Forecast AI

The intelligence powering predictive performance across the ProfitFirstTM platform.

Hunter Conference

If you attended the Hunter Conference and would like a copy of Cindy Estis Green’s 2026 Outlook, simply complete the form here to receive a copy of the presentation.

Forecast AI is Kalibri’s proprietary forecasting engine designed specifically for hospitality revenue performance. It produces Kalibri’s annual national forecast and powers predictive components across Hummingbird Commercial and Hummingbird Market— delivering forward-looking insights into occupancy, ADR, and RevPAR.

Built on decades of hotel demand intelligence, Forecast AI combines granular property-level history with forward market signals and machine learning modeling. The result is a modern forecasting methodology that is transparent in approach, rigorous in design, and uniquely built to reflect how demand actually behaves at the hotel and market level.

The Data Components

Forecast AI integrates multiple structured demand and pricing signals that collectively describe both historical performance and future market conditions. These inputs are engineered into predictive features and dynamically weighted across forecast horizons.

Star

NEW Enhancements: Event and Alternative Supply Signals

The most recent addition is the integration of one-time event demand forecasting and alternative lodging (like short term rentals) indicators as additional inputs that may influence performance expectations at a market or submarket level or even for individual hotels.


Revenue and Booking History

Property-Level Revenue and Booking History

The foundation of Forecast AI is detailed hotel-level performance history, including:

  • Rooms booked by date
  • Booking lead time
  • Length of stay
  • Rate category segmentation
  • Channel/source-of-business mix

This enables Forecast AI to model demand behavior at the most fundamental unit of hospitality forecasting: the individual property.


Forward Pricing

Forward Pricing Intelligence

Forecast AI incorporates forward-looking rate shop data for individual hotels from multiple publicly available booking channels. Future pricing signals are highly predictive of both ADR trajectory and demand elasticity.




On-the-Books Reservation Pace

On-the-Books Reservation Pace

For some partners, Kalibri integrates on-the-books data, providing real-time visibility into booking pace shifts that may influence near-term forecast accuracy. In testing our models, the data has shown that with the relatively short lead time and high cancellation rates, the impact and value of the on-the-books data is not material, but it is still considered as it may improve outcomes in select markets or timeframes.

Supply Pipeline Forecasting

Supply Pipeline Forecasting

Forecast AI models new hotel openings as forward competitive impacts — forecasting how incremental supply alters market occupancy and rate dynamics over time.



Market Drivers

Macroeconomic Market Drivers

Kalibri integrates forward economic projections including employment, GDP, and market health indicators. These inputs allow for structural adjustments based on long-cycle demand drivers refreshed monthly.

AI Models

Forecast AI is not built on a single algorithm — it is an ensemble forecasting system that blends multiple machine learning architectures depending on forecast horizon and signal type. Kalibri’s modeling framework includes:

  • Gradient Boosting Models – Optimized for high-frequency market responsiveness and nonlinear feature interactions
  • Additive Time-Series Models – Designed for structural trend modeling and seasonality decomposition
  • Exponential Smoothing and Transitional Blending – Provides stability across changing signal dominance.
  • Dynamic Model Weighting – Emphasizes different models and signals corresponding to different timeframe components

Forecast AI in Kalibri Products

Hummingbird Commercial

Hummingbird Commercial

Forecast AI delivers predictive hotel-level performance forecasting with unmatched granularity:

  • 18-month forward projections
  • daily-level demand forecasting
  • rate-category specific outlooks
  • forecasts for both the market, submarket, subscribed property and its selected comp set

Forecast AI also underpins the ProfitMixTM model, serving as the predictive base layer for forward strategy decisions around a hotel’s most profitable mix.

Hummingbird Market

Hummingbird Market

Hummingbird Market applies Forecast AI to market-wide forecasting, delivering a forward view of:

  • Revenue
  • Occupancy
  • ADR
  • RevPAR

Forecasts are provided at the monthly level across a 12-month horizon, with optional segmentation by rate category and channel.

Flag

National Forecast

Kalibri’s annual national forecast is built using a fundamentally different methodology than traditional market forecasters.

Rather than projecting macro performance top-down, Forecast AI generates forecasts bottom-up: starting with thousands of individual hotel-level forecasts and aggregating upward into submarket, market, and national outlooks.

This property-first architecture provides a structurally more defensible national forecast — rooted in real hotel demand behavior, not broad macro assumption layers.

On average, most U.S. hotels sit on at least $200,000 in untapped opportunity