How are the price recommendations calculated?

The algorithms used for forecasting and price recommendations in Atomize are unique, developed by an experienced team of data scientists in Sweden. These algorithms take a scientific approach to demand forecasting and pricing, which can be summarized as follows:

Demand Forecasting
To recommend optimal prices, the system first builds a demand forecast that predicts remaining demand at any point in time. This forecast also accounts for group allocations, cancellation probabilities, and other key data points.

Data Sources
The demand forecast is generated based on multiple data sources, including historical data, on-the-books (OTB) bookings, and current market trends, to forecast demand for the next 365 days (or longer). The forecast also incorporates expected cancellations and no-shows, both for individual reservations and group bookings. This information is combined with market intelligence to produce an optimal price for each room type at any given moment, up to a year in advance.

Atomize uses the following data from the PMS in its pricing calculations:

• Reservations
• Inventory
• Pricing
• Booking Pace
• Cancellations
• Group blocks and allocations

Additionally, Atomize includes competitor pricing and is in the process of adding broader market demand data to enhance its forecasting.

Each hotel’s demand model is trained on its historical data, ensuring that the pricing optimization aligns with the unique conditions of each property. This allows Atomize to offer recommendations that reflect the specific characteristics and booking patterns of each hotel.

Price Optimization
To determine optimal prices, Atomize simulates demand at different price points, calculating the ideal combination of rates across room types while considering room hierarchy and minimum/maximum price constraints. Recommendations can adjust in real time as new data emerges, helping maintain competitive pricing strategies.

Atomize optimizes each room type individually while taking overall hotel performance into account. You may occasionally see a recommendation for a particular room type that yields lower revenue than your current pricing; this is because the system identifies opportunities to shift demand to other room types, ultimately maximizing total hotel revenue.

Revenue Optimization
The primary aim of Atomize’s recommendations is to optimize Revenue Per Available Room (RevPAR). For properties that don’t rely heavily on ancillary revenue sources, Atomize may suggest a higher room rate, even if it results in slightly lower occupancy, based on analysis of price sensitivity. This approach maximizes revenue from each room type.

If your property generates significant ancillary revenue, such as through F&B or spa services, Atomize can factor this into its calculations, aiming to maximize Total Revenue Per Available Room (TRevPAR) per room night or stay. Contact support if this is the case for your property.

Recommendation Frequency
The optimization engine runs in two modes:

• Standard Optimization: Recommendations are generated three times daily at 04:00, 11:00, and 21:00 UTC and delivered to your property shortly thereafter. The exact time new recommendations appear in the app may vary slightly depending on the volume of recommendations processed by the PMS.
• Real-Time Optimization: For properties in highly dynamic markets, real-time optimization enables prices to be updated as soon as changes in the data are detected (e.g., reservation pick-up). This mode helps ensure that your hotel’s prices reflect current market conditions, providing timely adjustments in fast-changing environments. Standard scheduled optimization, on the other hand, incorporates all data changes since the last optimization.