Contextual add-ons and seat selection at checkout on travel pages

• WHO: Digital, growth & CRO teams  
• INDUSTRY: Hospitality, Travel & Recreation  
• IMPACTED PAGES: Mobile checkout, checkout page 

Experiment 1: Contextual bandit to surface personalized add-on offers at checkout

Hypothesis: Using Contextual Bandit to automatically surface relevant add-on offers based on real-time conditions (e.g., weather, time of day) will increase add-on attach rate and lift average order value.

Results:

• +12.4% Average order value uplift

Takeaways: 

1. Personalized add-ons increased average order value by 12.4%. 
2. Contextual recommendations feel more relevant to customers, driving stronger engagement with suggested add-ons. 
3. Most add-ons were low value items and impulse purchases. 

Experiment 2: Giving the seat map more screen space on mobile 

Hypothesis: If the seat map is given more screen space by hiding the order summary, then customers will find seats more easily because reduced clutter improves focus on the selection task. 

Results:

• Decrease in mobile drop-off rate

Takeaways: 

1. Seat selection view was redesigned to prioritize the map, reducing friction. 
2. Resulted in higher conversion rates, showing customers completed more purchases. 
3. Mobile drop-off decreased, confirming streamlined seat selection kept users engaged. 

Travel checkout is not one experience

A seat map on mobile. An add-on offer timed to the moment. A payment step that flows or frustrates. Each one is a test. The Experimentation Ideation Agent surfaces what to prioritize, and the Experiment Summary Agent tells you which changes travel across markets and which stay local. 

And here are more use cases from the Hospitality & Travel industry:

• Geo-personalisation and kiosk sign-in on hospitality experience pages