What next: stepping into the future with new forecasting systems

Hotel companies have to be accurate with their forecasting process in order to consistently evaluate business demands and come up with the corresponding pricing recommendations.

Today this can be done using automated updating of forecast patterns. However, from a results’ perspective, it isn’t only a hotel company that has to be well equipped for assessing its performance on a daily or a weekly basis. The system also has to be integrated with property management and central reservation systems, so that data can be managed across disparate systems.

As part of their role, revenue management (RM) executives have to evaluate historical stay-pattern data and booking-pace information by market segment. They work out arrivals for specific days, for each length of stay, market segment, and special events and so on.

Here EyeforTravel’s Ritesh Gupta talks to Dhiraj Trivedi, director - group e-marketing, The Park Hotels, about the significance of accurate forecasting in assessing the efficacy of the information.

EFT: Can you explain what accurate forecasting is all about and why it matters to RM executives?

DT: Only with an accurate forecast can one make correct decisions. For instance, the forecast says the hotel will be 50 percent occupied, but then the hotel ends up at 90 percent. All decisions would have been based on the 50 percent forecast, i.e. rates are set too low and business is accepted at much lower discount rates. There is also an impact on the operational side, too. There might not be enough staff in the house to clean the rooms, at check-in or in the restaurants, or enough amenities or food etc.

EFT: Is there any external development that you think is currently impacting forecasting in a big way?

DT: The big development in forecasting and revenue management has been automation. Many hotel companies are rolling out some sort of automated forecasting/revenue management system. The reasons for this are increased complexities in the market place, lack of trained manpower, as well as the ever-increasing power of analytics in computing.

EFT: What benchmarks have been set for assessing the success of forecasting?

DT: In an automated environment, one should look at forecast performance, which in essence is measuring how good the forecast is performing by taking the inherent volatility of the market out of the picture. In a manual environment – as forecast performance is a bit more difficult to calculate –many companies look at forecast accuracy, i.e. comparing the forecast to the actuals.

EFT: How do you process information to arrive at a particular decision?

DT: The data that one should be looking at would be actuals as well as OTB- pace information. This information should be looked at for all segments, as each segment behaves differently and should include information such as no shows, cancellations, etc.

EFT: When it comes to data extraction functionality, what do you look for?

DT: It’s worth considering established interfaces with many PMS providers. The extraction is simply a CVS file from the PMS that is put into a certain location on a server. This process is fully automated and doesn’t need any manual intervention. The extract includes information on actuals and OTB, plus room revenue, cancellations, no shows, departures, arrivals, out-of-order rooms etc.

EFT: What role do tools play in accurate forecasting?

DT: A few of them would be:

1) A system can do a forecast every day for the next 365 days by market segment group by day, which is impossible in a manual environment.

2) With advancements in analytics technology, a system forecast is more accurate than a manual forecast and reacts to changes in trends much faster than in a manual environment.

3) With an automated forecast, a revenue manager can now concentrate on his or her actual role, i.e. strategic decision making instead of spending hours on simply coming up with a forecast.

EFT: Can you explain what automated updating of forecast patterns is about?

DT: In today’s environment, no season is the same as the past, and things change very quickly. This is very hard to predict in a manual environment and sometimes the changes are too subtle to manually detect, whereas with the advanced algorithm in an automated tool, patterns will be detected faster and then adjusted to be able to deliver an even more accurate forecast.

EFT: What sort of customised reporting functionality do you consider?

DT: One can set up email daily/weekly/monthly information. Additionally, revenue management systems (RMS) offer a data-extraction programme, which allows the user to get all the data that is in the system exported and used as the data input for other internal reports.

Related Reads

comments powered by Disqus