What Data is Most Valuable To Hotels When Forecasting?

“Pre-pandemic and post-pandemic forecasting differ a little in the sense that past data is not really relevant anymore. Past data and segmentation have been the foundation for so many revenue managers and sales teams I know and they used to be an important factor to be taken into consideration for forecasting.

Forecasting has also become even more important nowadays as many hotels are running with reduced staffing and this will allow operations teams to better plan their schedules.

Right now, we are dealing with a very volatile market, and trends are not consistent at all. Travel restrictions are being constantly changed/ updated – which also affect the consumer behavior. Therefore decision-making in terms of pricing and strategy can be a challenge for many revenue managers these days.

I can’t say that there is one most important piece of information for forecasting. Revenue Management is the efficient utilization of data from different sources to be able to take correct pricing and forecasting decisions.

The basics, of course, stay the same: on the books, demand, future trends, lead time, pickup, length of stay, holidays, special events, general seasonality, nationality seasonality, segments, group business, corporate business (mainly for city hotels), search traffic for the destination / for the area where the hotel is located, brand website traffic, OTA insights on market and consumer behaviour, day of the week vs weekend demand / trends etc.

Forward looking data and information related to the pandemic has become an even more important aspect for forecasting these days.

• Observing how different markets adapt to travelling with the pandemic – does it affect LOS, Lead time, etc?
• We are also taking into consideration additional data like travel restrictions, visa requirements etc related to the pandemic.
• Monitoring cancellations.
• Has the main market changed? If the principal market has shifted to domestic or neighbouring countries, how has this affected ADR and for how long will it continue to be like this?
• Are your Top Accounts changing? What does this represent for your rates and occupancy?
• How are travel restrictions changing for the destination in the next months? Which countries will be able to travel to the destination? Or are people preferring to travel domestically/ in their own region as well?
• Are people preferring destinations without quarantine? Does the destination have a quarantine upon arrival?
• Are people booking longer stays, and prefer more spacious rooms/ suites or villas?
• Are people willing to spend more, stay longer and splurge on higher room categories (since they might not have travelled for a long time) – there are many surveys available about feedback from people around the world.
• What are the upcoming trends: Country searches, Flight search data, website analytics etc.
• Any changes in demographics?”

“Unlike pre-COVID, the most important factor to use for your forecast is the current demand pattern experienced over the last 21-28 days. You cannot use last year or even the previous year’s data as the status quo has changed. Day of week data is still very relevant under the current demand pattern, as well as the behaviour by market segmentation groups. Public / Bank Holidays and events still play a role.

It may not be as significant as before, but it will generally have a bearing on your forecast. An influencing factor that is new to our forecast module that now plays a role is travel restrictions from your own country and supporting markets, as well as COVID case numbers and experts’ opinions on new waves of cases, etc.”

“Historical data (Same Point In Time and Past Results), Events Calendar, On The Books, Booking Pace and Segmentation. From a top-level perspective, being able to see the Booking Pace is the most useful information. This allows you to see when and how fast bookings are coming for any date. and by the time you track the data, you can compare and forecast dates. With Pace information, you can split to have this per segment, per event date and room category and any other “micro” analysis.”

“In these COVID transitional times, we can no longer rely on our past data (previous pricing strategies, historical demand trends from OTB to achieved revenue), and we need to concentrate on future demand data, looking at the current situation and understanding the demand trends.

When looking at On The Books it’s really necessary to look at what is performing well by segment, assessing new customer behaviour patterns, utilising customer database and social media to assess current and upcoming needs in order to promote at the right times to secure the demand required.

Look at current statistics on bookings, cancellations, changes to assist in predicting the reliability of the OTB data. Market intelligence, rate shopping of competitors, understanding airlift factors (for Island resorts) all have an impact on the potential future demand as it’s necessary to analyse your market and both internal and external factors that may impact.

The data accuracy in the systems is important, especially as we need to prepare scenarios to ensure that everyone is prepared for potential impacts and can react quickly. All the information that we assess in creating a forecast is important, otherwise we wouldn’t spend the time on gathering and analysing. However, ensuring the current data OTB is accurate would be a priority for starting assessing future demand.”

“Forecasting depends on the relevance and availability of true historical data & the period that needs to be forecasted. It is very important to develop a forecast process for accurate forecasting. Unknown market conditions, demand and supply, price sensitivity and comp set rate fluctuations are all taken into consideration whilst forecasting for a specific period.

Forecasting has always being a mixture of personal motivation and interest followed with some mathematical process that analyses large amount of data. It is also very important to understand the purpose of the forecast and what you want to achieve from your forecast numbers.

Special events need to be taken into consideration whilst forecasting, looking at same time last year and required budget numbers. Revenue Management Systems nowadays have forecasting tool in-built to help revenue managers/directors to accurately forecast. It’s always Garbage in – Garbage out. The more accurate data entered in the RMS, the more accurate forecasting results you will get.

From my point of view, when I forecast for a hotel/resort, I also share with the relevant teams in order to understand how we can achieve those forecasted numbers. For example, if there is any promotional offer needed to be done in order to get to that forecast figure or any incremental gain that can help contribute towards the bottom line figure.”

“At this point, the short term forecasts (next 4 weeks and below) are driven by historical performance with the difference that only the last few weeks are taken into consideration (rather than comparison to previous months or even years). The pace and pick up of the last few weeks, together with government policy announcements on restrictions and measures. are the most important indicators right now.

This will determine access to the destination and if/how many people can meet and possible labour restrictions. Forecasts beyond 3-4 weeks are very fluid at the moment. Overall assumptions need to be made based on government policy.”

“Forecasting has changed dramatically post-covid as historical data has lost importance and the focus is on new trends: search intentions and conversions (how many visits and searches on our BE or website vs conversions), which geo sources (travel restrictions and travel corridors), which air routes are opened and flight maps, what type of rate plans are booked (flexible vs promo vs prepaid), segment and target analysis (from individual segment to target couples, families, solo travellers only) and obviously the booking window.

The history is not completely forgotten, but for example, I tend to use very close periods of reference like the past weeks and/or months in order to trace a trend. It is also great to use the benchamark comparing the growth curve with the market and how far we are from the pre-pandemic indexes.”

“Recent events have highlighted that relying on historical data for forecasting and pricing was far too risky, so revenue managers had to adapt and change their forecasting methodology. We should rely more on forward-looking data as opposed to historical data. Since we can’t rely on the past trends when forecasting, we can adopt Short-Term Forecasting instead, by putting heavier weight on short-term history. Apart from tracking and analysing the performance from the last two to three weeks, we should also consider:

• Global and national travel restrictions, lockdowns, vaccine distribution and social distancing measures
• Air travel levels
• Changes in visitor behaviour, spending power and expectations
• Regrets and Denials
• Search trends and Google analytics
• Booking pace and lead time
• Analyse groups pipeline, calculate the conversion and allocation wash
• Market trends
• Demand-driving events”

“In pre-covid times, forecasting has been defined as the analysis of historical and future data in order to predict demand for each day of the year and then make wise strategic decisions on prices, commercial offer, staff management, services to be offered and costs optimization.

The pandemic has challenged this pattern, yet the analysis of a mix of historical and future data remains critical. What can change is the type of data taken into account. If before covid the historical data (occupancy rate, ADR, segmentation by channel, room type etc.) taken into account was related to the same day previous year, during the pandemic it was more related to same day previous week, if not the day before.

However, in the case of countries returning to almost normality thanks to vaccinations, the historical data of great importance can be the same day of 2019 or 2018, the last normal years before covid. Especially in the case of events that were held in the pre-covid era, cancelled or postponed during the pandemic, and now gradually resumed.

In the case of future data, before covid, we considered, in particular, the pickups, the occupancy on the books, the average booking window, the speed of reservations entry, events, and a general overview of competitors. Today, in times of pandemic, we inevitably tend to take a look at medical and scientific parameters such as the Rt index (which determines the expansion or regression of the pandemic depending on whether the value is higher or lower than 1, that is, how many people are infected on average by a person), the vaccination rate, the hospitalization rate, the case fatality rate, along with seasonal, environmental, social and political factors that may have a direct relationship with the evolution of the pandemic and therefore with the volume of demand foreseeable for a certain period in time.

And this analysis must be carried out both within one’s own micro or macro-destination and in all those destinations that can represent source markets.
Additionally, it’s important to consider course the level of supply in a given territory, i.e., how many and which hotels are temporarily or permanently close due to covid and how many new openings will replace the closures.”

“I consider the following to be some of the most important data:

• Pick up trends
• Same point in time data ( OTB vs how much pick up LY)
• Pricing of competition vs your own pricing
• Event calendar including details on attendees / guests/ visitors
• Groups on the books ( definite): wash vs pick up expected
• Groups on the books ( options)

But with all this data and the way the landscape has changed during the past 2 years of Covid, it all comes down to how do you interpret this data, and focus on the most recent trends, e.g., what was pick-up like during the last past weeks, when do we see a recovery, and what are the latest restrictions not only on your location but also on the locations of your top client origin.

When the market recovers, look for new travel trends as the industry has changed. I would still always use the above data but with every forecast it comes down to what we do with this data and purchase systems (RMS) that can help you track abnormalities and support the forecast.”

“Following the recognition of the fact that the pandemic has made historic data useless for demand forecasting as the result of everchanging booking behaviours, the focus on forward-looking demand data has increased instantly and thankfully technology vendors have reacted to this requirement quickly.
As to forward-looking data, there are different options to provide an indication of demand for the future.

• Own business on the books, broken down into segments or channels
• Business on the books of the competitive set, broken down into segments or channels
• Demand prediction based on search behaviour from GDS, OTAs or search engines for specific destinations

Deciding what data to use for the forecasting will depend on the position, location or marketplace of a hotel. A hotel with a significant unique selling point will likely not be concerned much about other hotels’ forward-looking occupancy but instead will be more interested in the expected demand for the region as a whole and their own pace. On the contrary, for hotels in very competitive markets, it matters a lot what kind of business is on the books in their competitive set and how they compare.

When correlating different data from multiple sources at the same time, the transparency will probably be impacted. The more data is mingled for the generation of demand forecast, the more complex it becomes to understand the drivers of demand predictions.”

“The most important data are the pickup numbers as well as the past numbers. With the past data, patterns can be recognized which can be adapted to the following months. Also, demand analyses which are related to a market give important indications for an accurate forecast.”

“There are several points to consider while forecasting. One of the data points vital to revenue management is the booking/ reservations data. This pertains to when a reservation was received and when it was consumed. While forecasting, I prefer using the advanced pick-up method, wherein I consider all booking data (i.e., for complete and incomplete stay nights) and then run a combined forecast.

Other interesting touchpoints to keep in mind are the pace reports and the booking curve. Both give me a fair idea of my hotel’s performance as compared to the previous year’s.”

Ask a Question & Join Our Expert Panel

Would you like a question to be answered by our Industry Expert Panel? Or would you like to join our community of experts and share your experience, insights, and knowledge with fellow industry professionals? Via the buttons below you can submit a question or submit a request to become part of our expert panel.