Designing the Future of Forecasting

In the face of constant changes in the hospitality landscape, forecasting stands as a compass directing businesses towards success. Forecasting has historically relied on market trends as well as personal intuition, which has been used to weight each data source, often without testing or validation. Today, AI can combine those elements more effectively to produce more accurate forecasts.

To make this transition easier, FLYR for Hospitality has added a new forecasting feature to its revenue planning offering that can revolutionize the way you create, manage, and share monthly forecasts with your stakeholders.

We caught up with FLYR's Maria Garbea, product manager and a former hospitality revenue manager, and Matilde Adan, product designer. Their inside perspective sheds light on how this new solution will benefit everyone in the organization.

Laying the Groundwork

"We do a deep-dive with our users into their daily tasks," says Maria. "We ask what they are struggling with, and what they are trying to achieve. Based on that information, we design solutions that make our users more efficient & effective."

"Once we understand the problem, my job is to determine how we can design a solution that works best for our users," says Mati. "Not only should it be easily usable, but it should help our users to do things in a more efficient way. The product has to make lives easier – and perhaps make their day more fun."

The Challenges for Forecasters

"The biggest problem was that they did not have all of their data in a single place," Maria explains. "Previously, customers had to export data sets from multiple sources into Excel spreadsheets. Then they would have to comb through previous forecast versions and recent trends in the data before compiling a new forecast version to share with stakeholders."

Additionally, revenue managers and others responsible for forecasting were facing these challenges:

• Multiple revenue streams that were difficult to track in advance
• The need to collaborate with multiple departments, having to align different goals, and coordinating efforts
• Getting to the expected forecast granularity in the allotted time

"Based on industry standards, the accepted error or deviation an RM's forecast can have against actuals is about 5%," says Mati. "Considering the few resources they count on to create it and the recent uncertainty in the hospitality industry, it's incredible RMs can work within those margins."

Disrupting Conventional Methods

"Our solution aims to produce the most accurate forecast," says Mati. "But it's essential for RMs to control the narrative of forecasts and what gets distributed. The RM will know when they want to purely follow the data versus when they want to enrich the system forecast with information that is only available to them."

"We want to make it easy for our customers to take their base forecast and feel confident in committing to that, whether on a monthly or weekly basis," adds Maria. "We're transforming what used to be a very manual process."

A Glimpse into the Future

The product continues to evolve as customer needs change. Maria notes the possibility of holding multiple versions of a forecast. "Our users want this feature so they can compare and then lock it in as their final forecast, which would not be able to be changed so that it can be used for accuracy tracking."

"We're also looking at enabling clients to visually customize their forecast views," she adds. "Different users have different needs depending on their properties and how they like to read data."

Cultivating Innovation

"We have a lot of freedom to create," says Mati. "We can base the work that we do mainly on customer feedback. Seeing people happily using the product and getting value from it makes my work very gratifying."

Maria cites the company's organization. "Our teams have the power to make decisions. It's not a top-down approach, so you feel empowered, and also accountable for the company's success. We listen, we build, and we iterate based on feedback."