Destination revenue management

Destination revenue management (DRM) is a business intelligence method intended for destination management organisations (DMOs). While business intelligence is defined as “the set of concepts and methods for informed decision-making based on factual arguments” (Power, 2007, p.128), its application to the tourism industry raises some questions: the product marketed is a complex system called “destination” (d’Angella et al., 2010), and the highly fragmented market restricts the ability of each individual player to see fluctuations in demand. Destination revenue management, conceived and implemented by the French company RMD Technologies (https://www.rmd-technologies.com/), addresses this issue of business intelligence to serve destination governance and its socio-professional players.

Business intelligence for DMOs: decision support or justification?

The model developed by the University of St. Gallen (Beritelli et al., 2015) models the flow of visitors to a destination, with the aim of facilitating decision by institutional players (designing itineraries that fulfil the desires of visitors, etc.). Mobile operators, for their part, offer big data enabling DMOs to throw light on the departure and arrival points of visitor flows, to distinguish between day-trippers and tourists, and thus to provide quantified evidence on the attractiveness of their destination (Berlingué and Contet, 2017).

However, the challenges of business intelligence go beyond the mere justification of effective public action. In keeping with the new public management approach (Hood, 1991), DMOs are now considered “accountable” (Huron and Spindler, 2019). They must be able to quantify their action in terms of promotion, reception, information, and the coordination of socio-professional players by using physical flow measurements or other barometers of the popularity of a destination on social media, etc.

For a destination, the challenges of business intelligence go beyond the mere justification of effective public action.

With such observations, the DMO is however not in a position to assess the effectiveness of its action, for example, to establish a causal relationship between a promotional campaign and a tourist’s decision to visit a destination. A real business intelligence tool must enable the DMO, by studying the volumes of bookings by region of origin of customers, to highlight the main places of origin associated with the bookings, which should then mobilise their promotional efforts. In the case illustrated below (Ill. 1), the study of bookings in Charente-Maritime camp sites by region of origin of the French clientèle indicates that operators should focus their promotional action on the regions along the Atlantic coast (Nouvelle-Aquitaine, Pays de la Loire and Brittany), rather than indiscriminately rolling out expensive communication campaigns on social media or in the Paris metro.

Ill. 1. Number of bookings in Charente-Maritime camp sites by region of origin in 2019 and 2020 (source RMD Technologies and Charentes Tourisme, a French DMO)

Big data, a fundamental component of business intelligence in tourism

What information is really likely to facilitate optimal decision-making by DMOs and hosts (selling price of a room, targeted promotion of the destination, etc.)? Other than contextual data (weather conditions, calendar of cultural events etc.) and those relating to the civil status of the traveller, it is above all behavioural data that must be observed: what date does the traveller make the booking, for what date and what duration of stay, what type of accommodation do they choose, and at what type of rate (flexible or non-refundable). These data flows contain valuable informational primitives (Pigni et al., 2016): who is travelling, what, when, where, why and how? Ill. 2).

Ill. 2. The primitives, the elementary informations contained in a digital data flow (Pigni et al., op.cit.,p8)

These primitives can be found by studying a host’s booking data (Ill.3). Their aggregation over several successive years, and over a sufficient number of establishments in the same destination, makes it possible to predict future events. If, for example, there is a delay in the usual rate of ramp up of bookings (Ill. 3, column D) for a specific customer segment, should we not consider a promotional campaign targeting that segment so as to stimulate demand?

Ill. 3. Illustration of the wealth of digital data flows. PMS is the property management software for managing the operation of a hotel, camp site, etc.: booking, etc. (source: Excerpt from the property management software (PMS) of a camp site).

Within an business intelligence tool, data must meet conditions of velocity, reliability, and representativeness (Baggio, 2016). Particularly critical is the fact that the degree of latency (Hackathorn, 2002) in the digital data flows (Ill. 4) used will determine the ability to make a decision in response to a change in the behaviour of demand. We can distinguish latency in the capture of data (delay between the manifestation of an event and the transcription of its primitives in the form of digital data), the analysis of data (delay in transforming data into information e.g., am I ahead of or behind last year on the same date in terms of bookings received?), and finally in the use of data (delay between the availability of information and decision-making).

Ill. 4. Latency in capturing, analysing, and using data flows (Hackathorn, 2002)

This phenomenon of latency serves to highlight two possible and complementary strategies in terms of exploitation of data flows. The first is to create value by making decisions based on real-time data flows: for example, deciding on a targeted advertising campaign via social media, when the data shows a delay in taking bookings from customers from a given origin (e.g., Belgian customers). The second strategy aims instead to extract value by analysing multiple data flows. By comparing the ramp up of bookings over several successive years, the aim is, for example, to produce algorithms that can predict a level of demand for a future date (Schwartz et al., 2016). In this second strategy, the analysis makes it possible to gradually form a refined vision of future demand.

Data flow strategies in a fragmented market: the power of tourism distribution platforms

The tourism industry is highly fragmented: there are 18,000 hotels in France, two-thirds of them not affiliated with hotel groups (Coach Omnium, 2020), nearly 8,000 camp sites etc. And that is just accommodation providers. In such a market structure, each provider has only minimal market power and therefore has access to very limited information on demand.

The Internet, the main tool for the organisation and marketing of travels (Echo Touristique, 2020), further facilitates the concentration of intermediaries between supply and demand: online travel agencies, community platforms and search engines. For example, Booking.com, with an estimated market share in Europe of over 30% (Caccinelli and Toledano, 2018), has a certain market power to extract considerable value from booking data flows.

The tourism market must therefore be seen as a “two-sided market” (Evans, 2003), where intermediaries (OTAs such as Expedia or Booking.com; community platforms such as Airbnb or HomeAway etc.) operate as platforms simultaneously selling services of a different nature to both sides of the market. The volume of demand coming from one side of the market (the hosts) depends on the level of demand on the other side of the market (the travellers using the platform to rent accommodation), and vice versa. Hosts are only willing to pay a commission for the platform’s services if they are certain that it will give them access to buyers. The business model of these intermediaries therefore requires being able to attract hosts and travellers simultaneously to the platform, guaranteeing at all times and for any destination an abundant supply of accommodation, while also ensuring a constant flow of visitors to its site looking for a good deal.

The highly fragmented tourism industry is a “two-sided” market, controlled by distribution platforms that generate intermediation bias leading to non-optimal decision-making by hosts and DMOs.

The platforms benefit from information asymmetry at the expense of each side of the market. They generate an intermediation bias (Calvano and Polo, 2021) by providing hosts with recommendations (price monitoring on presumed competitors, trends in demand), whose real objective is not to provide business intelligence for an optimal decision (selling price of a room, etc.), but to guarantee an abundant supply of rooms at promotional prices at any time on the platform. Although distribution platforms guarantee, thanks to their significant investments (for online advertising (pay per click, metasearch, affiliation etc.), Booking.com allocated the following budgets: 4.42 billion dollars in 2019, 4.44 in 2018, 4.14 in 2017 and 3.48 in 2016 (ARTIREF, 2020)) excellent visibility to hosts, it is not in their interest to inform them of the best strategic recommendations (e.g., for determining the price of a room).

The host-platform relationship – and by extension the destination-platform relationship – must therefore be analysed as a principal-agent relationship with adverse selection (Akerlof, 1970), where the agent (the platform, the search engine) has an informational advantage. The services offered by the platforms to the hosts– referencing, visibility, etc.– (Autorité de la Concurrence, 2015) are fundamentally irreconcilable with a promise of decision-making support in the interest of the host, particularly in terms of anticipating demand and determining prices.

Thus, observation has become the de facto prerogative of platforms with interests that do not necessarily converge with those of the destination and its socio-professional players. Destination Revenue Management (DRM) is a response to this tourist market structure that deprives hosts and DMOs of the possibility to observe demand in order to drive a commercial strategy.

Revenue management is a strategy for forecasting, optimising, and controlling the capacity, prices and revenue of service companies (hosts, airlines, etc.) having perishable assets, limited capacity (non-scalable stock) and subject to fluctuating demand. Revenue management addresses three major issues: demand forecasting, demand modelling, and price optimisation (Yeoman, 2016). The ability to collect and analyse booking data is a fundamental prerequisite for the execution of any revenue management strategy. There are three types of data: past bookings, newly received bookings, and data about phenomena that could affect the volume of bookings (weather, events at the destination, etc.).

In order to increase the accuracy of hotel management revenue forecasting models, research has attempted to integrate other types of data. Thus, Schwartz et al. (2016) suggest that organisations in the same competitive space should share their strategies with others. The data thus shared is then used as input data for each establishment’s forecasts. The underlying idea is that a forecast becomes more accurate if it takes into account not only past individual data (past demand behaviour) or present data (ramp up of bookings in real time), but also collective data.

While this suggestion has the advantage of attempting to address the problem of imperfect decision-making in a highly fragmented accommodation sector, it also raises two questions. The first is whether the forecasts provided by individual accommodation providers are the most relevant data flows for improving accuracy. Is it not preferable to use data flows that capture primitives (Pigni, 2016), i.e., behavioural elements such as booking date, length of stay, type of service, group size, etc.? The second objection relates to trust issues inherent in the sharing of information among members of a competitor set. To consent, each individual host must accept that their consent to information sharing is the necessary counterpart to better informed decision-making. They must also be prepared to entrust the extraction of data within the competitor set to a trustworthy third party.

Instead of a platform, which is a source of intermediation bias, a destination – a group of socio-professional players, including hosts – therefore requires an intelligent agent (Sheehan et al., 2016) who is responsible for the compiling of booking flows at the collective level, from which it extracts (demand) forecasts, which it then communicates to the socio-professional players, so that they can use them to determine their individual strategies.

Destination revenue management aims to enrich the revenue management strategy of a host or a destination as a whole, by incorporating into the forecasting models – without any intermediation bias – the ability to forecast future demand for the destination as a whole.

Destination revenue management consists of aggregating data destination-wide to anticipate future demand. This business intelligence technology thus allows the DMO, the destination’s trusted third party, to guide the socio-professional players while managing the destination’s offer as a whole.

Destination revenue management: the business intelligence of a destination embodied by a trusted third party

By choosing to fulfil this role of data flow extraction for the purpose of business intelligence (Béal et al., 2021), the DMO, as the trusted third party tasked with defending the general interest of the destination and of its components, builds up social information assets of the destination. In addition to the detailed, exhaustive, and up-to-date inventory of the components of tourism supply (accommodation, events, attractions, etc.), the social information assets include real-time knowledge of demand (seasonality, booking windows by customer segment, purchasing behaviour at destination, etc.).

The DMO can thus assert itself as an intelligent agent of the destination (Sheehan et al., 2016), able, for example for promotion purposes, to mobilise its resources, with no latency effect, on the most contributory pairs between place of origin and booking windows. The DMO uses the social information assets to guarantee the competitiveness and sustainability of the destination (Ritchie and Crouch, 2003), in particular by managing the destination’s supply in time and space (attractions, events, infrastructure, accommodation, etc.).

Ill. 5. Destination revenue management: the four components of business intelligence to serve a destination.

It is worth mentioning, as an example of business intelligence technology available to a trusted third party (DMO, departmental federation of outdoor accommodation, etc.), that developed by RMD Technologies, based in La Rochelle and founded by Jean Laherrere, a revenue manager specialised in the hotel industry. One of their tools is shown in illustration 6. For the reference destination in this example, anonymised raw data from some 60 hosts (campsites) over several years, representing some 2.5 million bookings, is collected and analysed. As shown in the figure, the individual host can compare its future performance with that of the destination; it can thus make confident decisions (for example, reduce its price for a future date if it is lagging behind with regard to the destination). If, on the other hand, the tool shows that the delay concerns the destination as a whole, it is the DMO that will roll out ad hoc promotional actions.

Ill. 6. Business intelligence tool for camping, populated by digital data flows on bookings (source: RMD Technologies)

Conclusion…

In a tourism market fragmented to the extreme, the emergence of the Internet has radically changed the buying behaviour of tourists and the organisation of distribution. The power of platforms today undermines the sustainability of destinations and their stakeholders, while every year we see intense competition between DMOs for expensive promotional actions, without real targeting based on the temporal and spatial specificities of demand. As their natural trusted third party, only DMOs can embody and build tourism intelligence to serve the destinations.

Luc BEAL

Bibliography

  • D’Angella Francesca, De Carlo Manuela et Sainaghi Ruggero, 2010, «Archetypes of destination governance: a comparison of international destinations», Tourism Review. vol. 65, n°4, p. 61-73.
  • Akerlof Georges, 1970, «The Market for “Lemons”: Quality Uncertainty and the Market Mechanism», Quarterly Journal of Economics. n°84, p. 488-500.
  • ARTIREF, 2020, Les chiffres clés de Booking 2019, [consulté le 20 avril 2021].
  • Autorité de la Concurrence, 2015, «Décision n° 15-D-06 du 21 avril sur les pratiques mises en œuvre par les sociétés Booking.com B.V., Booking.com France SAS et Booking.com Customer Service France SAS dans le secteur de la réservation hôtelière en ligne?», [consulté le 22 Mai 2021].
  • Baggio Rodolfo, 2016, «Improving Tourism Statistics: Merging official records with Big Data», dans Fuchs Matthias, Höpken Wolfram et Lexhagen Maria (dir.), Big Data and Business Intelligence in the Travel and Tourism Industry. Östersund, Mid Sweden University, p. 89-92.
  • Béal Luc, Sheehan Lorn et Zaman Mustafeed, 2021, «Destinations, données, décisions: le besoin impérieux d’un tiers de confiance», dans Botti Laurent et Spindler Jacques (dir.), Organismes de gestion de destination : Stratégies et pratiques pour un management responsable et durable des territoires touristiques. Paris, L’Harmattan, coll. «GRALE-AFMAT», p. 159-188.
  • Beritelli Pietro, Reinhold Stephan, Laesser Christian et Bieger Thomas, 2015, The St. Gallen Model for Destination Management. Saint-Gall, IMP-HSG, 200 p.
  • Berlingué Christel et Contet Jean-Michel, 2017, «Flux Vision Tourisme. Le big data au service de l’observation des flux touristiques», Revue Espaces. n°335, mars.
  • Caccinelli Chiara et Toledano Joëlle, 2018, «Assessing Anticompetitive Practices in Two-Sided Markets: The Booking.com cases», Journal of Competition Law & Economics. vol. 14, n°2, p. 193-234.
  • Calvano Emilio et Polo Michele, 2021, «Market power, competition and innovation in digital markets: a survey», Information Economics and Policy. vol. 54, mars, en ligne.
  • Coach Omnium, 2020, Panorama 2020 de l’hôtellerie en France. [consulté le 22 mai 2021].
  • Echo Touristique, 2020, «E-tourisme: les chiffres clés 2019 à retenir», [consulté le 20 avril 2021].
  • Evans David S., 2003, «The Antitrust Economics of Multi-Sided Platform Markets», Yale Journal on Regulation. n°20, p. 325.
  • Hackathorn Richard, 2002, Current Practices in Active Data Warehousing. Bolder Technology.
  • Hood Christopher, 1991, «A Public Management for all Seasons?», Public Administration. vol. 69, n°1, p. 3–19.
  • Huron David et Spindler Jacques, 2019, Management et Finances publiques – Les marqueurs du New Public Management. Paris, L’Harmattan, 300 p.
  • Pigni Frederico, Piccoli Gabriele et Watson Richard, 2016, «Digital Data Streams: Creating Value from the Real-Time Flow of Big Data», California Management Review. vol. 58, n°3, p. 5-25.
  • Power Daniel J., 2007, «A brief history of decision support systems», DSS Resources.com. 3, en ligne.
  • Ritchie J. R. Brent et Crouch Geoffrey I., 2003, The competitive destination: A sustainable tourism perspective. Wallingford/Cambridge, Cabi Publishing, 304 p.
  • Schwartz Zvi, Uysal Muzaffer, Webb Timothy et Altin Mehmet, 2016, «Hotel daily occupancy forecasting with competitive sets: a recursive algorithm», International Journal of Contemporary Hospitality Management. vol. 28, n°2, p. 267-285.
  • Sheehan Lorn, Vargas-Sánchez Alfonso, Presenza Angelo et Abbate Tindara, 2016,«The use of intelligence in tourism destination management: An emerging role for DMOs», International Journal of Tourism Research. vol. 18, n°6, 2016, p. 549-557.
  • Yeoman Ian, 2016, «The history of revenue and pricing management – 15 years and more», Journal of Revenue Pricing Management. n°15, p. 185–196.