Learn how to use A/B image testing on your hotel website widgets. Real examples, step-by-step setup, and best practices to increase direct bookings.
Find the image that gets more direct bookings.
Show different images to different visitors. Measure which one gets more clicks. Use the winner as your default. Repeat.
The widget text, offer and button stay exactly the same across all variants. The only thing that changes is the photo. Each visitor sees one version — you measure which image gets more clicks.
Variant A
Room interior — the classic shot
Variant B
Guests enjoying the hotel
Variant C
Pool / lifestyle shot
Real examples from hotels
These are the kinds of A/B tests that typically move the needle for hotels. Each example shows what changed, what won, and why.
Pool view vs room interior
A coastal hotel tested two images for their Image Campaign widget promoting a summer offer.
Why B won: Lifestyle photos (experiences, views, emotions) consistently outperform room interior shots. Guests book feelings, not square meters.
Seasonal vs evergreen image
A mountain hotel used a Call to Action widget with a "Book direct and get free ski pass" offer. They tested the image in December.
Why B won: The image matched the reason visitors were on the website — planning a ski trip. Relevant context wins. In July, the result would likely reverse.
Couple vs family imagery
A beach resort ran an A/B/C test on their Call to Action widget promoting an early bird summer deal.
Why B won: Most summer bookings at this property were families. Showing people who look like your target guest creates instant connection. Know your audience.
Food vs room for a breakfast offer
A boutique city hotel tested imagery on a Call to Action widget promoting "Book direct: free breakfast included".
Why B won: When the offer is the perk (breakfast), show the perk. The image should illustrate exactly what the visitor gets by clicking — not something generic.
How to set it up in 3 minutes
Open your notification settings
Go to your campaign → click the notification → open the Basic tab. Works on Image Campaign and Call to Action widget types.
Toggle on A/B Testing and upload Variant B
Scroll to the A/B Testing section, flip the toggle, and upload your challenger image. Optionally add Variant C. Save.
Visitors are split automatically
Each visitor is assigned a variant based on their IP and browser — deterministically, with no cookies. The same visitor always sees the same variant.
Open Statistics and read the results
Go to the notification → Statistics tab → scroll to the A/B Testing table. The variant with the highest CTR is marked with a green checkmark. That is your winner.
Understanding the results table
| Column | What it means |
|---|---|
| Variant | Which image was shown: A (main/control), B (challenger) or C (optional third). A is always your original image. |
| Impressions | How many times this variant was displayed to visitors. All variants should reach at least 100 before you trust the result. |
| Clicks | How many visitors clicked the image or the button when shown this variant. |
| CTR % | Click-Through Rate: Clicks divided by Impressions, times 100. The higher the better. The winning variant is automatically marked with a green checkmark. |
Best practices for hotels
Change only the image
Keep the title, offer and button text identical across variants. If you change multiple things, you will not know which one caused the difference.
Match image to offer
If the offer is a spa treatment, show the spa. If it is a sea view room, show the view from that room. Guests click what they see themselves in.
Retest seasonally
The winning image in winter is rarely the winner in summer. Run a new test for each high season and update your default image with the seasonal winner.
No credit card required. Setup in 5 minutes.