." A17: "And that's one of the advantages of the O&M business, you don't let anybody in, we have 100% market share of that installation for the length of the agreement" Learning opportunity creates room for relational advantage seeking A21: "So, they can be investment firms that may not have the knowhow of the details of energy markets but regard it as a good investment that could pay itself back in, well, a couple of years." A27: "…we must point out what is the customer's revenue generation model in it."

Primary focus on economic results
Lack of competence trust: means of the partner A13: "…customers want to focus on their core business and forget the engine room and power generation in these facilities." A2: "They don't think about any technical scopes or something so much, it's like third key solution for them." Secondary focus on technical issues A24: "…building and delivering machines, it has been the core for Alpha for ages. A10: "…to improve whole systems, it all emerges from the operations of the devices in the end." A22: "What should be understood is our solutions, our plants, how do they work, and how customers generate earnings and what's important for them, so it's extremely technical marketing." Interlinked opportunities get overlooked A18: "…to have control in-house so, that the dependency on intermediaries would not be severe." A6: "…if the customer buys the components from multiple different suppliers, then they certainly need to invest in the integration…" A25: "Different parties need different knowledge, and it may be needed for equipment monitoring and operating, or the invoicing, meaning what is the compensations based on." Intermediaries considered as costs Lack of competence trust: partner's ability to collaborate A25: "…for it to be actually scalable business, it may be that we really need the kind of parties that are not even in the picture today." A15: "And if we can't collaborate, it's just not going to fly…" Outcome provision efficiency and development depends on third parties A10: "It is very much so, that nobody really buys anything nowadays if it does not contribute in terms of added value for the system" A31: "It turns into this Steve Jobs kind of stuff that you need to tell the customer about things they do not know yet, and neither do we. So, it's nowhere near easy." Examples of the proposed developments are needed to convince the customer Lack of competence trust: partner's ability to communicate A15: "…you need to share knowledge… plans and opportunities to find some kind of interest for all to practice these things." A11:" Because solar panels, for instance, are not our own product or within our competences, but basically iron bought elsewhere to be integrated to our own (plants)" Exemplification requires admitting provider's past imperfections Appendix 2 Table 2 Discursive legitimation strategies.
Further examples of quotations contributing to first-order concepts Discursive resource Discursive legitimation strategies A15: "On the other hand, these should be viewed as opportunities rather than constraints and then to consider within the value chain, what to take on your own account, what are you trying to accomplish, and whether there is a business model that provides all the participants their respective share, so that everyone benefits." A24: "From our perspective, which is more beneficial, offering a plant repair or a new plant… Honestly, it should start from the customer value and what is best for them." Benevolence Trustification A15: "They want us to give promises, or they want that promises are kept. And they want that… if promises are not kept, it will incur harm for us and be of benefit for them." A19: "So there's always both the carrot and the stick." Fairness (continued on next page) Further examples of quotations contributing to first-order concepts Discursive resource Discursive legitimation strategies A27: "They can watch and follow what we are doing, so they have this kind of five-year university… But these are recognized issues in a sense…" A29: "And in a way, the need often comes from the customer perspective. If a company wishes to say they are customer-driven, then they really should sell… what they need, not what they ask." Honesty A13: "And electricity price is of course a good indicator, what is estimated that will happen to it in the next year and a half. It affects the investments and uptimes of power plant customers quite substantially." A7: "You have to understand where is the world going, where is the market going? What will be required in the coming years, what kind of fuel, what kind of capability of power plant" Forecasts Rationalization A7: "We also want to be our customer's best business partner and to create customer value through the creation of technology solutions that subsequently improve our customer's efficiency both economically and technically." A25: "The digital solutions must be built based on what's the business model." Technology economies A16: "So, our value proposition is an outcome of a massive calculative function" A29: "It requires that the automation engineer, IT engineer and the mechanical guys discuss with one another." Resources & Capabilities A17: "I mean we have been co-operating with the reinsurers… if you hire Alpha's operator, we will give a discount on insurance because we rely on them…" A25: "Financial institutions, investors, insurance companies, rating agencies and other parties like these may be needed." A17: "…a professional operator. It's an insurance for the insurance company." Financial authorities Authorization A15: "And might it be possible to predict some consumption spikes in the power generation in co-operation with the grid, which would give our facilities the best preparedness to react upon and predict them." A18: "One should remember that financers or international environmental organizations or legislative bodies have quite significant roles in these things." A17: "…in the western world, an owner or investor can find themselves competent manpower. But when you are in Africa it's very difficult." Operational authorities A29: "Most likely it will develop into a direction where the equipment itself contains the intelligence to monitor its own operation… A29: "And this is kind of the very core of digitalization, removing the silos and collaboration beyond borders." A17: "…requirement is that there is an experienced operator… preferably the manufacturer." Technological authorities A6: "…perhaps we just did not previously know how to sell it so that the integration would have some value to it." A31: "We understood that the plant is not just a plant, but a part of a system, it's part of the customer's investment portfolio."

Retrospective examples
Normalization A1: "In our eyes… solution is mainly conversion, for example, cash conversion." A15: "With these new digital services and measurement techniques, we end up in situations where we unambiguously have correctly measured things… everything we can measure we can guarantee." A17: "We have modules so we come with units and put it together as Lego." Ideal examples A29: "What is ahead of us is a certain integration of our installed base in terms of data. So that the devices communicate with each other and we can gather commensurate data of the operations." A16: "…to see what is their portfolio like; coal, nuclear what else is there? Then to optimize the portfolio for the future… in order to maximize the profits for that portfolio." Prospective examples