New business models in online hotel distribution: emerging private sales versus leading IDS

This article deals with online hotel distribution. Its purpose is to investigate the private sales formula, which has recently emerged as a new business model. Two case studies are presented: Expedia, a leading internet distribution system (IDS), and Voyage Privé, a private sales Website. The study highlights the main features of the private sales formula by means of a cross-case comparison between the two travel Websites. The article ends with some final remarks on the main strengths and weaknesses of the formula as compared to the leading IDS in the online travel market.


2011;
). This shift to online channels has also influenced tourists' buying behaviours (Schlosser et al. 2006), with the Internet being more and more the preferred means to seek travel information or make reservations. Considering this latter in the case of Europe-the world's most frequent destination continent-on average, 33 % of gross bookings revenues from travel products in 2010 came from online transactions, with online booking penetration varying widely from 21 (Italy) to 48 % (Scandinavian countries) (PhocusWright 2011a). In the USA, the percentage of travel booked online currently stands at about 39 % and will inch upwards to just 40 % in 2013 (PhocusWright 2011b).
Within the more general field of hospitality e-commerce, online hotel distribution, including reservation systems (hotel Websites) or third-party portals, which are broadly identified via the terms internet distribution systems (IDSs) or online travel agencies (OTAs), have emerged as the major players. However, third-party travel Websites gained dominance since the very early stages of the development of online distribution, while hotels were slower and less efficient in using the new online resources (Gazzoli et al. 2008), thereby losing much of their control over this channel (Phelan et al. 2011), for instance, in terms of price integrity and brand equity. Many new players specialized in Internet marketing of travel and related services emerged and rapidly gained leadership positions in the global competitive scenario, with increasing market shares in hotel distribution (Buhalis 2003;Bender et al. 2008). In Europe, for example, where on average 25 % of hotel bookings are made online, unlike any other European travel segment (such as airline tickets), hotel bookings made through IDSs exceed transactions via supplier Websites. Considering total online bookings, OTAs account for 63 % of online hotel sales, while hotel Websites account for only 37 % (PhocusWright 2011a). The combined market share of the top five leading IDS players in Europe amount to roughly 50 % of the total market share of IDSs. The remaining 50 % is held by a number of midand small-sized online travel agencies, mainly operating nationally. However, if we look at the shares amongst these five big players in 2010, Priceline Europe (which operates in Europe as Booking.com) held 40 % of the market, Expedia Europe (which operates in Europe through Expedia.com and Venere.com) 30 %, Travelocity Europe (Lastminute.com in Europe) 10 %, Opodo (Opodo.com) 10 %, and Orbitz WW Europe (which operates through Orbitz.com and Ebookers.com) 9 % (PhocusWright 2010). Moreover, considering that roughly 75 % of hotel inventory in Europe is still booked offline, and that the number of tourists booking online is expected to grow in coming years, the opportunities for IDS development are significant.
IDSs garner consumer preferences thanks to their competitive pricing and the possibility of easily comparing the prices of different hotels at thousands of destinations worldwide (Schlosser et al. 2006). At the same time, IDSs have gained favour with hotels because of their ability to globally market a highly perishable product (O'Connor and Frew 2004), namely, their room inventories.
These new players adopt different business models. Leading IDSs, such as Expedia (with Expedia.com, Hotwire.com, Hotel.com and Venere.com), which is the world's largest IDS, and Priceline (with Booking.com and Agoda.com), which is Expedia's main rival worldwide (PhocusWright 2011a), have received attention in some studies (Gazzoli et al. 2008; Barsky and Nash 2007;Schütze 2008). However, given the rapid growth of travel Websites, there is a need for a greater understanding of the specificities and recent evolution of such business models (Law et al. 2010).
It is to such an understanding that this article aims to contribute by building on previous study in the area of online distribution. In particular, this study attempts to shed light on the emergence of a new formula within the IDS, the so-called 'private sales formula'. Private sales in this context can be broadly considered as a community of select customers to whom an online travel agency proposes discounted hotel rates. Private sales Websites allow tourists to obtain cheap rates on select quality hotels. For hotels, the private sales formula represents a means to implement pricing disparities on the Web, while at the same time maintaining price integrity and rate parity, that is, setting the same rate structure across all distribution channels (Gazzoli et al. 2008). The success of the private sales formula is based on its competitive prices and ad hoc proposals, which are assumed to be key motivators encouraging tourists to buy online (Schütze 2008) and enablers of relationships with both consumers and hotels.
In order to shed light on the specificities of this formula, we propose a comparison with the leading IDS. In particular, this article aims to answer the following research questions: (a) Why do private sales firms exist in the online travel market? (b) What are the main characteristics of the private sales formula? and (c) What are the main differences with respect to leading IDS models?
In addressing these research questions, the article is structured as follows. Theoretical contributions on business models and online business models in tourism are presented in the first section. An empirical investigation is then presented and discussed. The article is based on a case analysis of two leading travel Websites: Expedia, the leading global IDS in online hotel distribution, and Voyage Privé, a leading travel Website in the emerging field of private sales. The two cases have been analysed in depth to highlight the main features of their specific business models. The article ends with a cross-case discussion, managerial implications and future research.
2 The online distribution of hotels: a brief overview of current business models

The concept of business model
Drawing on theory and empirical research in management studies, a business model can be broadly defined according to Teece (2010) as 'how a firm delivers value to customers and converts payment into profits'. The concept of business model has received renewed attention in recent years (Baden-Fuller and Morgan 2010), though it has in the past been used in a general and a broad sense since the seminal studies of Peter Drucker (1954Drucker ( , 1973. Although no common definitions can be found, scholars agree on the usefulness of the concept for management studies (Zott et al. 2010;Montoro-Sánchez 2009).
Over the last decade, new and innovative business models have often been inextricably associated with the onset of the Internet (Amit and Zott 2001). As New business models in online hotel distribution 185 Magretta stated provocatively (2002), with the advent of the Internet, a Web-based firm was considered to be the only thing a company needed to achieve success. Neither competence nor customers were needed to make a profit. Obviously, this was not the case, as the failure of many dotcoms demonstrated, and the search for a 'good' business model remains a key component of an organization's success (Perkmann and Spicer 2010). Moreover, innovation is increasingly viewed as relevant to the development of new business models, as well as new management practices, and there is now a greater recognition that novel ideas can transform any part of the value chain and that products and services represent only the tip of the innovation iceberg (Johnson 2010). A business model is like the telling of a story, whereby a new business model can be interpreted as variations on the story (Magretta 2002). This means that a new business model may represent a reworking of the general value chain of all businesses. Some academic contributions have attempted to identify the components of business models. Just as it can be stated that there is no such thing as a common definition of business models (Zott et al. 2010), at the same time it is also true that authors generally disagree regarding the components making up a business model: some such components are common to a number of conceptions, while others appear in only one definition. Shafer et al. (2005) propose a review of the existing literature, focusing in particular on the components of business models identified by several scholars. The authors highlight how the business model components proposed in the literature can be grouped into four primary categories: strategic choices, the value network, creating value and capturing value. Similarly, Morris et al. (2005) propose a six component framework, defining business models by answering the following questions: How will the firm create value? For whom will the firm create value? What is the firm's internal source of advantage? How will the firm position itself in the marketplace? How will the firm make money? What are the entrepreneur's time, scope and size ambitions? Thus, all definitions seem to converge on an analysis of the components of the business model by looking mainly at the value proposition, the way this value is produced and delivered to customers, and the revenue model (Doganova and Eyquem-Renault 2009).

Leading IDS in online hotel distribution
The range of business models adopted in Internet-based tourism is extremely varied and thus seems chaotic at first glance. In fact, the Internet has essentially transformed how tourism retailers compete in the marketplace (Varadarajana et al. 2010). In an attempt to build a classification of online business models, Järvelä et al. (1999) refer to three main aspects: the type of services delivered, the degree of innovation in adopting new technologies and, finally, the revenue model. Given these aspects, four main business models can be identified: E-Booking Services, E-Travel Agent, E-marketplaces and Meta Booking Engines (Järvelä et al. 1999).
E-Booking Services are e-mediaries operating through a 'booking engine' that enables travellers to make reservations guaranteed by credit card, but does not allow them to conclude the transaction online-payment will be due at check-out time at the hotel. The revenue model adopted is commission based. Such services can be both multichannel-provided through a combination of both on-and offline transactions (i.e. Tui.com, Alpitourworld.com)-or pure player, delivering services exclusively in the online environment (i.e. Booking.com, Hrs.com, Laterooms.com and Venere.com). E-Travel Agents are true multi-channel or pure player services that integrate their booking engines with e-commerce platforms, thereby enabling customers to conclude their transactions by charging a credit card. Different revenue models are encountered in practice: both commission-based and merchant (mark-up based). E-Marketplace is the business model that characterizes the leading IDSs. These horizontal operators can offer a wide range of tourist services, usually also offering options for dynamic packaging (i.e. the ability to build a tailored travel package). As before, here we find multichannel (e.g. Octopustravel.com, the IDS of Gulliver's Travel Associates) and pure players (such as Expedia.com, Lastminute.com).
Meta Booking Engines include the latest forms of online brokerage. These Websites support the buying process with functionalities such as price comparison and reviews systems. User-generated content (UGC) takes on a good deal of importance within the Metabooking Engines, which include tourism blogs or Websites such as tripadvisor.com and travelpod.com (Akehurst 2009). In recent years, such sites have seen exponential growth, mainly thanks to the possibility for tourists to share information, opinions and recommendations about hotels and other tourism services. Moreover, tourism organizations can achieve greater numbers of reservations from the often high click-through rates (Akehurst 2009). One further recent evolution of some online price comparison sites is to include price charts (line charts) showing a product's (i.e. hotel room) full price history to help consumers forecast future prices (Drechsel and Natter 2011).
It is possible to observe that the merchant revenue model is the most frequently used in the online travel market (Grau 2006). For instance, looking at Europe's online booking services, of the five big players (who as already mentioned account for 50 % of online hotel bookings), four, namely Expedia.com (Expedia Europe), Lastminute.com (Travelocity Europe), Opodo.com and Orbitz.com (Orbitz WW Europe), are all IDSs adopting e-marketplace type business models with merchant revenue models. Only Booking.com (Priceline Europe) has adopted an e-booking service business model with commissions-based revenue (PhocusWright 2011a).

The emergence of the private sales formula
Focusing our attention now on the revenue model, it is interesting to note that it is also the one that has exhibited the most interesting evolutionary dynamics (Talluri and van Ryzin 2004), in that it has shaped alternative classifications of business models, mainly within merchant players. With particular regard to the revenue models adopted, at least four basic types have appeared: transparent sales, last minute/inventory control sales, opaque selling, and flash sales/collective buying selling. Transparent selling (Jerat et al. 2010) refers to traditional sites in which all the attributes of a specific item, including of course the regular price, are displayed up front (i.e. Expedia.com). In this way, customers can perceive all the differences between the various hotel room rates offered by a given retailer and is moreover a position to easily compare prices offered by other retailers (online or offline) or by the same hotel itself. Last minute pricing is the revenue model adopted by retailers that try to sell distressed inventory that is offered late in a product's sales life cycle at a significant discount some days before it expires (e.g. Lastminute.com was started precisely to provide this kind of service). However, this particular kind of model poses a significant problem to hotels, as it often leads potential consumers to expect that there may be a much lower-priced deal available at the last moment, thus promoting consumer strategies that can eventually lead to reductions in hotel profitability (Jerat et al. 2010). The inventory control sales is a different strategy that-in contrast to last minute sales-aims to reduce distressed inventory by offering a limited number of low-priced allocations far in advance of check-in time, and reserving the remaining allocation for late arrivals at a higher rate (Jiang 2007). This method also presents a serious drawback: the risk of not exhausting all the high-fare allocations close to the consumption period (Saliba 2000). A number of different solutions have recently been adopted by retail players in order to avoid the issues related, on the one hand, to hotels' need to reduce customers' inclination to price comparison and, on the other, consumers' enacting strategies behaviours and/ or the risk of residual distressed inventory. The first is represented by so-called opaque selling (e.g. Hotwire.com, Priceline.com and Cheaptickets.com). In this model, consumers, before purchasing the 'opaque' product from an intermediary, does not know some of its attributes, for instance, which company will ultimately provide the service or when exactly the service will start (Jerat et al. 2010). Contrary to the transparent/last minute/inventory control sales, here the intermediaries keep some attributes of the offer hidden from consumers, revealing them only after the purchase. This mechanism represents a more efficient alternative to solving the problem of distressed inventory. It offers four main advantages. First of all, the system makes it possible to generate incremental revenue by selling distressed inventory cheaply without disrupting existing distribution channels or retail pricing structures (Wang et al. 2006). Secondly, it allows hotels to maintain an efficient market segmentation based on customers' different sensitivities regarding price and service level (Jiang 2007). Third, it improves profits, as the hotel can benefit from the fencing-off between segments, making it possible to simultaneously reduce distressed inventory and gain revenue from last-minute sales in the high-value segment. Finally, it avoids both consumer strategic behaviour and the risk of residual distressed inventory. The formula does, however, suffer from some limitations. On the one hand, it involves the risk of commoditizing hotel services, as it erases product differentiation. In fact, the customer chooses what to buy based on price alone (Shapiro and Shi 2008). On the other hand, the hotels run the risk of losing control of intermediaries by fostering a sense of store loyalty in customers by offering products that are guaranteed by intermediaries as if they were private labels.
The recent rapid evolution of this opaque selling model is exemplified by the growth of so-called private sales or flash sales or collective buying selling (i.e. Groupon.com, Livingsocial.com, Guilt.com, Voyageprive.com, Letsbonus.com) (Zmuda and Patel 2010). This model is based on virtual communities, which are emerging as an important marketing channel for a variety of industries (Kim et al. 2009), and offer significant benefits to such communities' members. In fact, the main characteristic of this model is that the discounted 'opaque' prices are visible and can be purchased for a very limited period of time only by members that have previously registered with the service. Owing to the restricted availability to members and the confidentiality of the proposals, this model fits into the trend of the broader phenomenon of 'social commerce' (The Economist 2010), which is in turn related to the currently large and widespread demand of online users to participate in social networking sites (e.g. Facebook.com, Twitter.com). The sale of hotel products in this context can take place first, rather than in travel sites, in horizontally structured 'social commerce' sites, in which the offerings of flash deals regard a wide range of products and services (from restaurants to fashion, and from personal care to health). Of the general social commerce Websites, Groupon.com, which in 2010 rejected Google's $6 billion acquisition bid, is one of the fastest growing Internet sales businesses in recent years, with tens of millions of registered users and sales expected to top 1 billion dollars in 2011 (Byers et al. 2011). Alternatively, hotel offerings can be spread via vertical social commerce sites that are proper 'travel' sites in that the flash deals exclusively regard tourism products. Of these, the French group Voyageprive.com, with over 8 million members worldwide and 2.7 million distinct consumers per month, represents one of Europe's leaders, while Jetsetter.com is the leading site on the American market (PhocusWright 2010).
The main characteristics of the private sales formula is evident on both the hotel side and in the IDS itself (The Economist 2010). For the IDS, this approach is much more efficient with respect to the traditional revenue models, as it reduces costs by attracting visitors through the 'viral' spread of registered users, boosted by personal invitations. It also improves conversion, given that belonging to a community or private shopping club cultivates more willingness in customers to buy. Moreover, it also seems to lead to significant improvements in loyalty. From the hotels' point of view, it appears that, apart from the enhanced possibility of selling distressed inventory and benefiting from the so-called billboard effect (Anderson 2009), using flash-sales intermediaries makes it possible to achieve greater exposure specifically to the youngest market segment, who rarely buy off-line (Bates 2010).

Research method and case studies
Our aim is to analyse various aspects of hotel distribution business models in depth and thereby gather some insight into the private sales formula. Considering the exploratory nature of our research, we employed the case study as the research method.
The case study method represents one of the most commonly used qualitative methods for carrying out research in management studies. According to Easton (2010), the case study is a 'research method that involves investigating one or a small number of social entities or situations about which data are collected using multiple sources of data and developing a holistic description through an iterative research process' (p. 119). By using case study methods, theory is emergent, in the sense that it is situated in, and developed by recognizing, patterns of relations amongst constructs within and across cases (Eisenhardt and Graebner 2007).
Robert Yin (1994), who along with Kathleen Eisenhardt (1989), can be considered one of the most influential contributors to the methodological aspects of the case study in management studies (Piekkari et al. 2009), defines case studies as 'research situations where the number of variables of interest far outstrips the number of datapoints' (Yin 1994, p. 13). Consequently, one of the main reasons behind the widespread use of the case study method in management research lies in enabling the researcher to study a phenomenon in a real-life setting, where often it would be otherwise difficult to grasp its dimensions (Yin 1994;Stake 1995;Gibbert and Ruigrok 2010). A general case for resorting to case studies in management research can moreover be advanced, as it has been claimed that articles building theory from case studies are amongst the most highly cited (Eisenhardt and Graebner 2007).
On the other hand, case studies have been heavily criticized as well. In particular, it has been claimed that they do not allow 'researchers to effectively verify the effects of those hypothesized advantages' (Levitas and Chi 2002, p. 958). Additionally, scholars have criticized case studies in terms of validity and reliability, as they do not allow for generalization, as do instead quantitative methods (Gummesson 2006)). Although the case study method has certain limitations (as does any research methodology), management scholars generally agree on the rationale behind why such research lines can be followed. Barratt et al. 2011, for instance, point out that the use of case studies can be justified as a research approach in the following situations: when existing theory suffers from gaps in explaining the phenomenon under investigation; when we are faced with questions of 'how' and 'why' in situations where the context and the experience of actors are critical; and lastly when the research is exploratory.
Following such reasoning, the case study research approach was considered suitable to investigate the current topic of new business models in online hotel distribution. Moreover, in order to better highlight the specific features of new online business models, a cross-case analysis has been conducted to compare and contrast the two selected cases, specifically, the two travel Websites: Expedia.com and Voyageprive.com (Eisenhardt 1989;Yin 1994;Eisenhardt and Graebner 2007). To this end, we first present a case analysis of each and then proceed to a cross-case comparison to outline the main features of the private sale formula as contrasted to the leading IDS.
The choice of the cases was dictated by the standing of the two online systems, both of which can be considered emblematic of the article's subjects of investigation. Expedia.com (part of the Expedia Inc. Group) is the world's leading IDS. It holds a commanding position of leadership in both the American and European markets, and is as a whole representative of a business model in widespread use amongst the most important IDSs in the online market. Voyageprive.com is instead a leader at the European level in the sale of hotel products through a specialized social commerce approach, and is thus representative of similar players that are finding fertile terrain in the online market, albeit with far less bookings in comparison to the leading IDSs.
To ensure data robustness and hence sound propositions, following the aims of 'triangulation' (Eisenhardt 1989;Barratt et al. 2011), the cases presented have been investigated using data collected from different sources. Semi-structured interviews and archive sources such as documents, historical records, organizational charts and statistics have been used as data sources.
Regarding the semi-structured interviews, in particular, they were organized in such a way as to collect a large amount and variety of information on the following topics: (a) enterprise profile elements: the main stages of its historical development, entrepreneurial model and organizational structure; billing records and end market; its revenue model and strategic development paths; (b) channel relationships with hotel facilities: guidelines followed in defining the hotel assortment, price positioning, variety and types of accommodation facilities in their assortment; stages of the relationship with accommodation facilities, from defining the contract to implementation of the relationship; (c) consumer-tourist profile: segmentation of the clientele and major dynamics of the segments; management of relationships with tourists; (d) competitive dynamics: main competitors and the bases for competitive advantage; marketing strategies and policies adopted by the competition.
The semi-structured interviews were conducted with representatives of the two travel sites in Italy: for Expedia, the Market Management Director for Italy (responsible for central and southern Italy), while Voyage Privé was represented by the Managing Director (country manager for Italy). For both, the survey involved preliminary contact to introduce the research and its aims (both face-to-face and by telephone), while later surveys included semi-structured, face-to-face interviews. The interviews were conducted at the Italian branch offices of the two companies, both located in Milan. The interviews were recorded to ensure robustness of the research procedure. Table 1 illustrates the methodology followed.
The subject and aims of the empirical study were discussed in the semi-structured interviews as regarding the general travel market (in terms of both geographical markets and evolutionary trends), with the aim of highlighting the main  2008). Therefore, the Italian hotel sector is extremely fragmented: apart from a few medium or small-sized national chains, it is made up mostly of many small-sized independent facilities of which many are family managed. Finally, the need to consider the Italian case, as the main geographical setting for further empirical investigation, stems from the aim of studying the internationalization of the two players in question. In fact, apart from being one of the main tourist destinations in the world, albeit with rather particular characteristics, it was thought that close examination of the situation in Italy would enable highlighting further differences between the two companies' approaches to expansion, given that Expedia, an American company, adopts standardized approaches when expanding abroad, whereas Voyage Privé, a French company, proposes ad hoc penetration into each country it enters. Thus, for both, Italy represents a very fertile, fast-growing tourist market whose importance and peculiarities, make it worthy of particular attention.
3.1 Expedia: the leading IDS

The company
Expedia Inc. is a well-known American company founded in 1996 as a Microsoft initiative for the online sales of airline tickets. It essentially aimed to offer consumers global distribution system (GDS) assortments via the Internet, at the very dawn of its potential. Over the years the product portfolio sold through the site has expanded to include various other tourism services. During the last decade, the company changed hands a number of times. After its transfer from Microsoft to Interactive Corporation in 2004, Expedia is now an independent public company quoted on the Nasdaq in New York. The Expedia Group manages several brands in the online travel industry, amongst which Expedia.com, Hotels.com, Hotwire.com, Venere.com, The Tripadvisor, and Classic Vacations.
With particular reference to online hotel distribution, Expedia.com and Hotels.com, represent the leading brands within the group's portfolio. On Expedia.com, the hotel offering is marketed within a wider variety of tourism services (such as airline tickets, car rentals, resorts, etc.), while Hotels.com offers an assortment composed solely of hotels. Moreover, Expedia.com is considered to be the best-known brand in the final markets within the corporate portfolio, while Hotels.com, has a lower degree of brand awareness (except in the US), though it is more developed internationally. Sales from hotel services represent roughly 60 % of corporate revenue; 15 % is generated from online airline ticket sales and 10 % comes from communication of other actors inside the group's corporate brands sites (Expedia Inc. Annual Report 2010). Expedia.com adopts a merchant model with transparent sales formula.
From an organizational perspective, the company's headquarters is in the US (Seattle), while in Europe the main base is its London office. At the European level, they also maintain a local presence in those markets with the greatest potential. Thus, the structure adopted is one that seeks to reconcile the requirements of standardization at the central level with the need for adaptation to local circumstances in order to meet specific consumer needs. One example is the case of Expedia's Italian subsidiary, which is responsible for managing purchasing and its relationships with their various tourism product suppliers. The marketing department is instead based in the London office, which oversees all European markets.

The value proposition
It is interesting to note how Expedia formulates the hotel inventory proposed on its site. Expedia.com aims to include in its assortment all types of hotels in all potential destinations because the target segments are all potential consumers seeking hotel rooms worldwide. However, as the manager interviewed pointed out, 'there are priorities, as we tend to include hotels in those geographical markets and destinations which are considered to be more attractive'. Consequently, prime destinations are covered quite extensively, while secondary ones are present to a somewhat lesser degree. Several potential sources are used for selecting hotels to include in the inventory. For hotels that contact Expedia directly through joinexpedia.com, the results of an analysis based on their structure, characteristics and location, are matched with the potential demand and the current inventory that Expedia has in that destination (how many people are looking for a hotel, how many hotels they have in that destination, etc.). Then the applications are considered by market managers, who play a crucial role as they are responsible for hotel contracting and relationships in specific countries (such as Italy) or specific regions within the country (such as central and southern Italy). In particular, based on qualitative judgments regarding the market and its specificities in terms of potential performance, the market manager decides whether any further hotels could be included in the inventory. The final objective is to maximize sales, so the occupancy rate of the specific market (in terms of numbers of rooms) is considered as a key measure. Regarding the types of hotel, Expedia includes all positioning levels from low-medium to high rates and all ownership types, including multinational and national chains, as well as independent hotels. Obviously, continues the manager interviewed: 'there are differences among countries', as for example the case of Dubai, where the great majority of the assortment is composed of chains, in contrast to Italy, where most of the hotels are independent. In general, the relationships with hotels are stable and based on a general contract. The contract contains details on product pricing, special offers and many other conditions. When signing the contract, which is standard, the hotel is automatically available to consumers in all Expedia country sites and other corporate brand sites. Clearly, given that the hotel room volume generated by Online Travel Agencies is increasing (in 2010, over 79 million gross room nights were booked on Expedia's site portfolio), Expedia's bargaining power is extremely high. However, the more a given hotel represents a strategic partner at a given destination, the more cooperative the relationship with that hotel will be. The relationships with hotels are mediated by technologies, as Expedia's connectivity solutions are tailored to the needs of global chains or independent hotels and regional chains and can provide connectivity to all hotel systems, whether they are a computer reservation systems, property management systems or a channel manager. Then, the more strategic the hotel is, the more the market manager gets involved personally in managing the relationship. The hotel assortment is the same for all Expedia sites the world over (Expedia.co.uk, Expedia.fr, Expedia.de, Expedia.es, Expedia.it, etc.), though it is adapted to the demands of each country's markets. Take, for example, a nationwide preference for certain 'sun and beach' destinations: if Greece is a typical such destination for Italians, it will be more valorised on the Expedia.it site.

The role of listing and opaque selling
Looking specifically on how the hotel inventory is proposed on the site, it is clear that the main aspect is related to the listing. The hotel listing on the site actually functions just like a shop window. Expedia sets up its listings using a weighted mathematical formula that combines many factors, amongst which hotel capacity and sales potential. Moreover, the listing also depends on the opinions-hotel reviews created by Expedia travellers when using an accommodation. It is moreover dynamic, as it takes into account who is looking for a room (the country of origin), the specific query entered (e.g. the number of nights) and the hotel's promotional policies. Regarding this latter point, take for example a hotel with a medium price positioning that offers a room for one night at a price higher than the average for similar hotels in that market. Expedia's algorithm will rank this hotel very low in the listing that displays when a user searches for potential hotels at that destination for one night. However, if that same hotel offers a lower promotional rate for three nights, it will advance in the listing displayed on a search for three nights. Consequently, an essential role in adapting Expedia's inventory to consumer needs is played by technologies that enable the firm to match user characteristics (IP address, query, language, etc.) with hotel characteristics (sales potential, marketing policies, etc.). This is because Expedia covers all possible types of targets. In this regard, an important role is also played by Web statistics and metrics, which can be used to better manage the assortment and maximize profits. Hotels can moreover purchase advertising on the site, for instance, by creating sponsored listings that appear in the search results or that simply seek to engage targeted consumers. Search engine optimization strategies are also implemented to increase brand awareness and ultimately drive tourists to the site. Moreover, in some countries, Expedia invests in offline local advertising, such as television commercials and billboard campaigns to increase overall brand recognition.
A real novelty in Expedia's merchant revenue model is the practice of opaque selling. Expedia.com began integrating sister company Hotwire's opaque hotel offerings (discounted inventory) within Expedia's hotel-booking path in some markets. The opacity is aimed at price-sensitive, rather than location-sensitive tourists. In essence, on Expedia customers can book a hotel room without knowing which company will ultimately provide the service; they know only the discounted rate, the star rating and the general location of the hotel. According to the managers interviewed, this offers advantages to both the hotels and the tourists. Tourists can find cut-rate accommodations, while hotels can make some very aggressive offers, without, however, running any risk of impacting pricing policies on the Web, where increasing number of big players are asking for rate parity. It thus represents an Expedia strategy to differentiate their offerings in consideration of the growing competitive pressures.

The company
Voyage Privé was founded in France in 2005 with the legal structure of an SAS (Société par actions simplifiée, corresponding to a Limited Liability Company in American law). It is the initiative of six members with experience in online tourism, amongst whom Denis Philipon, former CEO of Lastminute France, played the major role. It was he who first understood the potential of applying the private sales formula to the travel sector. In fact, this business model is an absolute novelty in online travel distribution, as sales revolve around a portfolio of club subscribers, to whom a selection of new tourism products is proposed each week (about 20-40 depending on season). The offers are announced by means of a newsletter at a special discounted price (30-70 % below list price, at least 20 % lower than the best published rate). The promotions must be purchased within 5 days of publication (called a flash sale) and their departure dates are limited to the following three to 6 months. The revenue model used corresponds to the merchant model based on net rates from suppliers. Currently, Voyage Privé, with Voyageprive.com, is the market leader in this particular segment, with a consolidated group turnover that reached 212 million Euro in 2010 and has risen by over 100 % each year over the past 4 years. Voyage Privé boasts over 8 million members, of which most (about 4 million) are French with higher conversion rates (up to 7-8 times) than the average for the leading IDSs.
From an organizational point of view, the group's headquarters is located in Aix En Provence, France, where the technology team also performs various functions, such as analyses on demand, Web marketing and management of the entire group's sites. The group also have four international offices in Spain (Barcelona), United Kingdom (London), USA (San Francisco) and Italy (Milan) and new openings in Germany and Poland are planned for 2011. All national offices have their own legal autonomy, while, however, remaining fully controlled by the French parent company. The group's had a total of 150 employees in 2010.
From the strategic point of view, Voyage Privé seems to follow a growth strategy of focalizing on demand segments with distinctive features as compared to the general market. On this point, the management interviewed emphasized an important difference between Voyage Privé's strategy with respect to that of leading IDSs. In particular, he pointed out that leading IDSs tend to avoid a segmentation approach, with regard to both hotels and vacation packages. With regards to the latter, mass market operators have reportedly been buying such packages from the same consolidator-wholesaler (i.e. Easymarket by TUI), thus providing absolutely undifferentiated products. Actually, the leading IDSs have also tried to launch private sales initiatives, though, in the management's view, such attempts have led to great difficulties because of the big players' inflexibility in both organizational structure and strategic choices. In this sense, Voyage Privé's competitors are not represented, at least at this stage, by the main players in the market, such as Last minute and Expedia, but by those operators that have adopted a similar business model. These include Italy's Saldiprivati.com, America's Groupon.com and Kgbdeals.com, and Spain's Groupalia.com, even if these are all characterized by a wide-ranging assortment that often excludes travel products, though they can offer tourism services directly or by means of partnerships with tourism retailers. Amongst Voyage Privé's direct competitors in tourism, the manager cited Italy's Hotelyo.com, Spain's Letsbonus.com and the American Jetsetter.com.

The value proposition
It is interesting to observe how Voyage Privé formulates its hotel inventory. Its make-up is not stable, but undergoes continuous weekly changes that are based on both economic opportunities and analyses of consumer segments. In particular, the assortment proposed can be essentially described according to a specialized approach limited to only two product lines: hotel rooms and vacation packages. Particular attention is devoted to the quality of the deals included in the assortment. For example, with regard to hotels, only four-and five-star facilities are chosen, while travel packages include only high-end positioned packages. In both the cases, in fact, Voyage Privé aims to offer an 'aspirational' proposal to its members. Apart from the qualitative aspects, another element can influence the inclusion of an offer in the proposed hotel inventory-the discount offered by the hotel to the final customer. In fact, before deciding upon final adoption of any item, Voyage Privé subjects the associated discount level to rigorous analysis through comparisons with published rates in order to estimate its real amount.
With regards to the inventory make-up, it should be emphasized that the site displays all hotels with their own brands, while the travel packages are essentially presented in a 'blind' fashion. All hotels contact the head office directly, and contract managers in the national markets act on behalf of the group, but never of the national office, through an agreement that provides for recognition of net rates and a minimum of allocations. Although bargaining is centralized, in order that specific deals be included in the promotional set (the so-called flash sales), they are adapted to the characteristics of the individual countries. In particular, the choice of a specific facility to include, the time of inclusion in the set, selling prices, travel dates and ancillary services are decided in full autonomy by the individual national offices. Another specific aspect of the formula is represented by the fact that hotels contracted in a certain destination are always limited in number and, in any event, will never be marketed together in order to avoid a 'cannibalism' effect between similar products in the set. Such behaviour is particularly advantageous for hotels, which as a result, can in fact count on exclusive exposure each time they are placed in the set. In any event, the quality of the facilities in the inventory is carefully monitored by analysing both the opinions on major social networks and the assessments contained in their Website review system. On the hotels' part, the relationship is currently characterized by keen interest in the new distribution formula, which is perceived as an attractive alternative for diversification of risk, in contrast to the substantial oligopoly represented by the big players. This point was emphasized by management as the factor that led to rapid penetration into international markets, such as the Italy, in spite of the brand being as yet littleknown. For hotels the main advantages offered by Voyage Privé over the leading IDSs seem to be related to the exclusive visibility afforded their facilities during a promotion's validity period. According to the managers interviewed, Voyage Privé also ensures hotels a quality positioning, while at the same time keeping hotels rates opaque.

The role of club members
It is particularly interesting to note the distinction between club members and registered site users, a fact emphasized by the Voyage Privé management. The first category refers to individuals who visit frequently and carry out repeat purchases, while individuals in the second rarely go to the site after registering and do not make purchases. Users join the site by invitation, that is a member invites others to join the community. The business model is characterized by the predominance of the first category of individuals.
Moving on to an analysis of Voyage Privé's consumer segments, it should firstly be pointed out that in general analyses of online demand are performed mainly by means of tracking software such as Google Analytics, rather than with proprietary systems (Eulerian was cited as an example of this latter kind of software widely used by competitors). In particular, the management stressed that while the majority of mass market IDSs refer to a central cluster characterized by high education level, aged 30-45, upper-middle income and 60 % female, the target cluster for Voyage Privé appears to be substantially different. In contrast to the above central cluster, the target chosen by Voyage Privé is represented essentially by niche users looking for advanced, sophisticated offerings with a focus on quality and price. The management also pointed out that a distinguishing characteristic of this target group is also the higher value they attribute to participation in a small club of users. In light of such considerations, the predominant importance attributed by management to aspects of quality, rather than just the size of the customer base, is consistent.
Turning finally to the group's main marketing tools, the management stressed that current communications activities focus mostly on enhancing brand reputation and increasing membership. In this respect, Voyage Privé appears to use predominantly online tools, particularly non-conventional techniques such as viral/buzz marketing, e-mail and social network marketing. Regarding offline activities, the management interviewed cited mainly public relations at fairs or through trade magazines.

A cross-case comparison, discussion and final remarks
The literature review and the study of the specific cases of Expedia and Voyage Privé provide some useful elements for addressing the research questions posed at the outset of the article: (a) Why do private sales firms exist in the online travel market? (b) What are the main characteristics of the private sales formula? (c) What are the main differences with respect to leading IDS models?
As confirmed by the recent literature, a business model is a general scheme for defining the generation of value to the benefit of both customers and the enterprise. The emergence of operators such as Voyage Privé highlights a new business model, whose features are best appreciated by contrast to other existing enterprises currently playing the leading roles in Internet distribution systems, at least in tourism, as the case of Expedia. The reasons for the emergence and growth of formulas such as Voyage Privè's stem from the characteristics of both the demand and the supply in the tourism industry. As revealed in the foregoing, opaque discount price offerings through social networking present benefits to both the tourist, who has access to advantageous offers through a particularly appreciated channel, and the hotel, which can place unsold stock on the market, circumventing the principle of online rate parity, and at the same time access select target markets.
This said, in order to better grasp and understand the phenomenon of private sales in online hotel distribution, the sections that follow present, first, a comparison between the two business models, in order to better highlight the particular characteristics of private sales and its differences with respect to the leadings IDSs. Then the managerial implications are addressed, which enables better understanding the reasons for the existence of such formulas and the competitive pros and cons in comparison to the leadings IDSs. Lastly, the limitations inherent in the study are discussed and some indications for future research directions on the topic advanced.

Leading IDS versus private sales: three main aspects to compare
Given the characteristics of the two players investigated and the information gathered through the interviews conducted, in this section we will try to summarize the main features of the two business models and highlight the main differences between them, without, however, attempting to provide any statistical evidence. Three main aspects, namely, target segments, value proposition and revenue model, can be used to distinguish the two business models. Table 2 summarizes these differences.
Target market segments. Expedia appears to follow a strategy of totally free market access to end users. Voyage Privé, instead, places great emphasis on focusing on a segment of demand that is looking for 'specialty' tourism products, but at affordable prices. In this case, access to the site is limited to registered users, who must first be invited to register by an existing member of the community, after which they can login using a password issued to them upon registration. With regard to expanding their customer base, it should be emphasized that Expedia adopts an essentially hybrid strategy, using both online and offline channels, in particular, conventional mass media combined with Web marketing (i.e. Search Engine Marketing and Optimization). In the case of Voyage Privé, on the contrary, the strategy is focused exclusively on Web marketing and unconventional instruments, in particular, viral marketing and social initiatives, such as buzz marketing.
Value proposition. In formulating its hotel assortment, Expedia covers the widest possible offering in terms of both type-quality of contracted hotels and geographical coverage of destinations. Hotels in direct competition are included together in the inventory. The distribution agreement with hotels has an annual duration, automatically renewable for the following year, while the inventory allocation made available by the hotel is accessible at least for the next 12 months, according to a 'free sale' mechanism (though it is possible to specify some close out dates at the start of the contract). Therefore, in this case, a hotel's presence in the assortment is potentially unlimited in time and is, in any event, continuous throughout the duration of the contract. Regarding the way hotels are presented, it should be noted first that a very concise, standardized description appears for each facility and the ranking in the listing is processed through an algorithm that takes into account a number of parameters, such as price, guest reviews, the willingness of the hotel to recognize extra commissions, the average level of inventory allocation granted, and finally the presence or absence of promotions.
In the case of Voyage Privé, on the other hand, the composition of the inventory is severely limited in both hotel type and location. In fact, regarding the former, only high-quality accommodations (four-five stars or equivalent) are accepted, while from the geographical standpoint, each issue of the newsletter announcing the current promotions contains only one hotel for each destination, thereby avoiding cannibalization amongst the set. The choice of destinations and related facilities to be included in the inventory is carefully adjusted from time to time depending upon market demands and trends. Regarding the contract, it is divided into a general section, to which specific arrangements (in terms of rates, availability and close out dates) are added from time to time for inclusion of a certain facility in the assortment. Specific arrangements are necessary for each occasion. As regards inventory allocation availability, in this case, it is limited to a period of 3-6 months, while the selling dates are in addition limited to 5 days after issue of the newsletter. Therefore, hotel presence on the site cannot be considered continuous, as it depends firstly on compliance of the hotel with choices made by the retailers in terms of assortment, and also on the hotel's willingness to adhere to a given promotion. The presentation of a hotel in the inventory is in addition far more elaborate than Expedia's, with extensive descriptions and a sophisticated graphical layout reminiscent of a stylish catalogue. Further differences emerge in the two companies' policies for adapting their assortments to various international markets. In the case of Expedia, the hotel inventory is the same throughout the various sites of the group's network, with adjustments in relation to different national markets. Voyage Privé, on the contrary, uses wholly different hotel inventories assembled expressly for each country.
Revenue model. Although both use a 'merchant' revenue model, it should be emphasized that Expedia relies on a transparent sales formula, with opaque rate proposals limited to only a few 'top-secret' deals. Voyage Privé, on the other hand, follows a formula based exclusively on opaque rates in a variant specific to private sales.

Managerial implications
The foregoing comparison raises some significant managerial implications. Considering the foregoing points of comparison between the business models, these implications concern at least three distinct, albeit interrelated, aspects of how value is generated, that is, through (1) positioning; (2) discounting; and (3) profiling and new targets.
With regard to the first aspect, as we have seen, Voyage Privé aims to be a consistent channel for hotels offering products with high positioning. Naturally, the model is fuelled by the high-level positioning of the tourism products offered, but it is not limited to simply confirming such position. The travel products offered on Voyage Privé are, from the consumer perspective, clearly 'aspirational', as the site presents their offerings as 'dream' vacations and resorts. For hotels, being part of Voyage Privé's offers not only serves to substantially confirm their high-level position, but hopefully to eventually increase public awareness of such position. Moreover, the 'opaque rate' formula offers hotels the chance to circumvent the principle of rate parity and to propose aggressive promotional offers without jeopardizing the standing of the hotel brand.
With reference to the second aspect, discounted pricing necessarily goes counter to such high positioning. Also in this respect, adopting 'opaque rates' has meaningful managerial implications, as it offers hotels the possibility to pursue the natural aim of minimizing unsold products and at the same time monetize its inventory, unconstrained by considerations of positioning. To their community members (acquired, as seen, not in a free and open fashion, but via public relations, through co-optation and invitation), Voyage Privé offers a number of limited promotions-a process consistent with a 'showcase' of offers limited enough as to recall a 'luxury products display'.
With reference to the third aspect, recourse to 'flash deals' offers hotels the possibility to profit from a vast, highly-profiled customer base, thereby gaining access to marketing information that is otherwise difficult to come by for SMEs (such as Italy's many independent hotels). Flash deals provide the added benefit of possibly translating communications costs into sales commissions. By design, Voyage Privé's target customers are people who aspire to luxury products, but are seeking exclusive conditions by which luxury does not necessarily mean costly. The perception of luxury is a value clearly delivered by Voyage Privé to its customers in terms that clearly differentiate them from leading IDSs such as Expedia. The goal is above all to offer luxury defined in terms that dissociate it from price. Naturally, this is a key aspect not only for the customer-tourists, who in the case of Voyage Privé are looking for advantageous pricing conditions, but for hotels as well, as they are looking to monetize their own inventories. The values Voyage Privé produces for both are not, however, limited to price considerations. For the customer-tourists they offer the key to a 'dream' product at very advantageous conditions, or at the very least to a product of confirmed high positioning made accessible 'at a discount'. To hotels they offer the possibility to reduce the risks of de-positioning, not only through the 'opaque' nature of the offer, but also via the sales context represented by the channel and the select target reached.
Voyage Privé does not work through, or even enable, price comparisons, as is evident from the choices they make (limited number of offers, no competing offers to the same destination, limited time availability, etc.). Instead, the aspirational quality of offers tends to confirm the products high positioning-in this sense, producing a service for hotels that wish to go beyond simply monetizing their inventories.
The model does, however, suffer from various limitations. From the perspective of the customer-tourists, it is by nature a formula that cannot broaden its customer base beyond a certain point, without, that is, risking its generation of value through positioning and through profiling. From the perspective of the hotels, the drawbacks to 'flash deals' include very low margins, the risk (present in any event) of commoditizing their own offerings, the risk of vertical competition from distributors (brand loyalty, e.g. to Voyage Privé), and low customer retention. In general, there is the risk of image and price competition.
The growth of market schemes analogous to Voyage Privé's could come about by multiplication and specialization of the proposed formula. Such a growth hypothesis stems from the very nature of the Voyage Privé model, which itself emerged from a process of specialization with respect to unspecialized dealers, such as Groupon, for example. This process could continue to define new spaces of growth for 'clusters' of niches.
The identification of specialties and niches leads to the value generated through profiling and the definition of new targets stated above. As customer profiling becomes more refined, it will be possible to a identify groups of buyers to which to offer ever more specific formulas and thereby generate value and modify the business model. In the case of Voyage Privé, the group of buyers identified is more sophisticated and demanding, but it is also an exclusive group (in the sense that it is non-inclusive and without free access). Not only are the customers in this group the recipients of 'opaque' offers, but they may very well have a clear profile and even share a context that leads them not to define the positioning of a product based on its price. Other elements enter into play in defining positioning-elements that include not only the attributes of the product, but also its exclusive nature and the value produced by the channel.
In this regard, the leading IDSs can compete in generating value through discounts (2) and to some extent they can protect positioning through 'opaque' formulas, but they are very much at a disadvantage in terms of generating value through positioning, in that it is much more difficult for them to put forward exclusive 'showcases' that appeal to perceptions of luxury through an emphasis on the 'aspirational' aspect (1). But, most of all, it is difficult for the leading IDSs to produce value in terms of profiling and new targets (3), as their customer base is far more numerous and transversal, if not universal.
In the case of Voyage Privé, in comparison to Expedia, the business model is also extended by regarding hotels as the recipients of value 'to be delivered'. A service that proposes 'flash deals' also offers technical and professional support for defining the communications campaign (presentation of the offering), as well as the followup management of sales. Moreover, such value is not represented solely by the monetization of inventory, but also takes on the features of the rendering of commercial services aimed at respecting and consolidating the hotel's positioning, and furnishing access to less generic or transversal profiled customers, who are oriented to a product by its 'aspirational' aspects and more willing to dissociate its positioning from its price. Naturally, all this does not modify the channel's inherently discount nature, with the consequent implications on margins and brand image, but it is associated with services able to circumscribe such implications. In this sense, the business model extends to include hotels as 'clients' of the special services afforded by the merchant.

Limits and future research
Finally, it is possible to highlight some limitations to the present study and provide some indications for future research.
Some words are in order regarding the limitations of the research presented herein. One first shortcoming concerns the analysis conducted on only two cases. While, on the one hand, the two cases are the online market leaders and therefore representative of the two specific business models in question, on the other, it is reasonable to assume that a broader analysis encompassing more players would reveal further features useful to understanding the competition between IDS and private sales. A second limitation concerns the fact that the general empirical analysis conducted involved an in-depth geographical examination primary focus of which was the Italian market. Also in this regard, a comparative analysis of other and more varied geographical settings could shed light on additional features above and beyond those things which emerged in the current study.
These aspects represent not only limitations of the research presented here, but also stimuli for subsequent empirical studies. Such future research should address some potential trends that have emerged from current research. A first research direction to consider concerns adopting the evolution of the private sales business model and its characteristics over the time as soon as model itself proves to be no longer static, but subject to change in terms of the relationships between the players involved and in market trends. For instance, the weight of the 'last minute deals' on the total number of offers, as well as the emergence of opportunistic behaviours, can have significant implications on both the selection of target customers as well as the type of associated discounts.
In addition, future research will have to analyse the competition between tourism-oriented social networks, such as Voyage Privé, and social networks not specialized in tourism that nevertheless propose hotel assortments amongst their offerings (such as Groupon). Such a comparison between these two contexts would be useful in order to further appraise the business models of groups such as Voyage Privé, as well as to evaluate the prospects for the development and economic sustainability of such formulas over time.
A second research direction worthy of following concerns the competitive reaction on the part of leading IDSs to the phenomenon of social commerce. In this regard, future research should be aimed at analysing if, how and the degree to which leading IDSs will integrate the new 'social' sales methods into their practices. In particular, we must look at whether the leading IDSs will attempt to implement social sales methods within their own channels, and how such integration can be implemented-whether internally through organizational changes or acquisitions, or externally through agreements with social commerce sites, whether they be unspecialized or specialized in tourism. New business models in online hotel distribution 203 A third and final research direction concerns instead the impact that technological innovation can have on innovation of the business model. In particular, future research in this regard will have to study the effects on the competitive potentials of the two business models of the current generalized trend towards ever-increasing widespread use of mobile applications running on Web-enabled tools other than the PC (such as smartphones, tablets, e-readers).