Run your own social media experiments in 4 key steps

Content creators understand the importance of a defined social media strategy, but how well do they actually know their social audiences? 

We know that social audiences interact with content in different ways than search, but drilling down and evaluating your promotion strategies is the best way to understand the tactics that will help you turn casual social visitors into loyal audiences

Below, we take you through our recent social referral data and provide some tips on running your own social strategy experiments in Chartbeat.

Social referral data: The current landscape

As we near the close of the quarter (ending Sept. 30), Facebook still remains the primary source of social referral traffic to your content.

In fact, Facebook has referred a whopping 7.8 billion pageviews so far this quarter across our global network, with Twitter a distant second at 738 million.

Facebook continues to dominate the overall referral landscape as well, coming only second to Google Search for weekly pageviews referrals, as we see below.

This is why it’s so important to run a few simple, data-informed experiments. In just a few minutes, you can help your team confirm or refute some of the long-held assumptions you’ve had about your social traffic, and more importantly, make optimizations that can have significant, positive implications.

Related reading: Decoding Facebook data to understand the true drivers of engagement, loyalty

How to use your data to run social media experiments

Now that you know the social platforms (likely Facebook and Twitter) driving traffic to your site, how can you analyze social reader behaviors to ensure you’re reaching these massive audiences? Let’s show you how with a little help from Chartbeat.

1. Set up your dashboard to see the frequency of your social visitors

  • Double pivot on “New” under Visitor Frequency and “Social” under Traffic in your Real-Time Dashboard. The top pages list shows you which content is doing the best job of attracting new readers to your site.  
  • Then, switch from “New” to “Returning” and check to see if there’s a difference in the content driving traffic from specific referrers. 

Now, it’s time for further analysis. Three initial questions you can ask of your data can include: 

  1. What social channels are referring readers to your site? Is it the usual suspects (i.e., Facebook and Twitter) or is there anything that comes as a surprise?
  2. (When comparing new and returning visitors) Are there stories that are resonating with readers that are more familiar with your site, but not necessarily drawing in the new audiences?
  3. Is there an alternative way to promote these stories to appeal to new visitors based on what you already know about your audiences?

Based on our global referral data, you’re likely to see that Facebook drives significantly more traffic than your other social referrals such as Twitter. While your Twitter or LinkedIn audiences could offer a high level of unique engagement, when it comes to getting readers onto a story, our data suggests that Facebook is typically the better choice.

2. Let’s challenge that thinking with a bit of experimentation. 

The next step is putting those assumptions into action. Options include:

  • Posting the desired content across your social accounts and seeing if that post, created the way you think offers the most impact, actually brought more readers to your story. 
  • If there is low engagement on your stories after a readers clicks the link, try a different approach. For instance, our data shows that using quotes or trying a different image with your post can help. 

Then, repeat. By running a few of these small experiments, you can build future best practices for engaging your social audience moving forward.

3. Experiment with timing.

As we’ve written previously, our research suggests that social traffic substantially underperforms overall traffic from the morning through to the early afternoon, and substantially outperforms overall traffic during evenings and nights.

when to post on social media hourly chart

If your goal is to reach wider audiences, the data suggests that late afternoon through the evening is the best time to attract readers on social media and get them to click through to your site. That being said, this can vary based on your own content focus and promotion strategies – so make sure you experiment and find the times that see the most traffic from your unique audience.

4. Stay on top of your progress.

Keep everyone in the newsroom on top of social promotion strategies and success with a specific social view in your Big Board. Simply select the “Settings” button in the upper right navigation and choose “Social Traffic” from the “Traffic Sources” dropdown, as we show below.

Also, if you’ve suddenly found yourself working remotely and need to get your Big Board set up, we’ve got you covered with this quick setup guide.

Takeaways: Setting yourself up for social media success through experimentation

Social media experiments will only go as far as your organization’s culture of experimentation. Give yourself an opportunity to try new tactics or iterate on your successful strategies. Just remember to:

  1. Analyze your new and returning audiences to uncover unique behaviors among those readers.
  2. Take that data you discovered to experiment, say, with alternative images and your most interesting fact or quote from the content.
  3. Experiment with timing, based on what we know about social interactions (and any reader tendencies to partake in late-night scrolling).
  4. Report and stay on top of your progress, thoughtfully iterating upon (and of course celebrating) the successes, while also learning from any missed opportunities along the way.

And while we used Chartbeat to seamlessly find the information you need and take action, we think that instilling a culture of experimentation comes at no cost, yet it can offer an impactful complement to the data analytics and audience resources you may already have available.


Want to conduct your own social experiments using Chartbeat? Receive a free demo of our analytics and optimizations tools.


More in Research