Business Model Innovation: Strategic Expansion of German Small and Medium Wineries into Hospitality and Tourism

: This article provides insights into the modalities of business-model change and innovation. On the basis of an analysis of empirical data of small and medium enterprises, a transition from wine production centrism to its expanded use in hospitality and tourism is explored. Previous research on wine tourism and hospitality predominantly focuses on a destination perspective, neglecting the organizational winery perspective. The article deploys a mixed methods approach, combining netnography and a content analysis for data collection with grounded research and clustering for theory building. The sample size included 885 German wineries. Data stemmed from two distinct sources (websites and a secondary publication in form of a wine guide) and has been analyzed through a two-step clustering algorithm as well as a Principal Component Analysis (PCA). The two-step clustering algorithm resulted in nine different business models while the PCA analysis grouped the variables into the following two categories: basic winery business model (BM) and BM extension into hospitality and tourism, thereby validating the difference between the two constructs. The results point to the diverse nature of business model extensions of wineries in tourism and hospitality, depending on their organizational type and size. This study offers a classiﬁcation of small and medium sized enterprise’s strategic business model expansion, and explores the expansion of the wine industry through wine hospitality and tourism services, starting with the winery organizational perspective, which has not been done before.


Introduction
Business model innovation is about market needs and ways in which to create and maintain value for the market, while business strategy is about differentiating businesses from competitors by being different and unique (Abraham 2013). Beginning with the market needs and customer experience, the business model innovation (BMI) approach takes into account cognitive, physical, behavioral, emotional, social and sensorial drivers of innovation and change (Keiningham et al. 2019). Changing the process of value creation, delivery and capture through business model innovation is a radically different approach than merely innovating an offer through product and/or service innovation (Baden-Fuller and Haefliger 2013; Bashir et al. 2020). The business model innovation approach provides entrepreneurial orientation, tools and process information to leverage Schumpeter's (1934) entrepreneurial innovation paradigm, where profits emerge from the ability of entrepreneurs to create new combinations. Business model innovation thereby needs to assess the feasibility of certain technologies on the market. The most effective way to do this is by changing, adapting and perfecting the BM in the early stages of new technology's development (Druilhe and Garnsey 2004). However, entrepreneurs follow different types of (agro-industrial) conventions in their decisions making, formulating typical solutions to Adm. Sci. 2021, 11, 146 2 of 12 certain problems or challenges (Guthey 2008). Understanding the typical wine, hospitality and tourism business models can therefore elucidate these conventions and provide deeper insights into the decision-making patterns of wine entrepreneurs.
Wine tourism is a hybrid industry consisting of viticulture, wine making, tourism and hospitality (McRae-Williams 2004). More specifically, wine tourism experiences consist of a combination of aesthetics, services and hospitality (Carlsen and Boksberger 2013). The "new world" wineries, which focus on tourism, usually have a connection to the wider wine trail, organize regular events, have a high product variety combined with lower-than-average prices, and lower customer-relationship orientation, combined with higher social media usage (Brannon 2016). The "old world" wineries are rooted in their wine region's gastronomies, as a unique fusion of products, traditions and etiquette-three important elements for developing hospitality and tourism (Harrington and Ottenbacher 2008). Business models in wine tourism are usually dependent on the regional informal networks among hotels, restaurants, winemakers, guides, tourist information and similar destination actors (Harms 2017). Moreover, planning the infrastructure for the development of wine tourism should take into account legal arrangements together with the resulting business models for wine tourism (Cusin and Passebois-Ducros 2015). The reason for this lies in the fact that the wine-tourism cluster only partially intersects with the wineproduction cluster. The wine-tourism cluster therefore exposes wineries to a completely different set of institutional pressures, social legitimacy expectations and organizational performance expectations (Lavandoski et al. 2016).
The wine industry deals with the production and sale of wine and therefore primarily operates in the agricultural industry. Business model innovation research in the agricultural, and more concretely, the wine industry is scarce. Due to the specific nature of agricultural production, which takes place in rural areas, mainstream business model research is only partially applicable to this industry. This is why addressing the research gap on business model change and innovation at the intersection of agricultural industry and other industries, such are tourism and hospitality, is of outmost importance for wine business. It would be necessary to understand the entrepreneurial process of risk taking, business model change and creative experimenting for use as business growth strategies in the rural context. The connection between the agricultural industry, food industry and tourism industry for wine business has been previously identified in the literature (Porter et al. 2004). However, the link between business model innovation of grape and wine production on one side, and tourism and hospitality on the other side, has not been researched in detail.
Keeping in mind the aforementioned research gap, the present research aims to provide detailed insight into the entrepreneurial process of BMI into hospitality and tourism in German small and medium sized enterprise (SME) wineries. The paper aims at classiffying the major approaches to BMI through hospitality and tourism in German wineries and extracting the major components underlying this process. The paper thereby aims at providing detailed insights into the process of making new and unique combinations of winery, hospitality and tourism businesses.

Business Model Innovation
Business model innovation can be used for conceptualizing and mapping strategic evolution, managing process of change and identifying change outcomes and consequences (Bouwman et al. 2019;Foss and Saebi 2017). The literature on business model innovation is vast and deals with planned, substantial business model change-both related to overall architecture as well as to separate components (Foss and Saebi 2017). Although some changes cannot be characterised as substantial, business model change often begins with a low-profile BMI exploration phase, which can last for several years before leading to a substantial change (Sosna et al. 2010). Business model innovation can be implemented in different modes, depending on the original business model characteristics as well as Adm. Sci. 2021, 11, 146 3 of 12 future change plans. It can involve changing one or multiple elements of the existing business model, extending the existing business model into new, similar industries, as well as adding a completely new business model to an existing business model (Kim and Min 2015;Kleinschmidt and Peters 2017;Osterwalder and Pigneur 2010). The decision-making for business model innovation in SMEs depends on three strategic levers to influence goal attainment, including key partners, investment decisions and price (Cosenz and Bivona 2020). Regulatory and copyright aspects, although outside of direct SME decisionmaking, are important mediators between new, disruptive technologies and business model innovation (Dobusch and Schüßler 2014).

Business Model Innovation in the Wine Business
Previous research on business models in the wine industry has identified four types of wineries, regarding their market share in the wine market, markets they serve and the origin of investment capital, categorizing them as either a (1) Large player, (2) Lone ranger, (3) Wine group, and (4) Diversified conglomerate (Castaldi et al. 2005). Additional approaches to business model research in the wine industry include examining the process of rethinking the customer value proposition through various governance and coordination mechanisms in co-operative wineries (Pezzillo Iacono et al. 2016).
The most innovative business models in the wine industry have the ability to eliminate unnecessary industry conventions and provide completely new avenues where certain aspects are below industry standards, while others are above industry standards (Chan Kim and Mauborgne 2005). "New world" wineries increasingly exploit economies of scale and scope and are able to adapt to the changing nature of retail sales (Jenster and Cheng 2008). Online wine retailers, wine-crowdfunding platforms and digital wine business models currently spearhead business model innovation in the wine industry (Gebauer and Ginsburg 2003;Mariani et al. 2014). However, these innovative business models are often not in direct relation to the grape and wine producing companies. They innovate the business models of wine traders and wine investors, without necessarily changing the business models of winery entrepreneurs radically.

Wine, Tourism and Hospitality, Business Models
It appears that wineries are motivated to take part in global business through two strategic options, either by wine exports to international markets or through wine tourism combined with direct sales (Chan Kim and Mauborgne 2005;Charters and Menival 2011;Dressler 2017). The benefit of direct sales through tourism is that even in case of marginal sales, and increase in synergetic value is created (Bridge 2017).
Innovative business models transform and adapt value along the value chain. This change can transform the offer from commodities to goods to services to experiences, bringing the new perspective and value creation opportunities for classical wine production and sales business models (Dressler 2016;Priilaid et al. 2020). The departure from the relatively stable business models in the wine industry and venturing out into tourism bears certain risks for vintners, which need to be addressed. It has been proven in the literature that business models in both hospitality and tourism are much more volatile and dependent on the economic cycle, which means that business model innovation capabilities become a crucial resource for surviving on the market (Barth 2011). While wine production and sales are primary economics activities, where cost minimizers with a homogenous product offer thrive, wine hospitality and tourism are a tertiary economic sector, where profit maximisers with a heterogeneous product offer thrive (Bridge 2017). In this sense, major obstacles for developing wine tourism in wineries include a lack of managerial capacity and time, followed by a lack of capacity to mobilize large tourism-related investments as well as low regional visibility of tourism as an economic activity (Tafel and Szolnoki 2020). On the other hand, major profit maximisers for wine hospitality and tourism include the presence of a wine bar and shop or a restaurant at the winery, compared to only selling the wine to wine tourists (Alonso 2009). Previous research on winery business models that include wine tourism have specified two general strategies. First, wine tourism can act as a booster for winery branding in the global markets; and second, wine tourism can act as an additional source of income which can help the wine industry become more resilient (Hojman and Hunter-Jones 2012). However, beyond these two quite general types, there is little understanding in the literature on how these types function in detail, and whether other types of winery business model innovations into hospitality and tourism exist (Dressler and Paunovic 2019). Keeping in mind these differences between primary and secondary economic sectors and the risks associated with venturing outside of the wine industry into hospitality and tourism, the following research question has been posed: RQ1: What are innovative models of winery's expansion into hospitality and tourism? Only around 6% of wineries in Italy are equipped for wine hospitality with a designated area for sales and tasting, winery and vineyard tours, English-speaking staff and public restrooms (Colombini 2013). Keeping this in mind, an often-deployed alternative to developing hospitality and tourism in wineries is the development of close co-operation with regional hotels, restaurants and bars (Alonso and Liu 2010). Previous research has identified the link between wine and hospitality as a key to understanding tourism markets (Duhan et al. 2019). However, the only destination actors that can provide an authentic voice to the eno-hospitality and eno-tourism are vintners themselves (Bridge 2017). Therefore, it is important to understand how similar or dissimilar wine production, wine hospitality and wine tourism are as business model elements, thereby leading to the second research question: RQ2: How simmilar/dissimilar are business model elements of wine production and sales, wine hospitality and wine tourism?

Data Collection
The primary data were collected in 2020 from multiple sources, deploying a mixed method approach. The data regarding basic winery data were collected from a wine guide, while the offer design data were obtained from wineries websites. Keeping in mind this combination of primary data sources, the overall methodological approach can be described as mixed method, leaning strongly on netnography. Netnography is an often-used research approach in marketing according to Ahuja and Shakeel (2017). Websites are used as a primary data source method for netnographic research, equally developed as a proven mixed-methods research method, especially for content analysis (Heinonen and Medberg 2018).

Data Analysis and Interpretation
The sample size consists of 886 German wineries. The wineries cover all 13 German wine regions, they are all listed in an established winery guide for Germany (Gault and Millau) and have a functional website. The data for the variables (1) company type, (2) production volume and (3) vineyard size, have been collected from a published wine guide (Gault and Millau). The data for the dummy variables (4) Events, (5) Wine tour, (6) Wine bar and shop, (7) Guest rooms and (8) Cycling and wine have been scrapped from wineries websites. The data were analyzed through two-step clustering and a Principal Component Analysis (PCA) in IBM SPSS (IBM 2019). The objective of the clustering and principal component analysis extraction was to classify different BM innovation types. This approach is rooted in grounded theory, where classification is used as a legitimate methodology for developing scientific concepts (Bischof and Wohlrab-Sahr 2018;Grittmann 2018).

Results for Research Question 1
In the first stage, the data were analyzed through a two-step cluster analysis in order to classify the wineries and create a basic typology of wineries with regard to their basic business model as well as business model innovation into hospitality and tourism. The variables related to tourism (Guest rooms and Cycling and wine) have proven to not fit well into the overall clustering solution, as the cluster quality according to a Silhouette measure of cohesion and separation of 0.2 is rather low. The clustering solution, excluding the two variables, resulted in a Silhouette measure of cohesion and separation of 0.8, supporting a good value of validity with further analyses (see Tables 1 and 2). Table 1. Major extracted factors in the two-step cluster analysis, ranked according to their importance.

Rank
Factor (Indicator) Name Importance Factor Nine different modes of business models regarding hospitality and tourism offerings were identified. Small owner-manager wineries (around 15 ha) are represented by types 1 and 2 and have either no hospitality and tourism offer or a limited number of selected services, most commonly a wine tour, followed by a wine bar and shop. The mid-sized owner-manager wineries (around 18 ha) appear to have a very clear strategy for business model extension into hospitality and tourism. They either organize events or develop a wine bar and shop, but there is no mixed offer for both of these aspects. Owner-manager wineries in a size range of 19 to 21 ha can be distinguished in three distinctive sub-types based on the analyzed business model innovation. The two distinctive types with partial hospitality and tourism offer have a combination of events and wine tours or a combination of events and a wine bar and shop. On the other hand, the larger wineries in this group offer a full hospitality and tourism portfolio, spanning events, wine tours and a wine bar and shop. Furthermore, managerial aspects are important in business model innovation and organization. There are private wineries characterized by non-family management. These wineries, where the ownership and management are separated, belong to the larger entities in the winery population, averaging 53 ha. The majority of these wineries (more than 70%) offer the full spectrum of hospitality and tourism services (events, wine tours, wine bar and shop). Equally, cooperatively organized wineries are distinctive. Cooperative wineries are a legal form of vintners growing their wine as entrepreneurs with a jointly owned wine production facility, as well as joint marketing activities. These wineries represent larger players in the German market in the analyzed population, averaging 294 ha of the jointly organized co-operatives as mother companies. These co-operatives act as nodes for all the value-added or tertiary sector activity within the co-operative arrangement. The vast majority (above 80%) of these wineries offer a full spectrum of hospitality and tourism services including events, wine tours and a wine bar and shop.

Results for Research Question 2
For a further analysis of the business model innovation in German wineries, a PCA has been conducted to understand the major components of BMI in wineries and their major supporting indicators. In a first attempt, the PCA returned the solution with a rather low cumulative variance explained, of 49.8%, which is just under 50%, which is a recommended lower threshold value, as postulated by Sarstedt and Mooi (2014). This solution contained Adm. Sci. 2021, 11, 146 7 of 12 all eight original variables. In the Varimax rotation of the Rotated Component Matrix, only the loadings above 0.5 were retained, as suggested by Sarstedt and Mooi (2014). Based on this decision rule, the variables (7) cycling and wine and (8) guest rooms were excluded from further analysis and the PCA was rerun, rendering a final solution with 6 variables, which fulfilled all the threshold values postulated by the methodological literature. This solution is presented in the following paragraphs.
The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy of the second, improved PCA solution was 0.635, which is significantly above the threshold value of 0.6, postulated by Sarstedt and Mooi (2014), and classified as mediocre adequacy. The approximate chi-square showed results of 1829.65 with 15 degrees of freedom and a statistically significant result at the 0.05 level. The initial eigenvalues for Components 1 and 2 (Table 3) were 2.57 and 1.28, respectively, while the eigenvalues of the rest of the extracted components are below 1. Following the Kaiser criterion, only the two components with Eigenvalues above 1 were retained, as postulated by Sarstedt and Mooi (2014). The two factors cumulatively explain 64.21% of the variance, an acceptable value according to Sarstedt and Mooi (2014), as the extracted factors account for at least 50% of variance. The two-factor solution is supported as an acceptable solution applying the "elbow principle", recommended by Sarstedt and Mooi (2014), where the right number of components is determined by observing the components appearing before the elbow in the components and eigenvalues plot. Since the two-factor solution fulfills the abovementioned standard PCA decisionmaking criteria, a further analysis with Varimax rotations was conducted (Table 4). In the Varimax rotation of the Rotated Component Matrix, only loadings of above 0.5 were retained, as suggested by Sarstedt and Mooi (2014). The two extracted components confirm a clear-cut theoretical distinction between basic wine production BM (Vineyard size, Production volume, Company type) and a BM extension into tourism and hospitality (Wine bar and shop, Wine tour, Events). The component plot in the rotated space (Figure 1) graphically depicts the distance between the winery basic BM and BM extension into tourism and hospitality. This plot delineates the basic winery BM and BM extension into tourism and hospitality.
Wine tour 0.740 Wine bar and shop 0.692 The component plot in the rotated space (Figure 1) graphically depicts the distance between the winery basic BM and BM extension into tourism and hospitality. This plot delineates the basic winery BM and BM extension into tourism and hospitality. The data analysis deployed two different statistical methods, in order to increase the transparency and the reliability of the analysis. The two-step clustering and the PCA analysis present different statistical procedures, as the former provided the mutually exclusive and collectively exhaustive categories and clusters of combinations between wine production and wine hospitality and tourism, whereas the latter analysis extracted the two major components (core wine production BM and hospitality and tourism BMI), thereby confirming the theoretical and practical distinction between these two business models. It appears that both hospitality as well as tourism, are an important step for wineries in Figure 1. The difference between basic BM, BMI into hospitality and BMI into tourism (Component plot in rotated space).
The data analysis deployed two different statistical methods, in order to increase the transparency and the reliability of the analysis. The two-step clustering and the PCA analysis present different statistical procedures, as the former provided the mutually exclusive and collectively exhaustive categories and clusters of combinations between wine production and wine hospitality and tourism, whereas the latter analysis extracted the two major components (core wine production BM and hospitality and tourism BMI), thereby confirming the theoretical and practical distinction between these two business models. It appears that both hospitality as well as tourism, are an important step for wineries in innovating their business model. It is important to notice that both statistical procedures, the two-step clustering as well as the PCA analysis, excluded the variables of Guestrooms and Cycling and wine from the final, stable solutions. It demonstrates that these two variables, representing mobility and tourism, offer components without hospitality elements, represent a more radical business model innovation for wineries from wine production and hospitality, consequently constituting a business model journey rather than a business model extension.

Discussion
Several important theoretical implications for wine hospitality and tourism stem from the research. The presented types of BM-BMI interaction in winery SMEs demonstrate how winery decision-makers act on the two important strategic levers for BMI with key partners and investment decisions to create a winery business model configuration at the nexus of wine, hospitality, and tourism. The results contribute to the existing literature on business model innovation, presented by Cosenz and Bivona (2020). Furthermore, the insights based on the analysis extend the theoretical postulates of core, augmented, and ancillary services in a wine region (Byrd et al. 2016;Dressler and Paunovic 2019). The presented clustering solution groups wine tours, wine bar and shop, and events into non-core winery business model elements, thereby providing a winery-oriented classification, extending the wine industry into hospitality and tourism services. This novel classification is important for understanding the evolution of the winery business model into eno-hospitality and eno-tourism. It is noteworthy that the organizational perspective is quite different from the destination level perspective, which has been researched in the previous literature, and where events are part of ancillary wine destination services. In the present model, events are more relevant for the winery business model innovation into hospitality and tourism. The results of applying a two-step clustering method on the netnographic data collected from winery websites demonstrate that this method can be successfully applied to a variety of content analyses. The major advantage of this clustering method is that it works well with both categorical (binary and non-binary) and continuous data (IBM 2020), thereby allowing for higher flexibility in content analysis.
Previous research identified two basic and quite general strategies for wineries in wine tourism, namely, wine tourism for winery branding and wine tourism as an additional source of income (Hojman and Hunter-Jones 2012). The presented research extends this dichotomy on winery business models at the intersection of wine and tourism using a broader typology with innovation profiles in relation to basic wine production as well as hospitality and tourism offer elements.
The identified configurations of winery business model extensions safeguard entrepreneurial decision-making regarding a business model transition from a primary sector focused activity (grape and wine production) into a tertiary sector activity (hospitality and tourism services). Extracted types of business model extensions illustrate the diversity of business model change, and confirm the theoretical postulates that substantial as well as the less radical change as exploratory phases are relevant and important for conceptualizing and road mapping change in companies (Foss and Saebi 2017;Sosna et al. 2010).
The derived typology offers practical orientation. Observed configurations of winery business model extensions into hospitality and tourism, with its characteristics can serve as a useful tool for winery managers in managing the complexity at the nexus between wine, hospitality, and tourism. The framework created therefore presents a roadmap for navigating business model innovation and change on a continuous basis, confirming the high volatility of business models in the hospitality and tourism industries, as identified by Barth (2011). It also serves the purpose of sensitizing winery managers to the difference between the economic logic of the primary economic sector and the tertiary economic sector and the need to develop appropriate skills for each of the two sectors, identified as obstacles for developing wine tourism (Bridge 2017;Tafel and Szolnoki 2020).

Conclusions
Hospitality and tourism provide means for an extension of the winery business model, while accommodation and cycling represent winery BM migration, differing significantly from the original wine production model. A transition of the business model of wine production and sales into hospitality and tourism is a complex process which includes several important and interconnected aspects. It represents a transition from the primary sector to the tertiary economic sector. The rules of building a successful business model in the tertiary sector are quite different to those in the primary economic sector and require a different skill set from the wine entrepreneur, especially regarding the recurring need for business modelling and remodeling. It appears that the transition into niche tourism products is the next level of the transition into tertiary sector, which needs to be researched in future studies, since it is bound to a different set of rules than BM extension into classic wine hospitality and tourism. In this sense, future research should expand the knowledge on winery BM extensions into wine and tourism by providing case studies, either as multicase studies or with longitudinal case study approaches. These methodological approaches should help to obtain more in-depth insights into these entrepreneurial processes of transitioning from the primary sector to the tertiary economic sector for a single business. In addition, diverse examples from both "old" and "new" wine world countries are needed in order to take into account different economic and innovation systems in which wineries in different wine countries operate.
The theoretical contributions of the research primarily relate to the winery-and vintner-centric explanation of wine hospitality and tourism phenomenon by deploying the business model innovation approach. The research results demonstrate the richness of the strategic choices in wineries transitioning from wine production to wine hospitality and tourism, beyond the two previously identified strategic options. Furthermore, it also contributes to the literature on business model innovation and servitization, by presenting the multitude of modalities for business model extensions from a primary economic sector activity to a tertiary sector activity. The study results have clarified how size and therefore, the investment and reinvestment potential, influence the ability to create meaningful partnerships with stakeholders in the tertiary sector.
A research limitation to be taken into consideration is that the presented research has not been explored on the basis of cluster theory (close cooperation of the SME wineries with regional and non-regional hospitality and tourism actors). This offers potential for further research to explore cluster-oriented dynamic capabilities regarding the development of wine tourism and hospitality, while allowing businesses to maintain authentic wine production. This research stream seems of relevance for wine hospitality and tourism but has only been partially considered in the literature (Alonso and Liu 2010;Velikova et al. 2019). Furthermore, the method deployed can be developed, as every clustering and typology building method necessacitates simplifications in order to create mutually exclusive and collectively exhaustive categories. The complexity of each case of BMI in hospitality and tourism should be considered and the created clustering solutions serve as a framework. This research has relied on a sample of wineries published in a wine guide, representing an elite selection of wine producers in Germany, limiting the generalizability of the findings.
The practical contributions of the presented research mainly relate to sensitizing winery owners and managers to the business modelling approach, and regarding BM extension into hospitality and tourism as these economic sectors are dependent on experimenting and iterative business model adaptation, depending on the economic cycle. It also provides practitioners with a tool for a detailed analysis of suitable strategic options depending on the winery size and organizational type.