Data Mining in the Hotel Industry

Introduction

Data mining software examines existing information and analyzes it in order to find patterns and correlations (Han, Kamber, & Pei, 2012). Unbeknownst to some consumers, savvy technical software is tracking their daily spending and Web browsing. Data mining finds the relationships between who buys what and when and attempts to address those needs in order to encourage spending. In the hotel industry, data mining can gather information such as hotel feedback, reservations, and the length of guest visits in order to identify relationships and discover new ways of increasing revenue and decreasing costs.

Because the hospitality industry’s success is based on customer loyalty, data mining is a worthwhile process because it allows hotels to understand what their guests want and how to deliver their needs. Law, Leung, Au, and Lee (2012) have noted: “The hospitality industry is not technology-oriented by nature, but the increasing demands from sophisticated customers, together with the information-intensive characteristics of the industry, are prompting managers to embrace IT to meet present and future business needs” (p. 10). In other words, customers seek information, conduct business, or find entertainment on the go with smartphones, iPads, and laptops. If the hospitality industry hopes to meet technologically driven consumers, it is in their best interest to use what they already know about their buyers by tapping into what their customers look for.

Origin

Han, Kamber, and Pei (2012) noted that data mining evolved from Information Technology (IT) and Knowledge Discovery in Databases (KDD). In the 1960’s, businesses relied on file cabinets to accumulate information and maintain their records (Han, Kamber, & Pei, 2012). Subsequently, as an organization would grow, so would their records. In the 1970’s and the early 1980s, computer science experts accepted this growth and began to develop digital database management systems (Han, Kamber, & Pei, 2012). Developers created Query languages and Structured Query Languages (SQL) in order to communicate within a database (Han, Kamber, & Pei, 2012). Essentially, SQL allows an organization to assess and influence databases. Finally, in the late 1990s, IT developers introduced data warehousing and data mining (Han, Kamber, & Pei, 2012). Data warehousing stores the data from a variety of databases in order for management to assess their business as a whole.

Consequently, Han, Kamber, and Pei (2012) reported data mining is often known as “knowledge mining” (p. 6) because it supplies users with helpful information regarding their client base. Han et al. (2012) suggested knowledge mining follows seven steps: 1) data cleaning, 2) data integration, 3) data selection, 4) data transformation, 5) data mining, 6) pattern evaluations and 7) knowledge presentation (pp. 7-8). Data cleaning identifies irrelevant data, or “noise” (p. 6) so the user can separate useful data from unusable data (Han et al., 2012). Data integration uses a variety of databases in order to combine data and find relationships (Han et al., 2012). Data selection involves finding the most relevant data based on the user’s objective (Han et al., 2012). Data transformation transforms and consolidates data into summaries (Han et al., 2012). Data mining extracts patterns (Han et al., 2012). Finally, pattern evaluation and knowledge presentation identify and present the results (Han et al., 2012). Knowledge mining is an extensive process because the amounts of data are often large and complex. However, data mining sorts through multiple databases, the Internet, and data warehouses quickly and efficiently (Law, Leung, Au, & Lee, 2012). Essentially, hotels have a wealth of knowledge, regarding their customers, within their internal files.

Application of Data Mining Technology

Tzu-Ching (2012) emphasized that the application of data mining would allow the hospitality industry to improve its internal organization and overall performance by:

Improving productivity

Enhancing guest services

Increasing revenue

Reducing costs

Developing a competitive edge (p.46)

Tzu-Ching (2012) revealed that the hospitality industry could accomplish each of these goals because they have identified their customers’ needs. At the same time, in order to successfully use data mining, the hotel industry should decide upon the initial problem and what they hope to achieve. Data mining commonly uses anomaly detection, association rule learning, clustering, and regression (Han, Kamber, & Pei, 2012). In the case of the hospitality industry, executives may want to use the method of clustering. Han, Kamber, and Pei (2012) reported “clustering plays a central role in customer relationship management” (p. 64). Ideally, clustering will help hotel executives to plan customer loyalty programs because the customers are separated into groups that identify their similarities and preferences (Han, Kamber, & Pei, 2012). Fundamentally, data mining allows corporations to produce predictions and implement strategies.

Rapp (2002) has noted there are organizations that use minimal approaches to IT strategizing, but they are “not in the game” (p. 22) because their limited use does not allow growth. On the other hand, other corporations use IT as a means to develop their competitive edge (Rapp, 2002). With that in mind, data mining is a strategy that will allow the corporation to grow and outlast their competitors. In other words, in order to satisfy guests’ needs, Suite Spot Inc. can use data mining to predict their wants beforehand and provide them with customized incentives.

As an example, Suite Spot Inc. could prepare the data and divide the data into groups such as “room related, member segment, demographic characteristics, member preferences, and nature of complaints” (Nagadevara, 2008, p.493). Room related data consists of what type of guests occupy the rooms and whether or not they typically choose budget or higher-end accommodations (Nagadevara, 2008). In regards to member segments, data mining will categorize VIP guests, long staying guests, or overnight and infrequent guests who typically stay at the hotel for business purposes (Nagadevara, 2008). Demographic characteristics are similar to member segments; however, demographics include factors such as age, gender, and race (Nagadevara, 2008). Member preferences categorize guests by the types of rooms they request such as smoking, non-smoking or a room with a view (Nagadevara, 2008). Nagadevara (2008) noted the category of complaints is often the most important piece of data to consider. For example, complaints extend to service within the hotel, such as restaurants, lounges, housekeeping, and the front desk. In that way, data mining can find associations between guest satisfaction and the quality of their service.

Subsequently, management could identify how the data was distributed from each chain and compare similarities. Because each chain developed individual systems, Suite Spot Inc.’s main objective would be to clean out any data that is inconsistent with their goal. Essentially, data mining will allow upper management to understand booking trends, in slow and peak seasons, do not necessarily predict hotel rates. For instance, in the slow season, hotel chains may naturally assume that they should lower their rates in order to bring in guests. However, data mining will uncover previous patterns that may indicate hotels can actually keep their current rate but encourage guests to lengthen their stay. Consequently, guests will stay longer and bring in substantial revenue for Suite Spot, Inc.

Inside and Outside of the Organization

Depending on the application, IT will allow Suite Spot Inc. to address issues inside and outside of the organization. Law, Leung, Au, and Lee (2012) have found that “Various types of IT applications have been broadly implemented throughout the hospitality industry, including self-service kiosks, online check-in, and internet and email applications” (p. 11). The inside organization should consider using self-service kiosks. Self-service kiosks offer convenience to the guests and the staff. For example, when guests check-in, they are able to approach the kiosk and check themselves in without waiting in a long line. Similarly, online check-ins will allow guests to check-in from the airport and on their smartphone or laptop. Generally, it seems that hotel guests rely on quick and easy check-in and check-outs, so they can travel on their way. In addition, when guests develop loyalty, hotel chains should show appreciation with loyalty programs. Correspondingly, loyalty programs are viable sources of revenue for hotels because such programs effectively retain customers and provide them with incentives to return (Nagadevara, 2008).

Notable Successes to Emulate

Data mining encourages interaction between the customer and business (Tzu-Ching, 2012); however, Nagadevara (2008) revealed customers will often join more than one loyalty program, so a hotel should develop distinct incentives based on their guests’ preferences. Tzu-Ching (2012) reported, “Harrah’s Hotels and Casinos in Las Vegas introduced a trademarked loyalty-card program [called] ‘Total Rewards’” (p. 46). The “Total Rewards” program saves and tracks guests’ or clients’ habits while they stay or play. Basically, data mining allowed Harrah’s Hotels to understand their sources of revenue from their customers’ shopping activities, so the loyalty program will offer guests further rewards, based on those activities, which, in turn, will encourage their customers to spend more. Similarly to Suite Spot Inc., Harrah’s has multiple properties. Using the system WINet allowed Harrah’s to “link all of its properties, allowing the firm to collect and share customer information” by connecting and consolidating their guests’ details (p. 46). In other words, Harrah’s Hotels and Casinos were able to identify their guests’ patterns in order to establish relationships between each property.

In another illustration, Howard Hotels & Resorts joined forces with Fubon Bank in order to provide customized offers with loyalty cards (Tzu-Ching, 2012). Every time the guest or customer would use his or her loyalty card, that information was stored in a data warehouse (Tzu-Ching, 2012). Consequently, Howard Hotels & Resorts and Fubon Bank were able to use data mining to observe their customers’ spending habits and increase their spending by offering points, discounts, or reward redemptions (Tzu-Ching, 2012). This strategy allowed both corporations to relate to their customers’ needs, promote brand loyalty, and increase their overall revenue.

As another example of data mining within the hospitality industry, Hilton Hotels created the loyalty program “Hilton HHonors” (Nagadevara, 2008, p. 488) in order to collect information regarding their highest spending guests. The loyalty program allowed members to expedite check-in times or prolonged check out times. In addition to guest comfort, Hilton Hotels also provided cardholders with other incentives such as complimentary spouse lodging (Nagadevara, 2008). Moreover, the HHonors program connects their guests’ airplane travel miles so their guests will receive airline miles along with quality points (Nagadevara, 2008). In this case, Hilton’s strategy is a worthwhile one to emulate because it encourages travel and lodging. Another strategy to adopt would be categorizing guests into membership packages. For instance, HHonors gives substantial benefits to frequent guests such as free room upgrades and access to a private health club (Nagadevara, 2008). However, effective loyalty programs that draw in guests and subsequently keep them should be based on specific clients. Ultimately, Suite Spot Inc. should categorize their guests and identify their biggest spenders.

Notable Failures to Avoid and Areas of Concern

While cost is generally the first concern, the second concern is usually the language gap between hospitality executives and IT professionals (Law et al., 2012). Law et al. (2012) proposed “hospitality managers want to provide quality service and develop economically friendly customer relationships; however, IT experts are concerned about service quality and customer relationships, whereas the IT expert focuses on rapid application development and server protocols” (p. 19). While technical jargon may leave some hotel executives feeling lost in translation, ultimately, they are the key components to the hotel industry’s data mining’s success because they have recorded and kept every transaction each guest has incurred.

In addition, mining methodology and user interaction are also issues that may affect Suite Spot Inc.’s data mining success (Han, Kamber, & Pei, 2012). Han, Kamber, and Pei (2012) have noted data mining is “an interdisciplinary effort” (p. 29). In other words, data mining does not rely on one concept, but it requires multiple factors that may inhibit its use. For example, if data contains noise, such as errors,” those may lead to false patterns (Han, Kamber, & Pei, 2012). If an organization identifies the wrong pattern in the consumers’ buying patterns, they may end up wasting advertising resources on the wrong market. Subsequently, data mining’s methodology should utilize data cleaning in the process. Also, identifying users’ patterns relies on reliable user interaction. With that said, Han, Kamber, and Pei (2012) suggested that data mining be interactive. Interactive mining involves the relationship between the user and the computer. When data is interactive, Han, Kamber, and Pei (2012) have found that it allows “user[s] may like to first sample a set of data, explore general characteristics of the data, and estimate potential mining results” (p. 30). Interactive mining will hopefully reduce the level of human error. Predictably, data mining can predict patterns and associate relationships, but it is ultimately up to the user to decipher the data. Subsequently, the poor analysis will provide inconsistent results.

Best way to deploy Data Mining at Corporate level

As a guide, the American Hotel & Lodging Association’s Technology and E-Business Committee’s (2006) suggested hotel executives focus on revenue management. In order to say afloat, corporations at all levels and functions need to make profits. The American Hotel and Lodging Association (2006) reported the

Customer Relationship Management system…track[s] behavior at an individual level and the Revenue Management software have allowed the combination of data mining and revenue optimization tools to allocate rooms effectively, based on the potential profitability of each customer and the opportunity cost of each room. (p. 26)

Consequently, forecasting guests’ likelihood of staying at a Suite Spot hotel during weekends, weekdays, or holidays will allow the corporate level to optimize the demand and hotel occupancy (American Hotel & Lodging Association’s Technology and E-Business Committee, 2006).

Best way to deploy Data Mining at Individual Hotels

In order to effectively implement technology, all hotel employees should undergo training and managers should recognize their guests are the first priority. Law, Leung, Au, and Lee (2012) emphasized: “hospitality managers must retain a good relationship with their customers and business partners by using the right IT, by incorporating IT into their business strategies, and by enhancing their staff’s knowledge of and proficiency in using IT” (p. 20). In regards to customer relations, management’s main objective is to ensure loyalty. At the same time, fostering relationships with business partners such as online travel agencies or airlines will allow each corporation to rely on each other’s data and information. Ultimately, adding new clientele to the mix will inevitably increase the client base and revenue.

If Suite Spot Inc. uses loyalty programs for their guests, they will have to ensure each code redemption, gift, or other incentive is recognized. Curtis(2012) determined that “Data mining technologies provide managers with information and insight and help businesses as a whole anticipate and behave proactively in their environments” (p. 84). Ultimately, the environment includes the guests, staff, and management, but for the most part, the guests and the staff have to most frequent interaction. It bears noting that staff members should familiarize themselves with the key components of loyalty programs. Training such as online programs or group workshops would allow IT professionals and hotel management to teach staff members how to understand the correlation between the guest and the loyalty card. While every guest should be treated with hospitality, staff should be encouraged to recognize those who bring in substantial revenue. In addition to training staff, hotel management will have to understand and explore all facets of data mining because management positions will naturally work with or assist IT professionals.

Investing in Data Mining

At one time, data mining software was exceptionally expensive; however, technological advances have allowed developers to create programs based on a corporation’s size, budget, and need. Investing in data mining is a must in the digital age. Along with convenience, the data mining software will allow Suite Spot Inc. to cluster frequent guests from around the country. Thus, the Decision Support department recommends data mining software in order to improve communication, increase revenue, and obtain guest loyalty based on Suite Spot Inc.’s existing files. Internal records are only valuable when they are compared and examined. For the most part, sales in the hotel industry are largely dependent on the guests’ previous stays and feedback.

Conclusion

Data mining software offers the user convenience. While manually digging through old files is a concept of the past, old files are fundamentally important to a hotel’s growth. Ultimately, within those ancient files lie relationships. At one time, it seems that hotel management and sales were considered separate entities, but data mining joined the two forces. Hotel management will discover tangible leads whilst the sales department goes after them. In large part, hospitality industries develop relationships with their guests because they offer them the proverbial home away from home, entertainment, and rest. Data mining would analyze those relationships and track spending habits to reveal steady and loyal guests. In turn, Suite Spot Inc. can reward their guests with loyalty programs in order to show their appreciation. Today’s guests are technologically savvy consumers, so in order to predict what their customers need, the hotel industry should embrace technical advances too. Data mining is the ultimate source of identifying information that was already there. When the hotel industry applies that knowledge to today’s business model, they can offer consumers what they want and keep them coming back.

References

American Hotel & Lodging Association’s Technology and E-Business Committee. (2006). Revenue management: A technology primer [PDF]. American Hotel & Lodging Educational Foundation.

Custis, C. (2012). The role of business intelligence within the hospitality industry's information systems strategy: Historical concepts and future trends. Journal of Management Policy & Practice, 13(2), 82-94.

Han, J., Kamber, M., & Pei, J. (2012). Data mining: Concepts and techniques (3rd ed.) [CaféScribe Format].

Houran, J., Lange, R., & Kefgen, K. (2013). Industry Trends: Fascinating Rhythms in the Career Paths of Hospitality Executives. Cornell Hospitality Quarterly, 54(6), 6-9. doi: 0.1177/1938965512467602

Law, R., Leung, D., Au, N., & Lee, H. (2012). Progress and Development of Information Technology in the Hospitality Industry: Evidence from Cornell Hospitality Quarterly. Cornell Hospitality Quarterly, 54(10), 10-24. doi: 10.1177/1938965512453199

Nagadevara, V. (2008). Improving the Effectiveness of Hotel Loyalty Programs through Data Mining. In T. L. Lockyer (Ed.), Global Cases on Hospitality Industry (pp. 487-499).

Rapp, W. V. (2002). Information technology strategies: How leading firms use IT to gain an advantage. New York: Oxford University Press.

Tzu-Ching, L. (2012). A technique for enhancing customer relationships in the service industry. International Proceedings Of Computer Science & Information Technology, 30, 45-48. Retrieved from http://ipcsit.com