You can measure so many things on social media. Likes, shares, clicks, comments, link clicks… but what do those measurements really mean? Will knowing that our engagement rate increased over last month help us decide which photo to post tomorrow?

Content segmentation can help you put those metrics to practical use, optimizing your social media posts and informing your day-to-day content decisions.

When marketers talk about “segmentation,” it’s usually in reference to audience segments. And while audience segmentation is important, there’s still the question of how to best communicate with your target audience. Content segmentation fills in that piece of the puzzle. By breaking your content into different categories, it allows you to compare the performance of each type of post. That makes it easier to create and develop content that your audience will appreciate, which in turn will improve your overall metrics.

What Data to Use

For reference, the metrics we typically use for comparison are engagement rate, click-through rate, or engagements per post, depending on what we are trying to measure. For example, if you’re sharing a blog post, engagements are nice, but what you’re really looking for is a high CTR. If you just posted a photo, however, engagements are where it’s at.

A word of caution before we begin: When segmenting your content, you’ll probably end up with categories that only have a few posts. That’s why we try to work with the biggest data set possible, or leave out those categories entirely. You don’t want one extremely high- or low-performing post to skew the analysis.

Segmenting Posts by Type

When we say “type” of post, we mean how it appears on a timeline, because appearance can be the difference between someone noticing your content or scrolling right past. Facebook has very structured delineations — there are 4 main types of posts (photo, link, status, video) and each post can only be one of these. And on Facebook, post type tends to have a big impact on how people interact with it.

Post types aren’t always as straightforward on other platforms. Twitter doesn’t segment its types as neatly, and you can have a post that has a photo and a link, where neither is really dominant. Those will have to count in both “photo” and “link” categories. Or, if you have enough content, you could create a separate “photo and link” category, answering whether or not you get a better CTR when the link post also has a photo.

Many measurement platforms will do these post type comparisons for you, but there are other variables you can compare. For example, how long is the post? Did it specifically ask for engagement (“like if you agree” or “tell us what you think”)? Does it include hashtags?

Exploring post types on this level may seem petty and granular, but you’d be surprised — for several clients, basic insights on post structure have vastly improved engagement.

Segmenting Posts by Topic

This segmentation tactic is less structured than the previous one, but it can be extremely useful. Here’s what to do:

Step One: Create Your Tags
When developing your content strategy, develop a set of “tags” that describe the types of social media posts you should be sharing. These tags aren’t audience-facing; they’re just a way to categorize your posts for internal use. For example, if your brand shares a lot of funny content, “humor” will probably be a tag. If you will frequently share curated content, “curated” should be a tag. And so on.

These tags can be specific to your brand, identify a content series, or be broader and more topical. Just try to have fewer than 15 tags, or you could end up with a bunch of categories with only a few posts in each.

It’s also important that everyone on the team understands the tags and what they mean. Otherwise, you could end up disagreeing about which posts belong to which category.

Step Two: Use Your Tags
As your community manager creates posts, he or she should add those tags to the posts. To manage and track tags among various content contributors, we recommend implementing a content planning/workflow tool. We use Divvy to manage social media content approvals, which also allows users to add tags to content. It’s not perfect, but when we are analyzing content, we can export the Divvy calendar and align it to the native platform exports.

Step Three: Analyze Your Tags
Use the tags to break your content into groups and analyze. Be sure to have a large sampling of content — we try not to analyze fewer than three months’ worth at a time.

Once you get three months of content and analyze it, you should see a few distinct patterns. Maybe you find that “humor” content gets better engagement, but “curated” content gets more click throughs. Maybe content about “social media” gets shared more frequently, but content specifically about LinkedIn performs best.

Sometimes you have to dig a little deeper — is it that we always share LinkedIn content with a photo, or does our audience really just like content about LinkedIn? There are many variables in social media, and it’s important to consider all of them when drawing your conclusions.

Closing the Loop

Finally, no matter how you decide to segment content, be sure to share your findings with your content creation team. After all, they’re the ones who will benefit the most from this information. However, don’t just give them a report. Take some time to talk about what your insights mean.

Discussion is important, for two reasons: 1) You could have drawn the wrong conclusion. There are lots of variables, and nobody’s perfect. 2) You don’t want your content team to follow the numbers too closely. It’s easy to let creative guidelines push your content into a rut, creating the same content over and over and never experimenting.

The insights you will glean from these analyses should be treated as guidelines: what we’re shooting for is a mix of what we know works and new things we think will work. They are intended to guide creativity, not stifle it. In this way, content segmentation will drive your team to continually improve, creating better content for your brand and your audience.

What methods does your team use to better understand your audience? We would love to hear your thoughts in the comments below and make this an ongoing conversation.