Key factors for demand forecasting
Understanding future demand is particularly relevant when the tourism sector is at a critical juncture.
The strategic management of the hotel business, and Revenue Managers in particular, need to understand future demand in order to adjust their decisions, so that they can improve factors such as occupancy, revenue or operational needs associated with demand at any given moment. This is particularly relevant at times like the present, where the situation in the tourism sector has taken a major turnaround as a result of the COVID-19 pandemic.
The objective of any Revenue Manager is to capture the greatest amount of revenue for a finite number of accommodation units, which is why it is important to have the ability to calculate forecasts and anticipate demand needs as much as possible.
Sometimes, when projecting a point in time into the future and trying to anticipate demand in a given market, it can seem like making decisions based simply on experience or intuition. However, RMSs now help to interpret historical data, include forward-looking market data and propose the ideal situation for your accommodation within a given context. In situations of absolute normality of the market, and without factors that significantly affect demand, there are usually no major variations. Although it is always advisable to be alert to possible changes, it is precisely now when we must be more alert as demand is subject to sudden changes due to the health crisis situation.
With this in mind, let’s look at five factors we should always keep in mind when updating our demand forecasts, as they can affect demand behaviour:
- Sales conditions of every rate
- Segmentation strategies
- Trend deviations between historical and current data
- Lead time and booking pace
- Agility in the face of specific variations in demand
Sales conditions of every rate
A common mistake in forecasting demand is not having a proper constraint policy. Unrestricted demand is the actual demand for your product without taking into account the natural limitations of a property. Regardless of the actual (or perceived) constraint, with higher actual demand you can set higher price points, which maximises the revenue potential for the same number of rooms.
We should not instinctively focus our demand forecasting on the number of rooms available alone, as this would not give us visibility of total demand and limit the ability to maximise revenue.
At a time when the hotel sector is in the process of recovery from the COVID-19 crisis, it may seem pretentious to approach a demand forecast from the perspective of overall demand expecting it to exceed the hotel’s capacity. However, being able to capture this broad view of total market demand will give us a better understanding of the environment and the behaviour of our clients and therefore we will be able to offer them a better service at the most appropriate price, maximising our results.
A winning bet, nowadays, consists of data-driven product segmentation and personalisation. An interesting position could be to take advantage of the information generated by our RMS system to determine which traveller mix is sufficiently interesting, in terms of the revenue it generates, to support the personalisation of your product or the channels that need to be used to attract them.
Therefore, not having data is not only a weakness in recognising new business avenues but also a factor in identifying changes and improvements that we can implement.
Trend deviations between historical and current data
While very useful in establishing a baseline for future demand thinking, historical demand data is not always the best indicator of what final demand will actually be. In other words, if you think about what happened last year and, without accounting for what is happening in the market at the present time, you will result in a biased demand forecast.
This is especially relevant in the post-COVID era, where historical data has lost relevance, and it is necessary to incorporate new sources of information to be able to make a more accurate demand forecast.
The Machine Learning algorithms implemented by some tools, such as Beonprice, make it possible to adapt and balance the historical weight on the rest of the factors, adapting the historical data to the current context. As an example, it is important to apply a greater weight to the historical weight on long-term work while short-term forecasts are strongly biased towards current market conditions. In addition, the forecasting process has to be adapted to the new macroeconomic environment.
Lead time and booking pace
There are a large number of factors that can affect demand forecasts, such as economic, social, weather, etc. These factors may be circumstantial or they may respond to a deeper trend change. Technology has directly influenced these consumption habits in a very important way. A representative example can be found in bookings made via mobile devices, which are usually made close to check-in dates and even on the same day. Planning is being replaced by the immediacy provided by these devices. Therefore, the improvised booking factor emerges to the detriment of a more planned distribution in the face of a booking made further in advance.
Again, the pandemic has accentuated the change in these consumer behaviours, and demand has become highly volatile and unstable. Uncertainty reduces lead time and encourages last-minute purchasing decisions. We need to keep this in mind when analysing trends and observing our booking pace.
Agility in the face of specific variations in demand
A demand stream is characterised by travellers who are staying at the accommodation on a one-off basis through significant events. The information provided by this traveller may or may not represent anything, but we may also have a recurring customer for an event that will recur regularly in the future.
Each of these customer segments, or demand streams, has its own booking behaviour, and possibly even a product or fare that is specific to its typology. Anything less than being able to adapt to change will be a clear sign of loss of competitiveness.
Revenue Management should be considered as a mixture of art and science as we have to be able to interpret the situation and make quick and effective decisions. Taking advantage of and distributing unrestricted demand, reducing to a minimum the loss of sales opportunities, will be the key to optimising resources and revenue.