How-to Guide for Split Testing Your Content Marketing Strategy on Twitter

Do you know how to give your followers what they truly want? Now you can use this test to accelerate the process bit-by-bit! Split testing, also known as A/B testing, is a strategy to finding what content sparks the most interest in your audience.

Essentially, the strategy works by testing a specific type of content on one portion of your audience while another section sees a slightly different version of that content. This test can be used for your website’s landing page, blog content, and emails, but today we will apply it to social media.

You can try posting two tweets with different hashtags and use analytics to see which one performed better by weighing how many impressions it received or how many people clicked the link.

After the two content variations are posted, you can take the most engaging method to add to your strategy and go from there. But you wouldn’t want to post the same things twice because that would hinder your split test results, so here is where Tweepsmap can help you segment your content, and we will see how to not double down on similar tweets!

 

Segmenting Your Content

 

We started by scheduling our content into different pipelines of information so we could keep track of metrics like how often we post about a specific type of tweet. So, for example, it wouldn’t be good to see 6 tweets about Instagram when you follow us for our insightful blog posts and #30DaysofTips threads.

 

Here is how we divide our content using pipelines:

  • Twitter News,
  • Social Media News,
  • Posts for activity in the Americas,
  • Posts for activity in Europe and the Middle East,
  • Posts for activity in Asia and Australia,
  • Our threads optimized for everyone.

 

Separating our content by time zones ensures that we post during everyone’s peak activity times with an international audience in mind. So it’s not a surprise that this method showed that our followers enjoy and prefer Twitter-related news over other news articles since we offer services for audience growth, and this news is helpful to our followers.

Since everyone’s audience is different, you can use pipelines in various ways to see what content works best for your followers! For example, we used pipelines to try better headlines for posting news articles. There was an increase in natural engagement when we personalized them with a fact from the article, a question that seeks our followers’ opinions, or even just rephrasing the headlines to be more straightforward for our busy clients. The possibilities are endless.

 

Tweaking Your Strategy

 

Once you have divided your content types, you can further separate your content by time zone. Segmentation enables you to post different variations of content to some users while others are inactive, and you don’t have to stay up around the clock posting day and night.

Because 6000 tweets get published a second, your posts will virtually fade away in moments allowing you to test posts between time zones with differing peak activity times.

Since Twitter puts all your content on one feed, they are visible to everyone. Doubling some posts is okay because they are fed to your followers through Twitter’s algorithm that does not work on linear time unless specified. Knowing this, you’re A/B tests won’t really bother your audience unless you post the same content within the hour.

 

Creating Pipelines & Schedules

 

To create split tests, you can make different pipelines for various content. For example, with Tweepsmap, you can make pipelines that attach to an RSS feed so new articles or content from your blog get added to your posts calendar automatically, letting your Twitter followers stay in touch effortlessly.

Optimizing your pipelines is easy. Just specify the posting times you’d like to apply to each. You can choose from anywhere between all week long or specific hours of certain days. If you can’t decide, our algorithm will automatically select the best times to tweet every day.

 

How to Measure the Results of Your A/B Test?

 

After segmenting your content into pipelines and configuring them to post for different time zones, you can now test content without spamming your audience!

For example, we tested a post about our recent YouTube video on 6 ways to get new Twitter followers. Each post shared the #audience hashtag and tagged @JonWesselink. In our case, we split tested different video clips in the posts. Note: posting videos to Twitter receives higher engagements than embedding a YouTube video.

The first one was about posting during active peak hours and the second one was about testing different content to see what works. We tracked the success of each post through our Engagement Dashboard, which allowed us to quickly see which one was more successful.

When split testing on Twitter, it’s essential to wait around a week to surely know the results of your efforts!

 

Our A/B Split Test Results

 

This quick test we ran, for example, shows you which video performed better in a tweet. Once again, waiting a week would bring better results so here is what we gathered from 24 hours of each post being live.

Below, after 24 hours, the first one received a reply, 5 likes, 4 link clicks, 9 profile clicks, and 14,000 impressions. Not bad for 24 hours.

The second version, after a day, received 3 likes, 14 profile clicks, and 15,000 impressions. That’s roughly the same amount of impressions with a slight increase in profile clicks.

This is where the fun begins. Depending on what you’re trying to improve, you can go with what works best. The first tweet talked about optimizing posting times for a higher outreach, while the second mentioned how to understand your audience by varying your content and also mentioned a free tool to use right now.

After some thought, it makes sense that the first post gained more engagements as it gives helpful info about optimizing posting times because valuable tips are what people want.

The second post had more profile clicks which might correlate to the free tool mentioned in the video, so people are inclined to click on our profile to find out more.

Although this was a quick test, you can see that we now have a better understanding of how our audience reacts to info so we can optimize our strategy depending on what we want to push for! Of course, there are no definitive answers, and perhaps this test would bring more insight after a week.

 

If you want to try this for yourself, sign up for Tweepsmap to see what’s possible! Also, watch the video we are talking about here on Tweepsmap’s YouTube channel.