Back before the days of the internet, direct mail marketers used to perform tests on their campaigns before committing to printing and mailing the final copy.
Modern marketers have the luxury of doing the same thing on a much larger scale thanks to the internet.
And since the internet allows you to perform tests in real time, you can use it to switch up your strategies at the drop of a hat by split testing your content.
Split testing isn’t just limited to testing your web pages. You can even split test your social media posts to get the most out of your social media ad spend.
This can save you from wasting money on posts that just aren’t performing well. Research shows that testing can also help you increase ROI by at least 30%.
Here’s how to split test your social media posts on LinkedIn, Twitter, and Facebook by changing one variable at a time to see how to affects your metrics.
Why split test on social media?
When you first got started with social media marketing, you probably based your strategy on general statistics.
For example, you may have started posting more photos on Twitter after hearing that tweets with photos usually get 35% more retweets.
Or maybe you only post on Facebook on Wednesdays at noon after reading that global engagement on the platform is high at that time.
But how do you know that your strategy is the best one for your specific audience and brand? Split testing on social media can help you figure this out.
You can use split testing to figure out how to improve just about every piece of your social strategy across all social networks.
This is important because a strategy that works for your follower on Facebook may not be the best approach for your Twitter audience.
With social split testing, you can find out which messages are most engaging to your followers, what type of content pulls in the most engagement, which call-to-action converts best, and more.
That way, you’ll know which types of posts are worth paying to promote.
Once you find the best formula, you can continue to test even smaller variations of each element to make sure that your content is at its best at all times.
Here’s how to split test on LinkedIn.
Split testing on LinkedIn
The first step to any split testing experiment is to decide what element you want to test.
For example, if you want to see what the best targeting strategy is, create different campaigns that are completely identical aside from the target audience.
You could set one campaign to target entry-level professionals and set the other one to target junior-level ones.
Or you could target a few different locations.
If one post gains more comments or likes than another, use more of that variable in the future, whether it be a specific location, audience, or type of content.
LinkedIn recommends letting each of your split tests run for a minimum of two weeks before concluding a result.
You can perform split tests on LinkedIn with the following types of content:
- Sponsored content: sponsored content includes campaigns promoted via paid channels based on posts that you have already put on your company page
- Direct sponsored content: this kind of post doesn’t appear on your company page, but will appear in the regular LinkedIn feed area
With direct sponsored content, you can more accurately split test whether or not specific audiences are responding well to variables in your posts.
Instead of sharing a company post, sponsor your content directly in Campaign Manager.
To get started creating a direct sponsored post, click “Create Sponsored Content.”
Collect the overall results from each of your split tests to gain more insights about how to adjust your LinkedIn content strategy.
Here’s how to split test on Twitter.
Split testing on Twitter
On Twitter, you can monitor changes by split testing different variables and tracking your growth rate between each change.
For example, if you send the same tweet out twice but include a photo with one of them, you can track which one performed better over time.
This will help you determine if your followers prefer tweets with photos or not.
Here are some examples of a few split tests to try on your tweets:
- Tweet on a different day
- Tweet at a different time
- Hashtags vs. no hashtags
- The length of the tweet
- @ mentions vs. no @ mentions
- Use of media like videos, GIFs, emojis, etc.
- The type of URL used (a shortened Bitly URL vs. a full URL)
You can use Twitter Analytics to look at your 28-day summary. View “per day” or “per tweet” stats to easily uncover which specific variations are working better than others.
Here’s how to apply the same approach to Facebook.
Split testing on Facebook
To get started split testing on Facebook, decide if you want to make changes to your headline, copy, or type of media used.
Then, isolate the variable you want to test by creating a few posts that are the same except for the element you want to test.
If you want to split test your post’s copy, for example, create posts with the same content. Then, give each post its own unique copy.
Once you determine which post performed better in terms of likes or shares, you can create more posts in the future with a similar copy.
You should also be sure to put your posts in front of a broad audience. That way, you’ll have a clearer picture of how they’re truly performing.
And you’ll also be able to determine a more accurate target demographic.
Instead of promoting posts to a small demographic, like women between the ages of 30-35 within ten miles of your business, put it in front of a broader range of ages and locations.
Create several posts that target different audiences to determine which group of people responds best to your content.
Then, send out your future posts to that group of people for higher engagement. Try messing around with other variables like audience size, hobbies, gender, etc. too.
Run split tests during your regular publishing times so that you can gauge interest more accurately.
And don’t send out more posts than your audience is used to.
If you only post two or three times per day but then start posting every hour to split test content, you’ll overwhelm your current audience.
Thanks to the internet, today’s marketers can split test just about anything online, including their social media content.
And those tests can help you ensure that you get the most bang for your buck when it comes to spending money on ads across different platforms.
By changing one variable at a time to find out which version of your posts perform best, you can better utilize your social media ad spend to gain even more engagement that can build your brand.
Test elements like photos, headlines, and the type of copy used to determine what your brand’s formula for a perfect post is.
Track different criteria like location, type of operating system used, type of browser used, and more for a more accurate outlook on how your social media split tests are performing.
However, if you’d like to do split testing on your website, Freshmarketer can help you with just that.