Analytics has considerable untapped potential to contribute to the bottom line in the hospitality and gaming industries. There are several categories of analytics which may be underused in the hospitality industry including predictive analytics, operational analytics and consumer insight analytics.

Predictive analytics is the practice of developing a mathematical model based on historic data which can be used to predict future behavior. Predictive analytics has been successfully used to predict purchases (e.g. Target), high crime areas, job turnover (e.g. Hewlett-Packard) and more. In the hospitality industry, predictive analytics can more accurately forecast booking behavior, incremental spend, guest response to a certain promotion or the holy grail of marketing: new customer acquisition and retention.

Operational analytics involves the analysis of any portion of the business operation often aimed at developing precise demand forecasts which can be used to increase efficiency and reduce costs. Operational analysis can be used in workforce planning, service design, and hiring to reduce bottlenecks in operations, thereby improving performance and the guest experience. In the integrated resort, the number of operations is extensive and includes gaming, food and beverage, retail, hotel, entertainment and all back of house operations. Developing a base line of operational efficiency or using benchmarking data for the same purpose, efficiency can be measured and improved. Improving operational efficiency in all areas of operations can flow through directly to profit improvement and increased guest satisfaction.

Consumer insight analytics appear to be the most widely practiced analytics in the hospitality industry. Most hospitality organizations have guest survey systems with metrics such as net promoter, satisfaction, loyalty and intent to return. Analysis of guest satisfaction data allows the development of customer segmentation, as well as more sophisticated analyses such as clusterwise regression which can show how each customer segment cluster affects dependent variables of interest such as gaming spend, total spend, etc. In this article, we will show how a simple analysis of a promotion data set reveals potential cost savings without sacrificing quality and satisfaction of guests.

Consumer-facing organizations of all types develop marketing promotions and distribute those promotions to potential and former guests in a variety of ways. These programs can require considerable resources to develop. Evaluating the return on these promotions in terms of participation (i.e. redemption) is essential to focus on the most productive promotions, while culling those that are not attractive to or redeemed by guests. In this data set, we are measuring guest interest by evaluating the rate of redemption in both quantity and dollar amount.

The use of promotional marketing offers is nothing new. From an offer as simple as a discount coupon to motivate a customer to return, to a highly-involved “best customer” cloning strategy for new customer acquisition, marketing promotions and prizes are limited only by the creativity and insights of the professional marketer. While many promotions are quick and easy to execute, such as a birthday promotion, others involve substantial resource commitments of both people and budget. In any case, promotions are designed with a theorized return on the investment or an incremental cost of acquisition for a new customer campaign. The cost commitment varies greatly depending upon the total resources devoted to the campaign. Those resources include not only human capital, but data science, data management, and creative and distribution costs. Consequently, managers are increasingly held to higher and more accurate measures of the returns on these scarce resources.

Using even simple analytics and descriptives, managers can determine the return on the marketing and promotion investment and develop recommendations for moving forward with more productive, potentially less costly campaigns. The following analysis is based on a confidential data set from an integrated resort which includes all promotion programs along with guest redemption data. The dataset has approximately 1 million redemption records, totaling roughly $20 million in guest promotion redemptions.

Using descriptive analytics and data visualization, we determined that the top four promotions (out of 19 total) were providing 91% of all guest redemptions. This is consistent for both the frequency of redemptions and for the total redemption amount. It also becomes clear that 13 of the 19 promotions represent only 6% of both the redemption frequency and the redemption amounts, indicating low guest interest in these 13 promotions comparatively.

Eliminating the 15 lowest-redeemed promotion programs may reduce development and distribution expenses significantly, while potentially reducing the total redemption amount by 9%, if all things remain the same. While that 9% is a considerable amount of money (roughly $2 million), the corresponding reduction in spending on those 15 promotion programs may be equal or better. In other words, some of these promotions may in fact have a neutral or net negative return on the investment.

It would be ideal to have guest total spend and promotion development and distribution costs to tie to these promotion redemptions in our data set, but that data was not currently available. However, accounting for the additional revenue would necessitate a more rational cost allocation of all other variable costs associated with redemption. The recommendations and implications of this analysis may be considerably different if it was seen, for example, that the lower redemption quantity and/or amounts resulted in a higher guest total spend or profit per guest. However, given only the data present, and making the assumption that if the guest came into the resort and redeemed their promotion they also spent incremental revenue, the implications stand.

The strength of this type of descriptive is that it gives us insight into what’s actually working, what’s not, and what incurred expenses on development and distribution of these types of promotions are achieving the expected and desired return. Adding an incremental spend to and expanding distribution of the most productive promotions (e.g. Cash in the Mail, Recency, and Ad Hoc Direct), while eliminating the promotion programs that aren’t bringing in the guests for redemptions, may reduce costs and keep guests coming in and returning. Alternately, the cost savings gained by reducing marketing and promotions spend for the lower-performing programs may be reallocated to develop a methodology to test new promotions, variations on the successful promotions and even integrate those higher return promotions into new customer acquisition programs. The larger implication of this article is that using even simple analysis can reduce costs and increase effectiveness of the marketing strategy, drive a higher return on dollars spent, and decrease or eliminate costs on unproductive campaigns.

Complex analysis methods such as predictive analytics can be used to develop more nuanced models to predict guest behavior, such as booking rooms, gambling spend or integrating unstructured data such as social media. Interestingly, most properties already collect data which could be used to develop the predictive behavioral models. For example, current guest data could be used to develop a logistic regression with the dependent variable of interest as book/didn’t book a room, additional purchase/no additional purchase, etc. Another option would be a linear regression to predict total spend in various departments given many currently existing guest survey or social media feedback variables. Many properties could implement predictive analytics using data that they already collect.

For operational analytics, performance metrics are frequently collected and recorded and can be used to determine performance benchmarks or baselines and analyzed for trends, changes, improvements, and efficacy of training programs or coaching sessions. You can develop and expand your analytics department to take advantage of these practices by ensuring metrics are in place and that analysts have strong basic and industry-specific statistics knowledge, know how to manipulate data with simpler programs like Excel and Access, understand data visualization – both its creation and interpretation, and have the wonderful skill of story-telling, that is, using your understanding and visualization of data to make a case for a change or decision with that data and its analysis as evidence.

Dr. Heather Monteiro is the Market Research Analyst for the UNLV Hotel College’s Center for Professional and Leadership Studies (PLuS Center). Dr. Monteiro earned an MBA from Georgia College and State University and a Ph.D. at Georgia Southern University in Logistics and Supply Chain Management, with a minor in Marketing/ Consumer Behavior.

Robert Rippee is the Director of the Hospitality Innovation Lab at UNLV. The Hospitality Lab develops and designs innovative applications and ideas for global hospitality industry. The lab is a project oriented innovation lab addressing the top needs and opportunities within an entire industry. The scope of exploration in the lab includes technology, robotics, big data, eSports, mobile, experience design and process innovation. Along with providing vision and guidance to the multi-disciplinary students, Robert also actively researches millennial ethnography, behavior and eSports Gaming.


Comments are closed.