The following is a brief excerpt from The 2019 Smart Decision Guide to Hospitality Revenue Management. Now in its fifth edition, this title is widely considered to be the industry’s most authoritative resource on next-generation hotel revenue management. The new edition is now available for complimentary access.
With advances in technology innovation, hoteliers are getting closer to achieving the ultimate promise of revenue management: to sell the right space at the right price at the right time to the right guest. The following is a brief overview of some of just 3 of the key concepts related to modern-day revenue management.
The basic approach to pricing guest rooms has evolved from a technique that uses best available rate (BAR) pricing, which yields one primary rate with a tiered percentage or flat-amount-based discounts off BAR differentiating prices across distribution channels, to one that uses dynamic pricing based on demand forecasts that harness the power of big data processing, cloud computing, and advances in machine learning to enable continuous pricing adjustments in the face of constant market change.
Machine learning makes it possible to perform complex computations that automatically evaluate market change as well as to incorporate predictive modeling mistakes. The algorithms become progressively smarter while avoiding the possibility of making the same mistakes in the future. Dynamic pricing is the antithesis of rigid, slow-to-implement pricing adjustments. Latency in making pricing decisions and rate updates can be costly for hotels.
Revenue managers have an embarrassment of riches when it comes to data they can feed into their forecasting models. Yet more data can also simply mean more noise. So what data is most relevant in terms of moving the needle and should be incorporated?
The volume and depth of clean historical data related to occupancy, rate and revenue figures (booking dates, rate codes, arrival dates, departure dates and revenue by day) tends to provide the strongest basis for forecasting accuracy. Market-level data, including competitor rate information, is also a must-have. Web shopping data (the number of consumers looking at and booking rooms and at what price) may also provide some insights into current and future room demand as well as price sensitivity. The number of website visitors tends to correlate to the frequency of last-minute arrivals.
More hoteliers are utilizing customer lifetime value (CLV) data, applying different pricing strategies to different customers with different lifetime values. With rules-based solutions, a revenue manager might set a rate for loyalty members to always be at a 10 percent discount to the best flexible rate. AI-powered solutions can adjust the discount dynamically, depending on its demand forecast. Add to the mix competitive rate data, demand data, multi-market economic data, and even air traffic, if desired.
In the end, revenue forecasting accuracy tends to be a matter of quality over quantity rather than the more the merrier.
Data Processing Power
Thanks to advances in data processing power, AI-powered revenue management solutions are able to process increasingly large volumes of data, and faster than ever. That’s a good thing, because combining all the data sets that a solution may use for just one hotel could amount to several hundred million observations.
Generating the pricing and distribution recommendations could easily result in thousands of decisions being generated each day for every day into the future. Multiply that number for a hotel chain with dozens of properties and it quickly becomes clear that, more than anything, revenue management is a big data challenge.
One major hotel brand recently revealed that it generates more than 45 million forecasts nightly for each hotel, segment, room type, and channel for the next 365-day period. While even the global hotel brands may not have data processing requirements that are in the same league as Amazon, Apple, Facebook or Google, their data processing needs are certainly large enough to stretch the limits of on-premise data storage and computing capacity.
Buyers need to know that any revenue management solution under consideration can handle the rigors of big data processing and optimize pricing calculations in highly compressed timeframes. Data processing power will continue to fuel the evolution the revenue management, including the shift to what is known as probabilistic decision models, which some experts believe will produce even better financial outcomes.
The Smart Decision Guide was independently produced, providing for unbiased, fact-based information. The research is based on data collected in Q2 2019 from more than 250 qualified survey respondents. The underwriters of the new Smart Decision Guide are the following industry leaders: Atomize, Duetto, Infor, Rainmaker (a Cendyn company) and SHR. The 2019 Smart Decision Guide to Hospitality Revenue Management is now available for complimentary download. It can be accessed here.
Logos, product and company names mentioned are the property of their respective owners.