For definition purposes, interactive media marketing includes social, search and mobile media marketing. Here the consumer or agent (travel agent, travel manager or event planner) interacts individually with the supplier or intermediary without human interaction to generate immediately or eventually incremental revenue and financial contribution for a hospitality supplier.

The ultimate financial measure of marketing value is the traditional return on investment (ROI) from media spending. Measurement of that effect should capture the contribution (revenue less marginal cost) of incremental business driven to a supplier or intermediary by a specific interactive media marketing activity net of cost.

Several recent articles attempt to discuss such effects. For example, Chris Anderson estimated the value of being listed versus not listed in an online travel agent (OTA) display as incremental reservations growth between 7.5 and 26 percent.^[1] Rohit Verma and Ken McGill reported that senior lodging managers are making interactive media marketing decisions “without sufficient time to investigate the ROI.” “Yet, 60 percent anticipated further increases in their budgets for such activities.”^[2]

There are a number of very important reasons why deriving an accurate ROI for interactive media marketing is difficult. Specifically, it is difficult to measure the ROI effects of activities like search engine optimization, keyword buying, OTA positioning and participation, tweeting, blogging, positioning in peer-to-peer reviews (e.g., Trip Advisor) and mobile applications. Some of those reasons stem from the nature of research design; some from the nature of marketing itself; and some from the data available needed to capture accurate ROI values.

Research Design

A large portion of hospitality marketing research, and a likely approach for measuring the ROI from interactive media marketing, stems from a social, rather than natural science research design. In the natural sciences, identical subject and control groups are created (e.g., separate white mice colonies) where an incremental change (e.g., the introduction of a carcinogen) is made in one group and not in the other. Over time the incremental effect is measured between groups.

In the social sciences, such an experimental design is not so easy. Indeed, some hospitality marketing researchers have used human subject experiments to measure response to certain types of marketing activities and stimuli. However, extending the results to an explicit ROI for hospitality marketing is difficult. The reason, as we discuss below, is due to the highly interactive reality of hospitality marketing and service delivery. Even if marketing researchers are able to create control of, and subject groups of near-identical participants to, a research design that controls for all but one of multiple interactive media marketing stimuli, applying results to a highly interactive hospitality marketing environment would be difficult, if not, suspect. Marketing involves the creation and delivery of multiple stimuli. Customer response is likely to emanate from multiple stimuli as well.

Typical social science research uses sophisticated statistical techniques to isolate both the individual and interactive effects of subject independent variables on key dependent variables. Statistical analysis is applied to either survey responses or real time market activity. ^[3] In the case of interactive media for hospitality marketing, for example, such research attempts to capture incremental changes in revenue (ADR and/or occupancy) associated with changes in keyword buys, search engine optimization, mobile applications, OTA displays, and so forth. This could be done with surveys that use responses to carefully designed questions to ferret out the likely behavior of various types of consumers (e.g., business versus leisure, agents versus individuals, loyalty club members versus non-loyalty club members) to various interactive media marketing actions. The effectiveness of this approach in accurately measuring ROI suffers from the same issue as the quasi-natural science approach. How applicable are the behavior indicators to the estimation of an explicit financial return? How well does the survey structure control for the customer mix of a specific property?

That is to say, the use of results is limited by its applicability for suppliers and intermediaries who have a business mix (or customer population) different from the survey respondents and results may be too far removed to explicit financial measures of ROI.

The other general social science research approach is to use a set of time-series and/or cross sectional data that contains a select set of independent explanatory variables including measuring interactive media action (e.g. relative position in OTA displays, number of property or brand mentions in social media sites, keywords, and so forth). These variables plus other explanatory variables like prices, income level, service level and such are used in various types of multiple regression models to measure in a probabilistic way the independent and interactive effects on financial performance variables: revenue, ADR and occupancy.

This approach has the advantage of replicating the highly interactive hospitality market environment of media marketing. It also allows the researcher to directly link variations in interactive media activities to variations in financial performance. While there are advantages to this approach, as we discuss below, interpreting results and getting appropriate data for analysis can be difficult.

Nature of Hospitality Media Marketing

Hospitality media marketing creates results over time and in association with other key factors that drive ROI for suppliers and intermediaries. Key factors include such things as the set of competitors from which incremental business share can be captured and/or the size and growth of the market itself. It is also affected by target customer segments, loyalty program features, services, service quality, price (or cost), and the use of traditional media print, TV, cable, and radio by the supplier and the other suppliers in the competitive set. For hospitality suppliers, other factors include membership in a chain and/or reliance on external marketing organizations like representation companies. The contemporaneous effect of these key factors with interactive media activity, individually and effectively, can affect ROI. Disentangling the interactive effects among all of these factors is difficult and a statistical challenge.

Research design and the effective use of statistical tools can go a long way in producing reasonable and statistically reliable results. Describing those statistical approaches and tools is beyond the scope of this report. In summary, the most effective approach involves capturing the best available data using it in a combined time-series and cross-sectional analysis at the property and competitive set level to capture effects.

Regarding this last point, the temporal aspect of hospitality marketing makes capturing the independent and incremental effects of interactive media remarkably complex. Generating incremental revenue from a given consumer in the current period is often a function of marketing activities that occurred in the past to create consumer awareness, interest, and advocacy that leads to a current period purchase choice. Moreover, the network of activities in which that consumer and the supplier and/or intermediary participate will affect a current period purchase. This is further complicated by customer advocates who can generate even more business among a suppliers’ target market. This argues for a time-series and cross-sectional statistical approach among individual suppliers at an appropriate (i.e., market-relevant) competitive set level.

To illustrate this complexity, consider the following illustration. Here consumers’ choices are shown to make incremental purchases occur over time and follow a potentially bewildering set of possible channels of influence over an extended period of time.

Availability of Useful Data

Fortunately, the hospitality industry has access to Smith Travel Research performance data: ADR and occupancy, at the property and competitive set level and by property quality. As well, data related to search, social media and mobile activity has become increasingly available, again at the competitive set level.^[4] Further, explicit price information for various channels of distribution is also available at the property and competitive set level.^[5] Finally, information about regional economics and geographic data are readily available from a variety of national and state governments, particularly in the U.S.

Unfortunately, creating a sophisticated statistically reliable model to capture the independent effects of interactive media on incremental ROI is daunting. Was a change in revenue ADR and/or occupancy really incremental to the supplier or would that revenue been generated anyway? Could ROI have been the result of other factors; for example, loyalty membership, corporate negotiations, event planner suggestions, and so forth?

A classic example of this is attributing incremental revenue from property level keyword search reports. In such reports “keywords clicks” to a brand website or call center are assumed to create hotel supplier “conversions” to bookings at given prices and lengths of stay. This is sometimes assumed to represent ROI from keyword purchases. Yet, such values either over or under-state potential ROI.

Keyword reported ROI overstates the effect if some customers would have chosen the property anyway. This can happen when a customer is seeking to book a specific brand in a specific location. The customer clicks on the keyword; is sent to the brand site; and makes a booking. Here marketing actions to create brand awareness and value were the key factors in consumer choice, not the keyword selection. Alternatively, even if a customer does not “click” on a property keyword and does not convert, that same customer may, at some future time, contact the property via the brand site or call center to make a booking. In this case the keyword purchase “caused” an incremental booking. Hence, a keyword value is under-stated.

Conclusion

The effects of interactive media marketing on hospitality financial performance (ROI) will be measured someday. Such activities are simply too important not to be measured! Deriving such measures will take research, creativity, and the willingness of suppliers and/or intermediaries to participate in the research (i.e., to create quasi subject control over circumstances). Results will also come from several alternate research approaches to the measurement process. In the interim, we should: (i) be very skeptical of reported research results, (ii) be critical of the research approach and methodology, and (iii) be wary of any author’s interpretation of results. All that said, we should encourage and support research in this key area.

[1] Chris Anderson, “Search, OTAs and Online Booking: An Expanded Analysis of the Billboard Effect” (2011) and “The Billboard Effect: Online Travel Agent Impact on a Non-OTA Reservation,” Center for Hospitality Research, Cornell University, 2009.

[2] Rohit Verma and Ken McGill, “2011 Travel Industry Benchmarking,” Center for hospitality Research, Cornell University, 2011.

[3] See V. Kumat, J. Andrew, and Robert Leune, “How Valuable is Word of Mouth?” Harvard Business Review, October 2007 for an interesting analysis of referral customer value measurement survey approach.

[4] For an interesting article on Social Media effects and variables see Douglas Quimby and Deepak Jain “Social Media in Travel” PhoCus Wright, June 2010.

[5] See www.travelclick.com

A Ph.D. economist, author and recognized expert in the areas of travel technology in marketing distribution and decision taking. Dr. Carroll is a Senior Lecturer at the School of Hotel Administration, teaching courses in economics, yield management, pricings and marketing distribution. He is also CEO of Marketing Economics, a consulting firm specializing in travel industry pricing, distribution, yield management and strategic planning. Over 25 years experience in senior positions in the travel industry. He was the Division Vice President for Global Marketing Planning at Hertz, responsible for global pricing, yield management, marketing information systems, and counter sales. He implemented the first decentralized yield management system in the car rental industry and a comprehensive Executive Information System (EIS) that gained national recognition. Following Hertz, Dr. Carroll served as the Global Vice President for Reed Elsevier’s Travel Group, which included responsibility for Travel Weekly, the Hotel and Travel Index, the Official Hotel Guide and the Official Meetings and Facilities Guide.

Cayuga Hospitality Advisors
www.cayugahospitality.com