What is Cancellation Risk?
Cancellation Risk is a predictive tool that helps you understand which bookings are more likely to cancel before arrival. It analyses your property's historical cancellation patterns and applies them to current reservations to give each booking a risk score from 0% to 100%
Think of it like weather forecasting - we can't predict with certainty whether a specific booking will cancel, but we can identify patterns that suggest a higher or lower likelihood based on similar bookings in the past.
How Risk Scores Work
Risk Levels
Each booking is assigned one of three risk levels:
🔴 HIGH RISK (20% or more) - Bookings with strong indicators that match past cancellations
🟡 MEDIUM RISK (10-19%) - Bookings with moderate cancellation indicators
🟢 LOW RISK (Below 10%) - Bookings with characteristics similar to those that typically stay
What We Analyse
The system looks at your property's historical data - specifically the last 2 years of bookings - to identify patterns. For each current reservation, we examine:
-
Booking Source & Channel
How was this booked? (OTA, Direct, GDS, etc.)
What's the guest segment? (Leisure, Business, Group)
Which source did it come from? (Booking.com, Expedia, Direct, etc.)
-
Timing Patterns
How far in advance was the booking made?
When is the arrival date?
What time of year is the stay?
-
Booking Characteristics
What room type was booked?
Has the booking been modified?
How many times has it changed?
-
Historical Context
What happened to similar bookings in the past?
How many similar bookings have we seen?
What percentage of them cancelled?
Understanding Your Scores
Why a booking might show 100% risk:
A 100% score means that historically, every booking matching this profile has been cancelled. However, this doesn't guarantee this booking will be cancelled because:
The pattern might be based on a small sample size
Past behaviour doesn't always predict future behaviour
Guest circumstances vary
Booking conditions may have changed
Example: If you only had 5 advance-purchase OTA bookings for a specific room type in winter, and all 5 cancelled last year, the system will flag similar bookings as very high risk. But this year's guest might be completely different.
Why a "guaranteed" booking might show high risk:
You might know certain things about your guests that the system doesn't:
Personal relationships with repeat guests
Advance deposits or guarantees
Special event bookings
Corporate contracts
Local context or guest communication
The system only knows what's in your booking history. It can't account for:
Prepayment status (unless tracked differently)
Guest loyalty or VIP status
Personal guarantees
Recent communication with guests
Best Practices
✅ DO:
Use risk scores as one indicator among many in your revenue strategy
Pay attention to high-risk bookings approaching arrival dates
Review patterns over time to understand your property's trends
Combine risk data with your direct guest knowledge
Use risk scores to prioritise which reservations need attention
❌ DON'T:
Treat any score as a guarantee (neither 0% nor 100%)
Cancel or release rooms based solely on risk scores
Ignore your personal knowledge of guests and circumstances
Assume the score knows information not in your booking system
Questions & Configuration Support
To adjust your property's risk settings:
Contact your Right Revenue support team with:
Your property ID
Which thresholds you'd like to adjust
Your reasoning (helps us give better recommendations)
Common questions:
Q: Can I exclude certain booking types from risk scoring?
A: Currently, no, but this is based on your feedback being considered for future updates.
Q: Why do some bookings show "N/A" for risk?
A: This typically means there isn't enough historical data for that specific combination of characteristics.
Q: How often are scores updated?
A: Risk scores are recalculated daily at 2 AM UTC using the most recent data.
Q: Can I see why a specific booking got its score?
A: Yes, each booking's risk details include the factors that contributed to its score.
Comments
0 comments
Article is closed for comments.