How to Build a Data-Driven Social Media Strategy

Data Driven Social Media

As you’ve probably read on this blog (and all over the web as of late), social media has become almost entirely pay-to-play for brands. If you want to reach even your existing fans, get out your credit card. Got it? Good.

Now more than ever social media is an investment for brands. You can’t just have your intern send out a few tweets a day and expect results. But when armed with the right data, you can develop a plan to make the most of your social media strategy. Here’s how.

1. Find your audience

If a tree falls in a forest and no one is around, does it make a sound? You can equate your social media posts to this old adage – without an audience, you can post all you want, just don’t expect any results.

Start by auditing your existing audience. Facebook and Twitter both have great insights and analytics reports out of the box for brand pages/handles. These reports, of course, are less helpful if you’re just getting started and don’t have a following yet. But, have no fear! Facebook’s Audience Insights will help you build out what your ideal audience looks like. Whether you’re interested in building audience based on personal demographics (age, education, income) or interests and behaviors (shopping habits, lifestyles), Facebook will help you understand where and how those users interact on the network.

Facebook Insights

Once you’ve created your audience, you can easily save the target demographics and run ads and promotions to those exact users specified. Remember, social media is pay-to-play, so even a small budget will help to put your page in front of the right users.

Twitter has similar options within its Analytics reports. If you click on the Audience menu, you’ll see your current Twitter audience. Again, if you haven’t built a following yet this won’t be very insightful. The fun comes in when you add another audience to compare. In the below, you can see how you can build your ideal audience based on Twitter’s available demographics.

Twitter Analytics

2. See what’s worked (and what hasn’t)

Next up is the content. Do you find success sharing owned content, like your company’s webinars or eBooks? Or are your fans more likely to interact with a photo of your company’s latest happy hour? The answers to these questions will vary greatly from brand to brand.

If you look back within Facebook’s Insight and Twitter’s Analytics reports, you can see content broken down by post with engagement and reach numbers to go along with each one. At a glance, you can see what’s worked and what hasn’t.

We recommend exporting your posts and engagement metrics. In the data-visualization tool of your choice, you can now manipulate the graphs to give you quantitative support for your content strategy.

Let’s look at a simple example from Twitter. The below chart represents total engagements for tweets broken down by type of tweet – link share, photo share or just a text-based tweet.

Screen Shot 2016-07-19 at 2.23.56 PM

We can clearly see that links generate by FAR the most engagement. But you can’t take that at face value – what if the brand almost exclusively shares tweets with links? Just looking at the raw data like this would skew the graph, as we just saw. Instead of looking just at total engagements, let’s break down the same tweets by type, but measure the engagement rates.

Screen Shot 2016-07-19 at 2.49.15 PM

Now you can see that while text-based tweets generated relatively low engagement numbers compared to the entire corpus of tweets, when looking at engagement rates we see the opposite. Text-based tweets have the highest engagement rate, with link tweets following closely behind.

From a strategic standpoint, we now have quantitative data to support a decision around what types of content that should be posted to increase engagement for this account.

3. Get the timing down

Don’t let the internet fool you – there is NO universal “best time to tweet.” There is, however, a trend to be found within your own data to figure out what works best for your brand.

Using the same data from your social media posts, analyze performance of your posts based on the time and day you posted them. You can likely use the same data export you just used to audit the content itself.

Let’s revisit the last example again, but orient our data to show engagement rate by the hour when the tweet was published. Below we can see that this account sees the best engagement rates during the work day, with a spike on the earlier end of the day – likely during morning commutes.

Screen Shot 2016-07-19 at 4.05.13 PM

What about on a daily basis? The below visualization shows days with the highest average engagement rate as the darkest color. Here we can see that Wednesday had the highest average engagement rate. What does this mean for your strategy? To get the most out of your owned content, try sharing it early and around lunchtime on Wednesdays.

Screen Shot 2016-07-19 at 4.13.36 PM

These examples used simple averages for metrics over the course of the past year. When analyzing for your own account, you’ll want to account for seasonality and other extraneous factors that may impact the data over time.

Put your data-driven strategy to work

Although you now have the basics down for building your data-driven social media strategy, we want to offer one bit of advice: test, analyze, and re-test. As with any strategic plan, you’ll want to give yourself room to make adjustments on the fly. Find what works for your brand – with the data to prove it!

Tori Sabourin
Senior Marketing Analyst 

Download our new eBook, Reinventing Public Relations

Posted on August 2, 2016 in Analytics, Data-Driven PR, Social Media

Share the Story

About the Author

Back to Top
Subscribe to SHIFT Happens!

Subscribe to SHIFT Happens!

Want fresh PR and earned media news delivered to your inbox? Sign up for the SHIFT HAPPENS newsletter!

You have Successfully Subscribed!