seo split test result would removing social media icons improve clicks to the website

SEO Split Test Result: Would Removing Social Media Icons Improve Clicks To The Website?

Before you start: what do you know about SEO split-testing? If you’re unfamiliar with the principles of statistical SEO split-testing and how SplitSignal works, we suggest you start here or request a demo

First, we asked our Twitter followers to vote:


Here’s what other SEO professionals have to share about this test:

Mike Friedman, SEO Consultant |

Anytime you are removing links on pages site-wide, you are potentially strengthening the other internal links on the pages. Because of that, I would expect this experiment to show a boost in rankings at the expense of social activity.

Kevin Maguire, SEO specialist at OnMark

It’s not so much about “icons”, but rather the OBL (outbound link) count that would be reduced site wide. Removal of those links, should naturally increase the PageRank weight of the remaining page links. So yes, with the consolidation of PR, things should improve.
It is impossible to calculate how much of a significant impact removing them may have without knowing the % of any given pages OBL they represent. Also, the widgets and plugins that need to load can significantly impact page speed. It may potentially improve load times and provide a better user experience. If visitors do not use it, lose it.

Craig Lawson, SEO specialist at ClickReady Marketing

Lower on the page near the footer. Goal is to get traffic to your site. Social media should be a feeder in essence to your site. Blogs often have high bounce rates to begin with. Really want to keep people on your site and hopefully send them perhaps to a sales page on your site. So move them lower would be my vote.

Find out if our followers were right by reading the complete analysis of this test.

The Case Study

Happy day to all of you SEO pros out there; I’m excited about the results of this test, and I hope you’ll find it as interesting as me!

For today’s split-testing results review, we have a well-known SaaS programming languages runtime website. 

As is common, in their universal footer are social media icons linked to their social media profiles.


We hypothesized, by removing the social media icons, and any associated JavaScript, we could improve clicks to the customers website.

It seems simple, what do you think the end result will be? Read on and let’s find out!


The Test

The SEO experiment was set up and configured using our SEO testing tool, SplitSignal. This isn’t a large website (although it has high traffic) so we selected 262 for the test control, and 266 as the test variant, for a total of 528 representative pages. 

On all 266 test pages, we dynamically removed the aforementioned social media icons and any related JavaScript. The test ran for 21 days with a confidence level of 99% with 63% of the test pages being recrawled by Googlebot during the time period for testing).

The Results

After the twenty-one-day period, we observed a statistically significant, POSITIVE result for the pages with the social media icons removed. ð¤¯


We asked you earlier what you thought would happen. The enterprise SEOs at LOCOMOTIVE Agency have formed an esteemed opinion we’ll share below, but we’d love to hear your opinion in the comments of what you originally thought, and whether you think it could be something different?.

Overall, the test pages received 4.4% more clicks during the test period, a gain of 785 clicks.

LOCOMOTIVE Agency Analysis

It’s getting harder to surprise us, and as long-time technical SEOs, this result was one we would have anticipated for one of two possible reasons.

The first reason that probably had a correlative add-on affect is fairly simple and straightforward. By removing the social icons in the universal footer, it reduced the overall number of on-page links allowing more authority from incoming links to be passed to the now smaller number of links, giving a bit of a boost. However, this boost could have also impacted the control group in the same way, so while it probably did impact the results, it may have been canceled out by a similar rise in the control group.

The second, and likely most probable rationale for the observed result is by removing weight from the page in the form of small images and associated JavaScript, the overall total bytes of the page were reduced, and there were knock-on impacts to other elements of typical web page measurements like overall load times, contentful paints, and possibly others like total blocking time.

Because the performance in these areas would’ve improved in isolation for each test page, this is the most likely cause of the observable result we see.

What intentional or vestigial elements within your document object model could you remove in the next few weeks for a similar improvement? Iterate for the win!

Have your next SEO split-test analyzed by the technical SEO experts at LOCOMOTIVE Agency.

Leave a Reply