NASHVILLE, Tennessee — Forecasting hotel performance often is described as both art and science, but data experts say the real key is prioritizing qualitative analysis and justifying assumptions.

These five tips for improving forecasts from the "Best Practices in Advanced Forecasting" session at the 2024 Hotel Data Conference underscore the importance of knowing the data and applying an informed perspective.

STR analytics manager Kelsey Fenerty said any revenue managers building out forecasts will need to make sure they can "explain it to your boss, to your client."

"Someone very smart once told me that you can do what you want with the forecast as long as you have a justification for it," she quipped.

Fenerty said analysts and revenue managers need to make sure they come at the forecasting process with perspective and not "be afraid to be the one on the phone call asking really fiddly, finicky questions."

Jacob Bruno, senior forecast analyst for STR, said it helps to visualize data to help "discern patterns and extract [them] as best you can." He added it also helps to be aware of industry-specific best practices for forecasting.

"Know the best modeling techniques to approach each situation and be able to mitigate problems," he said.

Both Fenerty and Bruno said it's also valuable to have a deeper understanding of the peculiarities and history of a specific hotel or resort when forecasting its performance, taking into consideration things such as regular seasonality. Changing the time frame when looking at historic data can give you significantly different outlooks on the forward-looking trend.

Fenerty said she generally favors a qualitative approach for forecasting and described how that generally works for STR while building out forecasts.

"We look at the data, and we take a bunch of notes on those trends," she said. "Then we call our teammates in that market and ask them what's going on. Then we reach out to clients and ask, 'Hey, are you seeing this? What are you worried about? What are you thinking about?'"

Forecasting in 2024 has some unique challenges from a historical perspective given how much the COVID-19 pandemic and its years-long fallout on hotel industry performance have skewed the data.

"We had a year and a half of truly outlier performance in all aspects of this industry," Bruno said. "So the question is, what do you do with that component? You could just delete it and remove it from your data set. Some models would be affected by that while others would not."

He said this outlier data has also caused some issues for automated forecasting systems, and hoteliers should be careful how they're training those systems.

Read more news on Hotel News Now.