Hotels and Resorts
Machine Learning - Predicting Hotel Bookings Cancellation
Model Accuracy: 94%
Tags: #Hotel #Regression #Prediction #Indonesia #DataVisualisation
Background
Rising concerns about the Omicron Covid-19 variant has led to more hotel cancellations globally. This has impacted the Tourism industry's recovery from the Covid-19 pandemic. Over the past four years, hotel cancellation rate across all channels has risen by 6% (source). Moreover, hotel booking cancellations can represent around 20% of total bookings received by hotels (source).
The current business challenge is that once the reservation has been cancelled, there is almost nothing that can be done to salvage the situation. This creates greater adversity for hotel groups and fuels the desire for a solution that enable pre-emptive action. Therefore, predicting reservations that are likely to be cancelled ahead of time will give the relevant stakeholders significant agility in migitative measures to prevent these cancellations and ultimately maximise hotel revenue.
Solution
By setting up and optimizing on Machine Learning models, Analytico Asia generated a prediction model based on 73,270 bookings and achieved a model accuracy of 94%.
Using our model predictions, hotels can create an adjusted staffing plan based on the occupancy rate calculated from our model thus preventing high labour cost, revenue and occupancy loss