Industrial Ingredient Co-branding : A Brand Relationship Approach

Existing research on industrial ingredient co-branding alliances has largely focused on the functional and transactional value that ingredient brands provide to ingredient buyers and sellers. The current research draws on the brand relationship perspective to investigate how the relational assets of ingredient brands generate value for both ingredient buyers and for the ingredient supplier itself. For the former, such value manifests in ingredient co-branding value, and for the latter, the value lies in brand equity. The results are derived from a survey of 107 ingredient buyers from a multinational ingredient manufacturer aiming to initiate ingredient co-branding agreements in the energy component industry. The results of the structural equation model reveal that ingredient brand trust directly affects both ingredient brand loyalty and ingredient co-branding value, but indicate no significant effect between brand loyalty and ingredient co-branding value. Moreover, ingredient co-branding value proved to be a driver of ingredient supplier brand equity.


Introduction
In industrial contexts, an ingredient co-branding alliance is defined as the introduction and communication of branded ingredients in products, machinery, plants, and services that already carry the industrial buyer's brand (Erevelles, Stevenson, Srinivasan, & Fukawa, 2008;Norris, 1993).
Notable examples of ingredient co-branding alliances in the industrial context are ABB and Okuma America for the joint development of robotic industrial machinery, that between Qatar Shell and Manweir for the realization and operation of co-branded turnkey petrochemical plants, and that between FedEx and Cardinal Health for the development of a co-branded third party logistics service offering.
Current research on ingredient co-branding has adopted functional and transactional perspectives to explain how allied brands create value both for ingredient buyers such as original equipment manufacturers and sellers (Bengtsson & Servais, 2005;Erevelles et al., 2008;Ghosh & John, 2009;McCarthy & Norris, 1999;Pfoertsch, Linder, & Chandler, 2007).These prior studies investigate the ability of buyer and seller brands to provide reciprocal benefits in a co-branding alliance.Such benefits might include reduced marketing expenditure, rebates, reliable revenues, and/or, a competitive advantage in terms of brand image attributed to the spillover effect that stronger brands have on weaker ones.However, the ingredient co-branding studies available to date have underestimated the relationship that industrial actors (particularly buyers) establish with brands, which is the focus of the brand relationship perspective (Chaudhuri & Holbrook, 2001, 2002;Delgado-Ballester & Munuera-Alemán, 2005;Han & Sung, 2008;Nyadzayo, Matanda, & Ewing, 2011;Van Durme, Brodie, & Redmore, 2003).
The relationship between brands and buyers is particularly relevant in industrial ingredient-cobranding, where the ingredient brand plays a crucial role in the alliance for two main reasons.First, since ingredient brands extend into production processes, industrial buyers select supply brands not only according to the prospective economic benefits, but also on the grounds that they are influenced by relational assets (Fournier, 1994;1998;Van Durme et al., 2003).Second, in industrial ingredient co-branding strategies, ingredient brands can endorse the unobservable quality of industrial buyers' products and services (Bengtsson & Servais, 2005;Glynn, 2010;Simonin & Ruth, 1998;Andreini, Salo, Wendelin, Pezzotta, & Gaiardelli, 2015).This ingredient brand characteristic provides ingredient buyers with a competitive advantage over competitors, affecting their attitude to paying a premium price (Bendixen, Bukasa, & Abratt, 2004;Chaudhuri & Holbrook, 2001;Hutton, 1997).In addition, to consider the value created by ingredient brands only from a functional and transactional perspective is to underestimate the potential contribution of ingredient brands to brand equity (Van Durme et al., 2003).
Therefore, this research aims to demonstrate how in an industrial co-branding context, the core relational concepts-brand trust and brand loyalty-generate value from a resource-based perspective for both buyers and sellers in terms of the ingredient co-branding value and the ingredient brand equity respectively.
The purpose of this research is twofold.First, it intends to verify how, in an ingredient co-branding context, the relational assets of the ingredient brand affect the ingredient co-branding value perceived by industrial buyers.Second, it aims to show how ingredient co-branding value affects the ingredient brand equity.The research questions can be put as follows: How do the relational assets of the ingredient brand affect ingredient co-branding value as perceived by the ingredient buyer?How does this value affect the brand equity of the ingredient brand?
Focusing on a sample of industrial buyers with the potential to be involved in a co-branding initiative with a specific ingredient supplying firm, the results of the structural equation model reveal that ingredient brand trust has a greater influence on the industrial buyer value than does brand loyalty.Ingredient brand trust directly affects both ingredient brand loyalty and ingredient co-branding value.No significant effect has been identified as operating between brand loyalty and ingredient co-branding value, which, in turn, has a direct impact on ingredient brand equity.
The remainder of this research is structured as follows.First, we provide a review of ingredient co-branding from resource-based view and brand relationship perspectives.Second, drawing on those perspectives (Pfoertsch et al., 2007;Srivastava, Fahey, & Christensen, 2001;Van Durme et al., 2003), we develop a theoretical framework to model the influence relationships between brands and buyers have on ingredient co-branding value and ingredient brand equity.Subsequently, we detail the research methodology, results, and conclusions.Finally, managerial implications and suggestions for further research are presented.

Literature Review
According to the resource-based view, companies leverage tangible and intangible resources in order to gain a competitive advantage over competitors (Penrose, 1959;Wernerfelt, 1984).The resources have to be valuable, rare, inimitable and non-substitutable, in order to create a long-lasting competitive advantage (Barney, 1991) and value for stakeholders (Amit & Schoemaker, 1993;Grant, 1991;Hunt & Morgan, 1995;Peteraf, 1993).
The resource-based view has been applied in different disciplines including human resources management, entrepreneurship and innovation (Bettinelli, Bergamaschi, Kokash, & Biffignandi, 2016) finance, marketing and international business (Barney, Wright, & Ketchen, 2001).Viewed from the resource-based perspective, marketing has to produce and manage market-based assets, such as customer-based knowledge, customer relationship, and brand equity in order to create customer value, which in turn improves performance in companies in terms of purchases and higher profits (Hunt & Morgan, 1995;Srivastava et al., 2001).
Researchers in the industrial ingredient co-branding field have studied how industrial companies can use the functional and transactional attributes of their industrial brands as a source of value for buyers and sellers.Such attributes can be of value in situations such as entering new markets, the creation of barriers for new competitors, determining the share of marketing costs, and in the generation of profits and revenues (Bengtsson & Servais, 2005;Erevelles et al., 2008;Ghosh & John, 2009).The same authors have also considered brand strength a characteristic of ingredient brands able to leverage spillover effects: defined by Park, Jun, and Shocker (1996) as the perceived impression that the images of the allied brands affect each other.This effect is considered one of the main sources of value provided to ingredient buyers and sellers in terms of brand image and thus differentiation and competitive advantage vis-à-vis competitors (Besharat, 2010;Ghosh & John, 2009;McCarthy & Norris, 1999).Erevelles et al. (2008) examined the cost benefits ingredient brands can offer to industrial buyers when entering a co-branding alliance, in terms of shared marketing investments and a price rebate.At the same time, sellers benefit from higher profits resulting from the barriers raised against new potential competitors by the co-branding alliance.Bengtsson and Servais (2005) studied resellers' perceptions of a co-branding alliance between two industrial brands in which the ingredient seller had a weaker brand than the buyer did.Although the authors suggested that the ingredient brand provided value to the ingredient buyer-facilitating the entry to new markets and reducing the marketing costs at the point of sale-they warned against the possibility of an ingredient brand damaging the brand of the industrial buyer.Ghosh and John (2009) highlighted both the functional and brand strength attributes of ingredient brands that are able to leverage value for both ingredient buyers and sellers.The functional characteristics of ingredient brands provide value in terms of responsiveness to specific buyers' engineering needs and thus a higher quality perceived by final customers.The brand strength of the ingredient brand was instead found to provide a competitive advantage for ingredient buyers in terms of differentiation from competitors.In the same vein, ingredient buyers could benefit from having a long-term relationship with the ingredient buyers and profit from a spillover effect when the brand image of the buyer is stronger in the market.
Based on the literature review on ingredient co-branding in the industrial context, we can detect the relevance of industrial brands as a source of value for buyers and sellers in a co-branding context, nevertheless, researchers have ignored a crucial aspect of co-branding: the relational assets of brands.This aspect is of major concern especially in industrial markets, where the dyadic relationship between buyer and seller (Bonoma & Johnston, 1978;Frazier, 1983;Ford, 1990;Håkansson, 1982;Meehan & Wright, 2012;Baxter & Matear, 2004), and industrial network relationships (Halinen, Törnroos, & Elo, 2013;Mattsson, 1997;Paliwoda, 2012) lie at the core of exchanges and of value creation.Industrial brand relationship studies (Chaudhuri & Holbrook, 2001;Delgado-Ballester & Munuera-Alemán, 2005;Van Durme et al., 2003) provide valuable insights in this regard, offering evidence of how relational assets of industrial brands create value for both buyers and sellers involved in a co-branding alliance.
Shifting the focus of analysis from the relationship between firms to the relationship between the industrial brand and the industrial buyer, scholars have provided evidence that industrial brands help to build and maintain both dyadic (Blombäck & Axelsson, 2007;Cretu & Brodie, 2007;Glynn, 2010;Glynn et al., 2007;Gordon et al., 1993), and network relationships (Glynn et al., 2007;Mäläskä, Saraniemi, & Tähtinen, 2011).
By drawing on this perspective, Veloutsou and Taylor (2012) investigated the brand as a person in business markets-similar to the approach often adopted in consumer studies-and found empirical evidence that industrial buyers attach personality traits to industrial brands, which hints at the existence of some form of mutual bond between brand and buyer.Han and Sung (2008) showed how the relationship industrial buyers establish with industrial suppliers' brands contributes to the quality of inter-firm relationships, cooperation, and mutual exchange, and highlighted the role played by brand trust and brand loyalty.Nyadzayo et al. (2011) illustrated the role of brand trust and brand loyalty in enhancing the quality of relationships between franchisors and franchisees and thus the franchisor's brand equity.On the other hand, Van Durme et al. (2003) chose to offer a conceptual model in which the relationship that buyers establish with brands was considered capable of affecting buyer and seller brand equities.
Although these brand relationship studies have shown how the bonds between brand and buyer are pivotal to the creation of high-quality, long-term, and stable relationships in the industrial context, and how (conceptually) the relationship between brand and buyer affects brand equity, it is not still clear which relational assets industrial brands might leverage in order to create value for both buyers and sellers in a co-branding alliance.

Relational Assets of Industrial Brands
Defining assets as any tangible and intangible features firms can use to develop strategic and operative marketing initiatives able to improve their performance in their marketplace (Barney 1991), the resource-based view considers brand equity a market-based relational asset that companies can leverage to create value (Hunt, 1997;Srivastava, Shervani, & Fahey, 1998;Srivastava et al., 2001).The authors agree with Van Durme et al.'s (2003) position, who illustrated how the brand-equity concept reveals only a functional and transactional aspect of the relationship between buyers and brands, that it is largely measured by a buyers' willingness to pay a premium price (Bendixen et al., 2004;Chaudhuri & Holbrook, 2001;Hutton, 1997). Moreover, Van Durme et al. (2003) underlined that the relationship between brands and buyers contains elements of trust, affection and a willingness to maintain a long-lasting interaction, that merit deeper exploration especially in the industrial marketing context (Ambler, 1997;Brodie, Glynn, & Van Durme, 2002;Delgado-Ballester & Munuera-Alemán, 2005;Glynn, 2012).
Brand relationship studies (Chaudhuri & Holbrook, 2001;2002) have identified brand trust and brand commitment as the main relational assets of brands, and as influencing the loyalty of customers.Encompassing past experience and prior interaction between brands and buyers, brand trust has been empirically demonstrated to be the major relational asset of a brand and has been established to be a predictor of customer loyalty (Garbarino & Johnson, 1999;Delgado-Ballester & Munuera-Alemán, 2005).
Suggesting that brand loyalty already encompasses brand commitment (Aaker, 1991;Assael 1998;Beatty & Kahle, 1988;Jacoby & Chestnut, 1978), Fournier (1994;1998) demonstrated how brand loyalty can be considered a dynamic concept capturing the relationship established between consumers and brands.Brand loyalty differs from brand commitment in going further and implying an enduring desire to maintain a valued relationship, typical of industrial interaction.Acknowledging Snehota and Håkansson's assertion that "a relationship develops over time as a chain of interaction episodes, a sequence of acts and counteracts.It has a history and a future" (1995, p.25), we consider brand loyalty a relational asset, able to express the lasting desire of buyers to maintain their relationship with the relevant industrial sellers (Geiger et al., 2012).
Thus, in this study we propose brand trust and brand loyalty are the main relational assets of an industrial brand, because they are able to represent the dynamic development of the relationships from the past buyers' experience with brands toward the future buyers' desire to continue the affiliation with the brands.

Theoretical Framework and Research Hypotheses
The current research adopts a brand relationship perspective (Chaudhuri & Holbrook, 2001;Fournier, 1994;1998;Han & Sung, 2008, Van Durme et al., 2003) marked by brand trust and brand loyalty being the core concepts.Hence, in this research, brand trust and loyalty are studied as the main relational assets of the ingredient brand able to create value from a resource-based perspective.That is, for the industrial buyers, the ingredient co-branding value-defined as the ingredient buyer's perception that adding its offering to the branded ingredients delivers greater value for its customers and for the ingredient seller, the ingredient brand equity-defined as a buyer's willingness to pay a premium price for disposing of the ingredient brand.
With regard to the drivers of the relational assets of the ingredient brand, the literature reveals that brand trust and brand loyalty are affected by both functional and intangible elements (Baumgarth & Binckebanck, 2011;Doney & Cannon, 1997;Morgan & Hunt, 1994;Singh & Sirdeshmukh, 2000).In this research, the functional driver is represented by responsiveness, a construct that describes the ability of industrial suppliers to quickly and incisively respond to industrial buyers' needs (Han & Sung, 2008;Nyadzayo et al., 2011;Van Durme et al., 2003).The intangible driver is ingredient brand strength, which can be considered the result of investments that companies undertake to support their brands in terms of awareness (Van Riel et al., 2005), reputation (Kuhn, Alpert, & Pope, 2008), and image (Jensen & Klastrup, 2008).The ingredient co-branding literature has also considered brand strength one of the most important elements in prompting ingredient buyers to enter into a co-branding relationship (Besharat, 2010;Ghosh & John, 2009;Simonin & Ruth, 1998;Van Durme et al., 2003;Andreini, 2012).
In summary, the theoretical framework proposed in this research is based on responsiveness and brand strength as antecedents of brand relationship assets (brand trust and brand loyalty), which in turn affect ingredient co-branding value.Finally, ingredient co-branding value influences the premium price of the ingredient brand.The following paragraphs reveal the detail of the relationships among the aforementioned latent variables and the underlying research hypotheses.

Ingredient Supplier's Responsiveness
Responsiveness is defined as the supplier's ability to quickly respond to and resolve buyers' problems, providing the right information and actions (Roberts & Merrilees, 2007).Finding the right supplier is a matter of finding a partner sensitive to the buyer's needs and able to quickly react to them.This is even more important in ingredient co-branding alliances where both brands are jointly communicated to customers, making the readiness of the supplier to react more important than in routine commercial transactions.
To be responsive, suppliers need to create a direct relationship with buyers to smooth information exchange and to develop the ability to always respond promptly to buyers' needs (Andersson, Holm, & Johanson, 2007;Ford, 1980;Han & Sung, 2008).Consequently, it is no surprise that responsiveness affects the trust customers place in firms and brands (Han & Sung, 2008;Roberts & Merrilees, 2007).Moreover, as responsiveness is closely linked to the development of competence and to the supplier's ability to be reliable, there is also evidence of its direct impact on brand loyalty (Han & Sung, 2008;Michell, King, & Reast, 2001;Walley, Custance, Taylor, Lindgreen, & Hingley, 2007).We suggest that responsiveness exerts a direct effect on industrial buyers' brand trust and brand loyalty, and therefore offer the following research hypotheses: H1a.The responsiveness of the ingredient supplier positively influences ingredient brand trust.
H1b.The responsiveness of the ingredient supplier positively influences ingredient brand loyalty.

Ingredient Brand Strength
In ingredient co-branding, brand strength is essential to persuade ingredient buyers to engage in a co-branding initiative (Besharat, 2010;Ghosh & John, 2009;Mäläskä et al., 2011;Simonin & Ruth, 1998;Van Durme et al., 2003).This is because in ingredient co-branding, the ingredient brand must be attractive not only to final consumers but also to industrial buyers who bundle the branded ingredient with their own offerings (Pfoertsch et al., 2007).
Brand strength, according to Srivastava and Shocker (1991), involves a company's performance, longevity, growth potential, efficiencies with product portfolios, and organizational objectives.In this sense, brand strength constitutes a benchmark against which ingredient buyers and ingredient suppliers may evaluate the potential benefits stemming from an ingredient co-branding initiative.Ingredient buyers in fact enter into a co-branding initiative if the ingredient brand endorses the unobservable quality of their products and services, gaining a competitive advantage over competitors (Bengtsson & Servais, 2005;Glynn, 2010;Simonin & Ruth, 1998).Hence, the strength of the ingredient brand must be evaluated against the strength of the ingredient buyer brand.Previous contributions have empirically validated the existence of a causal relationship between brand strength and brand loyalty (Baumgarth & Binckebanck, 2011;Van Riel et al., 2005).
Since the strength of ingredient brands constitutes a quality signal (Rao & Ruekert, 1994), in line with signaling theory (Kirmani & Rao, 2000;Nelson, 1974), in which brand quality influences customer trust (Caceres & Paparoidamis, 2007;Chenet, Dagger, & O'Sullivan, 2010), we also propose the existence of a direct relationship between brand strength and brand trust and offer the following hypotheses: H2a.Ingredient brand strength positively influences the ingredient brand loyalty of the ingredient buyers.
H2b. Ingredient brand strength positively influences the ingredient brand trust of the ingredient buyers.

Brand Trust
It was Chaudhuri and Holbrook (2001) who first introduced the concept of brand trust into industrial brand relationship studies, defining it as the situation where the buyer is confident that the brand will perform its stated function (see also, Chaudhuri & Holbrook, 2002;Han & Sung, 2008).
They demonstrated that brand trust influences brand loyalty, which in turn affects the industrial brand market share and the premium price in the market.Thus, brand trust plays a pivotal role in enabling the creation of long-term and stable relationships; and also improves a firm's market share and brand equity (Chaudhuri & Holbrook, 2001;Delgado-Ballester & Munuera-Alemán, 2005).
To date, only the research of Han and Sung (2008) has identified industrial brand trust as a specific construct distinct from trust in general.The consideration of trust in the brand domain requires a shift of the focus of analysis from the relationship between firms to the relationship between the industrial brand and the industrial buyer, and also necessitates viewing the industrial brand as an asset of the firm on whose behalf marketing strategies and action are implemented.
In extending the concept of trust to include brand trust, Han and Sung (2008) established the existence of a direct relationship between brand trust and brand loyalty.That study also hypothesized and tested a direct relationship between brand trust and the quality of inter-firm relationships.Accordingly, we propose the following research hypotheses: H3a. Brand trust influences brand loyalty.

Brand Loyalty
In the industrial branding literature, brand loyalty is presented as helping to protect the relationship between buyers and sellers, deterring switching to alternative partners, and reducing the perception of risk in the environment (Chaudhuri & Holbrook, 2001, 2002;Han & Sung, 2008;Rauyruen, Miller, & Groth, 2009;Taylor, Celuch, & Goodwin, 2004).
Brand loyalty is a multidimensional construct incorporating a behavioral and an attitudinal component (Aaker, 1991;Chaudhuri & Holbrook, 2001;Jacoby & Kyner, 1973;Van Riel et al., 2005).The presence of the former is indicated by repeated purchases of the same brand over time (Oliver, 1999), while the latter refers to the attitude to the brand and the attitude to the intention to repurchase (Bennett & Rundle-Thiele, 2002;Mellens, Dekimpe, & Steenkamp, 1996).The presence and measurement of behavioral and/or attitudinal loyalty depend on the industry, the product category, and the frequency of purchase.In many industries, repurchase behavior is indicative of long-term loyalty, as when the ingredient components are used in industrial machinery and plants.For these types of products, behavioral loyalty fully explains the willingness of industrial buyers to remain in the relationship with sellers.Thus, in this research, we will use the definition of brand loyalty introduced by Han and Sung (2008, p. 812): "the degree to which an industrial buyer has repeatedly purchased a supplier's particular brand during recent years, tempered by the significance of that expenditure in terms of the buyer's total outlay for that type of product."The admittedly rather scant literature on the industrial brand relationship suggests that brand loyalty affects the quality of inter-firm relationships and therefore their performance (Han & Sung, 2008).In a similar way, we suggest that in an ingredient co-branding context, brand loyalty affects the value of co-branding as perceived by industrial buyers, and therefore we hypothesize that: H4. Brand loyalty influences ingredient co-branding value as it is perceived by ingredient buyers.

Ingredient Co-Branding Value and Ingredient Premium Price
In this research, ingredient co-branding value is equated to the extent to which the ingredient buyer perceives that adding its offering to the branded ingredients provides a higher value for its customers than could be obtained by presenting the offering under its own brand alone.This definition underlines the most important element of the value of branded ingredients, that is, their ability to provide a competitive advantage to ingredient buyers, while increasing their economic value in the market.Accordingly, if the ingredient buyer obtains a higher value through a co-branding initiative, this is supposed to positively affect the ingredient supplier's brand equity (Van Durme et al., 2003).The willingness to pay a premium price for a branded product is the main evidence for the existence of brand equity (Bendixen et al., 2004;Hutton, 1997;Subrahmanyan, 2004).
Accordingly, this research will test the direct effect of ingredient co-branding value on the ingredient buyer's willingness to pay a premium price.Thus, we state that: H5.Ingredient co-branding value affects the ingredient premium price.(Anderson & Gerbing, 1982;Churchill, 1979;Steenkamp & van Trijp, 1991), the buyer's willingness to pay a premium price is often measured by a single item in the industrial marketing context (Leischnig & Enke, 2011;Michell et al., 2001;Smith et al., 1999).Previous studies have confirmed that a single-item measure can be cautiously adopted (Diamantopoulos, Sarstedt, Fuchs, Wilczynski, & Kaiser, 2012) when the object of the construct (price in our research) can be conceptualized as concrete and singular, and if the attribute of the construct (willingness in this study) can be designated concrete (Bergkvist & Rossiter, 2007).In this research, we embraced this view, because it offered a synthetic method for measuring the ingredient premium price.
Each observed variable was measured on a 7-point Likert scale.
The questionnaire was pre-tested on 10 master's degree level students with previous work experience in industrial marketing contexts, in order to refine the wording of some questions.Feedback at this stage led us to remove one inconsistent and redundant item (Churchill, 1979).
To distribute the questionnaire, the local subsidiary of the ingredient supplier provided researchers with a database of 343 decision makers who were prospective partners for an ingredient co-branding relationship with the company.We focused on this population, which was potentially interested in an ingredient co-branding alliance, but had not signed any formal agreements at that stage.The choice to focus on this population was justified because the main purpose of this study was to investigate the value as perceived by potential partners, and not the value delivered to buyers already engaged in an ingredient co-branding alliance.
These decision makers were contacted by email and asked to complete an online questionnaire.Of the 343 approached, 117 replied; a response rate of 34%.After eliminating 10 incomplete questionnaires, 107 usable responses remained.Early and late respondents were compared on the key variables in terms of means, and no significant differences were detected (Armstrong & Overton, 1977).

Results
The composition of the sample (n = 107) was as follows: 21% OEMs; 23% panel producers; 52% power plant designers/engineers; and 11% end users.Thus, the sample was in line with the population from which it was drawn, encompassing all the companies that could potentially be involved in co-branding alliance according to the data provided by the marketing managers.As for the positions held by respondents: 48% were technicians involved in the purchasing process; 20% were purchasing personnel; and 32% were owners or chief executive officers.Of the firms represented, 75% exported worldwide.Regarding relationship duration, the average was 6.46 years (standard deviation = 1.14).
The analysis followed the two-step approach to structural equation modeling (SEM), as recommended by Anderson and Gerbing (1988).SEM was implemented using the maximum likelihood (ML) procedure.The variance-covariance SEM analysis was chosen because it has been demonstrated to be a well-established and rigorous method for the validation of confirmative and low-complexity models (Rönkkö & Evermann, 2013).

Measurement Model
In order to reduce the risk of common methods bias for data derived from a single key informant, we employed Harman's single-factor test (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).We estimated a confirmatory factor analysis to compare our model to a constrained single-factor model.In case of common method variance, the single latent factor would account for all of the variables.The single-factor fit showed no evidence of common method bias, as it exhibited χ² = 568.807and df = 90.The measurement model instead resulted in  2 = 111.48and df = 76, demonstrating significantly improved fit (pb .001).
The results of the measurement model's fit statistics are summarized in Table 1.To identify measurement items contributing to poor fit, the largest negative and positive standardized residuals were considered, as well as the scores of items' multiple squared correlations.This research followed the common research practice that trait variables measured with the Likert scale are continuous, because if the scale has seven points or more, this has proved to result in only small biases in parameter estimates (Lubke & Muthén, 2005).The items excluded from the measurement scale were three items of the construct 'supplier responsiveness'; two items from 'ingredient brand loyalty' and two items from the 'ingredient brand trust' factor.
Each measurement scale was assessed as reliable: Cronbach's alpha values ranged from a minimum of .830 to a maximum of .936,higher than the .70threshold suggested by Nunnally (1978).In addition, the composite reliability (CR) and average variance extracted (AVE) of each of the constructs were above the recommended threshold of .6 for the former and .5 for the latter (Bagozzi & Yi, 1988;Fornell & Larcker, 1981).Working with the company is easy .845(std.)5.44 1.24 The company is responsive to emerging problems .862(9.03) 5.17 1.34 The company responds quickly to our needs .890(10.56) 5.36 1.26 The company management is skilled at working with us to solve problems .744(6.42) 5.55 1.17 Ingredient brand strength .827.830.711 The awareness of the ingredient brand is higher than our corporate brand .793(std.)6.00 1.66 The brand image of the ingredient brand is higher than our corporate brand image .891(5.80) 5.84 1.56 Ingredient premium price a --- We would be willing to pay a higher price for this brand over other brands 1.00 (std.)3.36 1.64 Note.M = mean; SD = standard deviation; CA = coefficient alpha; CR = composite reliability; AVE = average variance extracted.a Owing to single-item operationalization, the coefficient alpha, composite reliability, and average variance extracted cannot be computed.
We also tested for convergent validity by verifying that each item significantly and substantially loaded onto the expected latent construct by checking that all t-values were larger than 5.23 and that all standardized parameters were larger than .5,respectively.
Finally, following Fornell and Larcker (1981), to assess the discriminant validity of the constructs, we compared the AVEs with the squared correlations for all pairs of latent variables.As the highest squared correlation was .48 and the lowest AVE was .70,all pairs of constructs met this condition, and thus problems of multicollinearity are unlikely (Fornell & Larcker, 1981;Grewal, Cote, & Baumgartner, 2004).Table 2 shows the correlations among the latent constructs.Note.All correlations are significant at < .001.
In this research, as in many other studies, the data supporting the analysis are not normally distributed.The maximum likelihood method provides reliable results when there is a non-severe lack of univariate and multivariate normality in the data (Hu, Bentler, & Kano, 1992); however, we also estimated our model with the Satorra-Bentler scaling, which showed a perfect fit (df = 82; χ 2 = 97.69;[p = .095];RMSEA = .042;CFI = .993;GFI = .88;SRMR = .066).Moreover, considering the small number of observations (n = 107), we used the factor scores of the confirmatory factor analysis to estimate the paths between the single-item measurement constructs, in order to provide more robustness to our path model analysis (Anderson & Gerbing, 1988).In addition, in this case the model was estimated with the Satorra-Bentler scaling, which also showed a perfect fit ( sb 2 = 5.568, 5 df, [p = .351];CFI = .998,GFI = .987,RMSEA = .0331,SRMR = .0257).
The model explains 51% of the variance in the ingredient brand trust, 53% of the variance in the ingredient brand loyalty, 26% of the variance in ingredient co-branding value, and finally 26% of the variance in the ingredient brand premium price.The structural model supports six of the eight hypotheses formulated (Table 3).Turning to the proposed hypotheses, the data analysis provided evidence of a positive relationship between supplier responsiveness and ingredient brand trust (γ =.633; p < .001),thus hypothesis H1a is supported.The results did not establish supplier responsiveness has a positive effect on ingredient brand loyalty, so H1b is rejected.Hypothesis H2a is supported, indicating a positive relationship between ingredient brand strength and ingredient brand loyalty (γ =.194; p < .025).Hypothesis H2b, related to the positive effect between ingredient brand strength and ingredient brand trust, is also supported (γ =.201; p < .025).
As regards the effects of ingredient brand trust, H3a proposed a positive relationship between ingredient brand trust and ingredient brand loyalty (γ =.492; p < .001),meaningH3a is supported.Moreover, as proposed by H3b, ingredient brand trust also positively affects ingredient co-branding value (γ = .298;p < .001);thus, H3b is supported.
However, there is no evidence that supports hypothesis H4, related to a direct and positive relationship between ingredient brand loyalty and ingredient co-branding value.Finally, hypothesis H5, proposing that ingredient co-branding value positively affects the ingredient brand premium price, is supported by the data analysis (γ = .873;p < .001).

Indirect Effects
In addition to the direct effects of brand trust, the indirect effects also merit consideration.In fact, as Han and Sung (2008) suggest, not controlling for the mediating role of brand trust could lead researchers to overemphasize the functional and intangible drivers of the brand-buyer relationship (brand strength and responsiveness).It is therefore important to control for the total or partial mediation role of brand trust between ingredient brand strength, responsiveness, and the other constructs directly affected by brand trust (brand loyalty and ingredient co-branding value).Accordingly, two indirect effects have been tested.The Sobel test (Baron & Kenny, 1986;Sobel, 1982) corrected for non-normal formulation (Mackinnon, 2008, p. 73;Sampson & Breunig, 1971), and was statistically significant for the following relationships: responsiveness  ingredient brand trust  ingredient brand loyalty (t=2.78;p < .05),and responsiveness  ingredient brand trust  ingredient co-branding value (t=1.98;p < .05).

Rival Models
Finally, since our analyses relied on cross-sectional data, we tested two rival models using a structural equation model framework that allows comparisons between nested and non-nested models (Byrne, 1998).
The first rival nested model added two direct effects to the original model, verifying (a) a direct effect of ingredient brand trust on ingredient premium price; and (b) a direct effect of ingredient brand trust on ingredient premium price.This model did not offer evidence of positive relationship effects in either relationship, and showed an unacceptable fit.
The second, non-nested rival model was drawn from Sirdeshmukh, Singh, and Sabol (2002), who proposed that the inter-firm relationship value mediates the relationship between trust and loyalty.Therefore, this rival model considered (a) a direct effect of ingredient brand trust on ingredient co-branding value; (b) a direct effect of ingredient co-branding value on ingredient brand loyalty; and (c) a mediating role of ingredient co-branding value in the relationship between ingredient brand trust and ingredient brand loyalty.The model with the lowest Akaike information criterion (AIC) and most consistent Akaike information criterion (CAIC) is considered the best.The rival model showed a larger AIC and CAIC than the proposed model (AICrival1 =49.56,CAICrival1=130.37).Therefore, we conclude that the proposed causal representation between the latent variables fits the data better than the alternative models.

Conclusions
This research, in utilizing the theories of the resource-based view and brand relationship contributes to the ingredient co-branding literature by identifying and highlighting the role of the relational assets of the ingredient brand (brand trust and brand loyalty) in creating value both for buyers and sellers, in terms of a stronger competitive advantage for the ingredient buyer, and higher brand equity for the ingredient supplier.
We offer researchers three main theoretical contributions.First, with regard to the role of brand trust in an industrial co-branding context, our empirical findings revealed that it was determined by the seller's responsiveness, underlining the past experiential nature of the brand trust (Delgado-Ballester & Munuera-Alemán, 2005;Garbarino & Johnson, 1999;Morgan & Hunt, 1994).In a similar vein, we found that the ingredient brand strength affects the ingredient brand trust, indicating that the strong image of the ingredient brand, being a signal of quality, affects the brand trust in an industrial context (Doney & Cannon, 1997;Kirmani & Rao, 2000;Nelson, 1974;Rao & Ruekert, 1994;Singh & Sirdeshmukh, 2000).Moreover, consistent with the previous literature, and also in a co-branding context, brand trust directly affects ingredient brand loyalty, underlining the fundamental role of brand trust in creating long-lasting relationships between buyers and industrial brands.In addition, brand trust was confirmed to be a determinant relational asset of ingredient co-branding value.This last finding confirms the relevance of brand trust to creating value for buyers, in particular in building a competitive advantage in the market.While previous research has established the relevance of brand trust in creating inter-firm quality (Han & Sung, 2008;Van Durme et al., 2003), the current research advances our understanding by demonstrating how brand trust is a fundamental relational asset of the ingredient brand that creates value for buyers in a co-branding alliance just as other functional attributes can.Finally, with regard to the mediator role of ingredient brand trust, it proves to be significant only when it relates to the suppliers' responsiveness and not the brand strength.This last finding attests to the experiential meaning of brand trust, as something different and independent from the more transactional assets of ingredient brands, like the brand image.
The second theoretical contribution regards the role of ingredient brand loyalty.As expected, it was found to be directly affected by brand trust, as trust is a pillar of any long-term relationship (Garbarino & Johnson, 1999;Morgan & Hunt, 1994).Moreover, our findings also revealed a direct and positive relationship between ingredient brand strength and brand loyalty, demonstrating that in an industrial co-branding context, brand loyalty is not only a matter of affection (commitment), but that ingredient brands have to provide transactional and functional value to buyers if they are to support a long-lasting relationship.In contrast, a supplier's responsiveness has been found to be capable of affecting ingredient brand loyalty only through the mediation of brand trust, underlining the need to create a trusting relationship going beyond the technical prowess of ingredient suppliers.
In addition, interestingly, the relationship between brand loyalty and ingredient co-branding value was found not to be significant.This result is particularly important because in an ingredient co-branding context, merely being loyal is not sufficient to ensure an ingredient buyer perceives value from a co-branding alliance.Many companies, instead, search for their ideal partners for co-branding alliances from among their more loyal buyers, but as the current research demonstrates, loyalty is not a predictor of how buyers will perceive the value of a co-branding alliance.Specifying the type of value provided by ingredient brand loyalty to buyers, our paper enhances the knowledge offered by previous industrial marketing studies that found a positive relationship between brand loyalty and a general quality and value created by inter-firm relationships (Bennett, Härtel, & McColl-Kennedy, 2005;Cretu & Brodie, 2007;Kim & Hyun, 2011;Leischnig & Enke, 2011;Michell et al., 2001;Van Riel et al., 2005).
The third theoretical contribution is related to the influence of the ingredient brand value on the ingredient brand equity.In contrast to previous research, which focused only on end users' perceptions of co-branding value (Besharat, 2010;McCarthy & Norris, 1999), this study has shown that ingredient co-branding value, as perceived by ingredient buyers, positively affects ingredient brand equity.Its findings reveal that ingredient buyers are willing to pay a premium price when they perceive value in the co-branding relationship.This result is of particular significance because it empirically demonstrates that the value ingredient sellers can extract from the ingredient brands is not only related to the final customers' perception, but is created throughout the entire industrial value chain.
In summary, the present study is the first to highlight the role of relational assets of ingredient brands, revealing the relevance of brand trust as a determinant in the creation of value for a co-branding alliance.Specifically, in an ingredient co-branding context, ingredient brand trust exerts a greater influence on inter-firm relationship value than does brand loyalty.Ingredient brand trust was found to be very significant, not only because it directly affects ingredient co-branding value, but also on account of its mediating role when experiential and prior interaction (which create value for ingredient buyers) are in focus.This means that the value generated by brands in an industrial co-branding alliance is not generated only through opportunistic resource transfer (brand strength) or frequency (brand loyalty), but also through the trust relationships that industrial brands are able to develop with industrial buyers.
The results presented suggest three main managerial implications.First, industrial brand sellers planning an ingredient co-branding alliance must consider not only the final customer's perception of value, but also understand how value is created and perceived by ingredient buyers themselves, since that perception directly affects the ingredient seller's brand equity.
Second, the most loyal ingredient buyers cannot necessarily be assumed to be the best candidates for an ingredient co-branding alliance.On the contrary, ingredient sellers should consider which ingredient buyers trust the ingredient brand the most and can extract the most value from an ingredient co-branding alliance.Thus, the selection should not only be based on those opportunistic reasons related to the strength of the ingredient brand, but should encompass the trust relationship that ingredient buyers have developed with the ingredient brand.
Third, previous studies confirmed the strength of the ingredient brand as a determinant of ingredient buyer willingness to engage in a co-branding alliance (Bengtsson & Servais, 2005;Glynn, 2010;Simonin & Ruth, 1998).In contrast, this research reveals that it is not productive to provide ingredient co-branding value to ingredient buyers, unless those buyers perceive the ingredient brand to be trustworthy.This means that moving beyond the marketing challenge to enhance the strength of industrial brands, to support their ingredient co-branding alliance; managers should direct their marketing efforts toward the development of brand trust.
According to the model tested, developing brand trust means not only investing in strengthening the brand, but also guaranteeing a higher level of responsiveness, to quickly and effectively meet buyers' needs and requests: examples include 24-hour aftersales service, extended warranties, and adaptation.
Despite the merits of this research, its two main limitations must be highlighted.The first limitation lies in the sampling, which comes from only one company (albeit a multinational company) and one country.The second limitation is linked to the small number of factors included in this study, which was a choice driven by the consideration that a long and complicated questionnaire would not have been endorsed by the firm's senior management.
Our research illuminates two further interesting research avenues.It might be interesting to extend the knowledge of perceived ingredient co-branding value (Glynn, 2012) from the points of view of different industrial segments, by researching different customers in different contexts.Since industrial branding is strongly related to specific markets and contexts, extending these findings to other industries and other contexts should be undertaken with caution.Another fruitful stream of investigation might be the consideration of a more detailed survey with a broader set of variables as possible determinants of brand trust and brand loyalty on one hand, and of brand equity on the other, addressed to middle management in order to verify if this methodological choice provides different data.
Leischnig and Enke's (2011)mponent was measured throughLeischnig and Enke's (2011)single-item measure.Although multiple-item measurement scales are the accepted praxis for measuring latent variables

Table 1 .
Composition of measures

Table 3 .
Path estimates for the proposed model