Hospitality Revenue: 5 Things That Break a Demand Forecast

Author: Dan Skodol

In a previous blog post (“Revenue Management Forecasting: Outrun Your Pace Group”), I talked about the differences between Business Forecasts and Revenue Management Forecasts. This blog post digs a little deeper into Demand Forecasts, one of the cornerstones of revenue management. It is critical for revenue managers to understand future demand in order to project occupancy, revenue, and operational needs.

 A hotel revenue manager’s job is to capture all potential revenue for a finite number of “perishable” hotel rooms. Key to capturing this revenue is the ability to create revenue management forecasts. And regardless of the tools used to create these forecasts, this revenue manager must be able to generate an unconstrained Demand Forecast.

At first blush, forecasting future demand might seem to be just this side of guess-work. Except for using historical data, and knowing something about future (and likely local) conditions or events, the rest is just wishful thinking…right? Well, the art and science of demand forecasting is a little more complicated – and interesting – than that.

Starting with understanding the true definition of demand forecasting, let’s talk about five things that can break a demand forecast.


1. Lack of Booking Accountability


A common error in demand forecasting is failing to determine the unconstrained demand.

Unconstrained demand is the actual demand for your product without considering the natural constraints of a fixed property. In other words, selling out 100 rooms when you have 100 rooms worth of demand is very different than selling out 100 rooms when you have 500 rooms worth of demand.

Regardless of real (or perceived) constraints, with higher actual demand, you can set higher price points, thereby maximizing revenue potential for the same number of rooms.

Revenue managers often instinctively focus on forecasting demand with constraints in mind – for example, the number of rooms available.

This mindset (or software algorithm for that matter) might be good for operational forecasts, but by limiting visibility into the total demand opportunity, you limit the ability to maximize revenue. This runs contrary to the philosophy of why we do forecasting in the first place.


2. Insufficient Data to Support a Highly-Segmented Forecast


At Rainmaker, we advocate for granular and highly-segmented forecasts, as long as the data is available to support them.

For example, your forecast tools or processes might provide a way to exclude any number of customer segments, but if you do not have reliable data about those segments, you might be better off leaving those segments out of your analysis.

However, if the data is available (or can be quickly obtained), data-driven micro-segmentation of demand across your hotel’s unique business mix will yield significantly better revenue results.


3. Imbalance of Historical Data with Current Conditions


While useful for establishing a future demand baseline, historical demand data is not always the best indicator of what final demand will be.

In other words, thinking about what happened last year, without accounting for what is going on in the market right now, will result in a skewed demand forecast.

For example, if a new hotel has arrived on the scene in the last year, this must be accounted for. (This is distinct from macro-economic conditions, discussed next.)

Algorithms in software-assisted forecasting must properly balance the historical with the contextual. As an illustration, consider the rules of thumb that say long-term forecasts are heavily weighted by history, while short-term forecasts are heavily weighted by current conditions.


4. Not Allowing Macro Conditions to Affect Long-Term Forecast


It's essential to allow macro-economic conditions to affect the long-term forecast.

While it is a good idea to ground demand forecasts in history, you must revisit this demand in the context of current and broad economic conditions. And this should be done both for world markets that put pressure on everything as well as local economic conditions that affect the demand conditions in your area only.

A prime example of macro-economic conditions was the global financial crisis that occurred some years ago. A global recession can affect bookings originating from anywhere on the planet, while local economic disturbances might affect only bookings that originate from your own neighborhood.

Strong revenue forecasting tools are very good at ensuring the above distinctions are not missed like they might be by human analysis.


5. Inability to Treat Demand Streams Differently


A demand stream is characterized by customers who stay at your hotel to attend specific events, who stay at different times of the year, or who represent desirable demographics.

Each of these customer segments (i.e., demand streams) has its own booking behaviors and price points.

Revenue managers who are unable or unwilling to tease out particular customer segment data, will do a less good job of forecasting demand accurately.


In Conclusion

The art and science of forecasting is sophisticated and requires a real grasp on the complexity and requires the ability to take all of the variables into account.

These five things are common ‘breakers’ of the forecast.

Demand forecasting is essential to revenue management. Proper unconstrained demand forecasting will help you to not lose revenue opportunity, the capture of which is the reason we do revenue management.

New Call-to-action 

Learn how Rainmaker’s revenue optimization solutions can help you

Request a Free Assessment and Get Started Today!

Recent Posts