Hotel forecasting plays a vital role in strategic decision-making, shaping both the short and long-term success of a property or portfolio. By understanding the key components and dynamics of accurate forecasting, hoteliers can implement more effective strategies, optimizing efficiency and maximizing profitability.

This article is part of a series on hotel benchmarking data sets and use cases. Reach the other editions here: The hospitality industry’s historical KPIs | The role of profitability data in a comprehensive hotel benchmarking experience | Using business on the books in a complete benchmarking approach

The elements of hotel forecasting

Quantitative aspects:

  • Hotel metrics: Top-line performance indicators such as occupancy, average daily rate (ADR), revenue per available room (RevPAR) serve as a foundational element of forecasting.
  • Macroeconomic variables: Including economic predictors, such as Gross Domestic Product (GDP), inflation rates, and unemployment rates, can boost forecast performance by capturing underlying causes of performance movements.
  • Econometric models: There are multiple model designs for different types of forecasts:
    • seasonal (STL Decomposition, Holt-Winters)
    • classical (ARIMA and Multivariable Linear Regression)
    • more advanced methods (Gradient Boosting and Decision Trees)

Qualitative aspects:

  • Market insights: This incorporates non-numerical data, such as market trends and specific economic conditions from market analysts. This is important for factoring customer behavior and emerging trends.
  • Expert opinions: Engages insights from hotel management and operational staff to understand local market dynamics. Just as leveraging market analyst input is valuable, there’s no replacement for the firsthand insights provided by those directly involved in day-to-day operations. Their on-the-ground perspective can offer critical information that complements data-driven analysis.
  • Scenario planning: There are numerous situations, both economic changes and “black swan” events, where there is benefit to anticipating different market conditions. This serves to understand potential risks and opportunities that may not be evident through data alone.

Reasons for hotel forecasting 

Effective hotel forecasting provides multiple benefits, not only with strategic planning, but also at the operational level. These benefits are widespread and include:

Refined revenue management: Forecasting allows hoteliers to predict future hotel performance based on historical data, trends, and market conditions. Understanding these patterns allows hotels to optimize pricing and maximize revenue.

Improved operational efficiency: When hotels anticipate consumer demand, they can better manage staffing, inventory, and resources. This reduces operational labor costs by providing management with information that can be leveraged for smoother operations.

Better financial planning: Forecasting enables hoteliers to build revenue projections and anticipate expenses. This assists in better budgeting and financial planning by providing for more informed decisions about management decisions and operational adjustments.

Proper risk management: The hospitality industry is subject to various uncertainties, from economic fluctuations to changes in consumer behavior. Performance forecasting helps identify certain risks in time to develop contingency plans and hedge against future downturns.

Economic variables in hotel forecasting

Incorporating economic variables into hotel performance forecasting is crucial for creating a more accurate and comprehensive picture of future performance. Here are some key economic variables to consider and their role in forecasts:

  • Gross Domestic Product: GDP changes are strong indicators of the overall health of the economy. A growing GDP typically correlates with increased consumer spending on hotels and travel.
  • Inflation rates: Inflation affects consumer spending on hotels and travel. While higher-than-average inflation can lead to stronger ADR growth, it can also have a dampening effect on consumers’ willingness to spend.
  • Unemployment rates: Higher unemployment can lead to reduced discretionary spending on travel, which can have a negative correlation with hotel demand.
  • Exchange rates: More specifically for international inbound hotels, fluctuations in exchange rates affect travelers’ demand for certain markets or tiers of hotels. This can affect both demand and revenue for hotels.

Hotel data set components – trend and seasonality

When analyzing time series datasets in the hotel industry, understanding the concepts of seasonality and trends is integral for accurate forecast design. These two elements are part of the performance dataset, each shedding light on a distinct aspect of the overall metric being forecasted.

Trend:

  • The trend refers to the long-run movement or direction of performance over time. It indicates whether overall performance is increasing, decreasing, or remaining stable over an extended period.
  • Trends are influenced by long-term factors, such as shifting economic conditions, changes in consumer behavior, industry developments, and shifts in demographic patterns.

Seasonality:

  • Seasonality refers to patterns that are observable at regular intervals. These can be annual, quarterly or monthly patterns, and can be broken down further. It is not uncommon for a data set to demonstrate multiple seasonal patterns at once.
  • Seasonal patterns are often driven by predictable events or cycles, such as holidays, events, climate and local offerings. For example, ski resorts usually have strong winter performance, while family destinations show increased performance during summer months.

Sample size and its effects on forecast performance

The number of properties included in a forecast can influence results. Performing forecasts with larger sample sizes, such as at the chain scale and/or market level, more effectively capture traditional patterns such seasonal behavior and economic trends. Larger aggregation of properties in a data set reduces the impact of performance outliers or anomalies by averaging out the irregular data points from the overall set. Smaller sample sizes can be at increased risk to skew from statistical anomalies and can lead to less reliable prediction.

In addition to the increased performance of larger sample sizes, model decay tends to be lessened as the property counts included increase. Model decay – or drift – is the real-world application of conditions changing over time. It is normal and expected that forecasts should provide more accurate projections in the short-term than the long-term. The important takeaway is that forecasts should have a stable error that increase over time with little to no sharp changes.

Events

Events in particular cities, markets, or states can have profound effects on hotel performance, influencing everything from occupancy rates to pricing strategies. Accurate event forecasting is a crucial component of the overall forecast process and can be difficult given the uniqueness of each event. The effects of events on hotel performance include:

Increasing hotel demand and ADR: Large-scale events, such as conventions, festivals, sport matches, and concerts can significantly elevate demand for hotel rooms and can push up rates for all tiers of hotels. Some events, such as Taylor Swift’s Eras tour, were monumental for hotel demand and could affect the entirety of a city or submarket’s hotel performance during a given time.

Seasonal variations and the influence on consumer behavior: Certain areas of the U.S. host annual events that create predictable seasonal spikes in demand. While an event occurring at a regular interval is included in historical performance predictably, these events can boost both demand and rate performance compared to other areas.

Impact on extended-stay properties and group bookings: Events such as conferences or conventions often lead to increased group bookings. At the same time, longer events drive demand for extended-stay properties. This drives performance benefits from special offerings or services not affecting typical stays.

Conclusion

Hotel forecasting provides a useful tool for predicting performance on both a localized and industry-wide scale.  Utilizing proper quantitative approaches alongside valuable insight from analysts and hoteliers can assist with both improved operational and management decisions. Industry-leading methodologies incorporate long-term trends and observable, seasonal patterns for successful decision-making while accounting for market-specific economic conditions and events. Effective forecasting provides unique value for industry professionals and analysts alike.