Rounded/square corners - SeaWorld Orlando website

Last registered on December 10, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

Rounded/square corners - SeaWorld Orlando website

RCT ID

AEARCTR-0017385

Initial registration date

December 02, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

December 08, 2025, 12:42 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

December 10, 2025, 5:02 PM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

United States of America

Region

Primary Investigator

Name

Ron Kohavi

Affiliation

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Ravikiran Rajagopal

PI Affiliation

United Parks & Resorts

PI Name

PI Affiliation

PI Name

PI Affiliation

Linowski Interaction Design

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2025-05-19

End date

2025-06-30

Keywords

Behavior, Lab, Other

Additional Keywords

Online, website, user interface

JEL code(s)

C12

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Do elements with rounded shapes enhance click-through rates?

This is part of the Trustworthy A/B patterns project described at https://bit.ly/trustworthyABPatterns

The first two replications were done with Coop. The OBS trial was approved: https://www.socialscienceregistry.org/trials/17349
The second is at https://www.socialscienceregistry.org/trials/17384
This is the third replication on a US site: SeaWorld Orlando

Consumers frequently encounter digital design elements intended to influence their behavior, from call-to-action buttons to visual cues that guide online decision-making. A recent study published in the Journal of Consumer Research (Biswas et al. 2023) reported that using rounded visual elements increased online click-through rates by 55% in an A/B test—a finding that, if reliable, would have significant implications for marketers, designers, and digital platforms seeking to enhance consumer engagement

We suspect that the paper suffers from the winner's curse and is highly underpowered. See post at https://www.linkedin.com/posts/ronnyk_do-elements-with-rounded-shapes-enhance-click-through-activity-7183554027773734914--ugr/

Paper reference: Curvy Digital Marketing Designs: Virtual Elements with Rounded Shapes Enhance Online Click-Through Rates co-authored by Biswas, Abell, and Chacko (2023)

External Link(s)

Discussion of paper on LinkedIn

The paper: Curvy Digital Marketing Designs: Virtual Elements with Rounded Shapes Enhance Online Click-Through Rates

Trustworthy A/B Testing project page

Registration Citation

Citation

Kohavi, Ron et al. 2025. "Rounded/square corners - SeaWorld Orlando website." AEA RCT Registry. December 10. https://doi.org/10.1257/rct.17385-1.1

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

SeaWorld had square buttons throughout their website experience, and we were able to programmatically change the buttons to round in the Treatment. This included buttons everywhere, including add-to-cart, the popup asking for email address, and dates buttons.

See https://bit.ly/roundedCornersSeaworldSummary for description with screen shots

Intervention (Hidden)

Intervention Start Date

2025-05-19

Intervention End Date

2025-06-30

Primary Outcomes

Primary Outcomes (end points)

Clickthrough rate

Primary Outcomes (explanation)

Each user who visited the site was assigned 1 if they clicked on any button that was programmatically controlled as round or square, and 0 otherwise (Boolean).
The clickthrough rate is the average over all users of the 0/1/indicator.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Users who visited https://seaworld.com/orlando/ were randomized into control/treatment.
They're assigned 1 if they (anytime during the experiment) click on one of the buttons that was programmatically controlled as square or round and zero otherwise.

Experimental Design Details

Randomization Method

Every user who visits the website is randomized into one of the two variants by Optimizely based on a hash of their cookie.
The assignment is consistent for all future visits.

Randomization Unit

User

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

No clusters

Sample size: planned number of observations

At least 2.4M users for both arms

Sample size (or number of clusters) by treatment arms

1.2M users in each arm

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

MDE was set to 0.5% on estimated conversion rate of 35%. Using 16*(0.35*0.65)/(0.35*0.005)^2*2 = 2377143 We estimated 4-5 weeks, so agreed to run for 6 weeks to be conservative.

Supporting Documents and Materials

Documents

Document Name

Trustworthy A/B Patterns

Document Type

proposal

Document Description

This is the version of our document published at https://bit.ly/trustworthyABPatterns, as it existed on Google Drive on 5/6/2025, before the experiment started. The key paragraph proposing the replication is:

Rounded/square corners – published result that some believe is highly exaggerated. See https://www.linkedin.com/posts/ronnyk_do-elements-with-rounded-shapes-enhance-cli ck-through-activity-7183554027773734914--ugr Proposed MDE for power calculations on conversion metric: Based on the results of the Coop experiment, we recommend an MDE lower than 0.56%, say 0.5%. This is a tough one to estimate, but several practitioners believe this is much closer to zero than to the 17%+ in the paper.

Relevant experiment by Evidoo: https://www.linkedin.com/posts/jurjenjongejan_ecommerce-abtesting-activity-7275790 931121901568-nXh7

File

Trustworthy A/B Patterns

MD5: ba3e8d9513afe8a81a7735db46757dee

SHA1: b5726c89419d6751cb990cfbaeccaa93cbe2583d

Uploaded At: December 02, 2025

IRB