Pattern #77: Filled Or Ghost Buttons

Pattern #77  Tested 6 timesFirst tested by Luke Hay Recently tested by Daria Kurchinskaia on Jun 20, 2024

Based on 6 Tests, Members See How Likely Version B Wins Or Loses And By How Much

LOSSES
-5
-4
-3
-2
-1
FLAT
+1
+2
+3
+4
+5
WINS

Measured by the sum of negative and positive tests.

A B
Filled Or Ghost Buttons (Variant A) Filled Or Ghost Buttons (Variant B)

Expected Median Effects Of B

?

Progression

(6 tests)

-

Leads

-

Signups

-

Engagement

?

Sales

(3 tests)

-

Revenue

-

Retention

-

Referrals

?

ANY PRIMARY

(6 tests)

Tested on

Tests

Pattern #77: Filled Or Ghost Buttons
Was Tested On Volders.de by Michal Fiech

Replaced

Isolated

Test # 307 on Volders.de by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Michal Fiech    Jul 17, 2020 Test link

Find Out How It Did With 53,618 Visitors

  • Measured by clicks on the secondary button

Zoom In
  • ?% (? successes out of ? visits)
A
  • ?% (? successes out of ? visits)
B
Thank You

This experiment measured a shallow click goal on a button that would encourage to repeated the action that was just completed (in this case a contract cancellation). In the control version (A) a thank-you screen shows a filled button style, and the variant (B) there was a ghost button. As a note, I also flipped the A-B in this experiment for the purpose of matching it to our ghost button pattern, which means that Volders in fact was starting  out with a ghost button to begin with. 

Get Access To See The Test Results

The Same Pattern Was Also Tested Here

Replaced

Isolated

Test # 538 on Volders.de by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Daria Kurchinskaia    Jun 20, 2024 Test link

Find Out How It Did With 56,025 Visitors

  • Measured by clicks on checkbox

  • Measured by completed purchases

Zoom In
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
A
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
B
Checkout

In this experiment, a less visible (ghost button style) legal confirmation box, was tested against a more visible one (filled state with higher contrast). Impact on error rates (from submitting an incomplete form) and sales was measured.

Get Access To See The Test Results

Replaced

Isolated

Test # 526 on Online.metro-cc.... by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Andrey Andreev    Apr 17, 2024 Test link

Find Out How It Did With 896,009 Visitors

  • Measured by completed searches

  • Measured by completed purchases

Zoom In
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
A
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
B
Global

In this experiment, a feint gray background search bar was replaced with a higher contrast one containing a blue border and a filled blue color button style.  Impact on searches and purchases was measured. (Inverted to fit the Filled-vs-ghost button pattern).

Get Access To See The Test Results

Replaced

Isolated

Test # 310 on Backstage.com by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Stanley Zuo    Jul 25, 2020 Test link

Find Out How It Did With 594,997 Visitors

  • Measured by visits to a detail page

  • Measured by membership sales

Zoom In
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
A
  • ?% (? successes out of ? visits)
  • ?% (? successes out of ? visits)
B
Listing

In this experiment, the style of a button leading to view detailed casting calls on a listing page was changed. In version the style was a filled high contrast blue background, and the the B variation there was a feint "ghost button" style. 

Get Access To See The Test Results

Replaced

Isolated

Test # 243 on Goodui.org by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Jakub Linowski    Jun 05, 2019 Test link

Find Out How It Did With 25,046 Visitors

  • Measured by visits to pricing page

Zoom In
  • ?% (? successes out of ? visits)
A
  • ?% (? successes out of ? visits)
B
Home & Landing

In this experiment, full red background buttons were changed to ghost buttons (red outline and transparent fill)

Get Access To See The Test Results

Replaced

Isolated

Test # 199 on Freshegg.co.uk by $conducted_test->test->user_->first_name . ' ' . $conducted_test->test->user_->last_name Luke Hay    Sep 13, 2018 Test link

Find Out How It Did

  • Measured by clicks on any of the primary buttons

Zoom In
A
B
Home & Landing

In this experiment, standard (filled) buttons were tested against ghost buttons.

Get Access To See The Test Results

Leaks

Leak #39 from Booking.com   |   Jan 3, 2020 Listing

Booking A/B Tests Ghost Buttons, But Their Hopes Fade Away As Predicted

A - Dec 18, 2019
B - Dec 18, 2019 REJECTED

Ghosts buttons. We know that this style of buttons is net negative based on this evidence-based pattern. So when Booking ran this experiment, it's no surpirse that it was eventually rejected - a predictable replication. View Leak

-0.5 Evidence

Leak #106 from   |   Apr 17, 2024 Global

Etsy A/B Tests This High Contrast Button Style

A - Oct 11, 2023
B - Oct 11, 2023 IMPLEMENTED

Etsy ran a little A/B test where they probably tried to encourage more searches with the help of a stronger visual cue in their search bar. Instead of using a  ghost-like icon only approach for the primary search, they stylized it as a high contrast button. Fast forward a few months and it now looks like the variation was implemented.  View Leak

-0.5 Evidence

For each pattern, we measure three key data points derived from related tests:

REPEATABILITY - this is a measure of how often a given pattern has generated a positive or negative effect. The higher this number, the more likely the pattern will continue to repeat.

SHALLOW MEDIAN - this is a median effect measured with low intent actions such as initiating the first step of a lengthier process

DEEP MEDIAN - this is derived from the highest intent metrics that we have for a given test such as fully completed signups or sales.