Social media metrics: which ones to track and how to calculate them

Metrics and KPIs are not the same

A metric is any number you can measure, such as likes, reach or followers. A KPI, a key performance indicator, is a metric you have chosen because it shows progress on a goal.

You can watch many metrics. You should have only a few KPIs. If everything is a priority, nothing is.

Engagement metrics

These show how people respond to your posts. They tell you whether the content is worth people’s attention.

• Engagement rate: total engagements divided by reach, multiplied by 100. Some tools divide by followers or impressions instead, so check the definition before you compare numbers.
• Save rate: saves divided by reach, multiplied by 100. A save usually means a post is worth returning to, which makes it a strong signal on networks that offer it.
• Share rate, sometimes called amplification: shares divided by followers, multiplied by 100. Shares put your post in front of new people.
• Comments and replies: the clearest sign of real conversation. Read them, not just count them.

Reach and awareness metrics

These tell you how many people see your content. They matter most when your goal is to be known.

• Reach: the number of different people who saw a post.
• Impressions: the number of times a post was shown, which includes repeat views by the same person.
• Video views: how many times a video was watched, with each network defining a view in its own way.
• Average watch time: total watch time divided by views.
• Completion rate: completed views divided by views, multiplied by 100.

Audience growth metrics

Follower growth rate is new followers in a period divided by your followers at the start, multiplied by 100. Compare it month to month. A small account growing steadily is healthier than a large one that has stalled.

Pair it with quality: are the new followers the audience you wanted? Ten buyers beat a thousand strangers.

Conversion and return

If social media is meant to drive action, measure the action. Track clicks to your site, sign-ups, trials and sales that start from social.

For paid posts, cost per click is spend divided by clicks. Cost per thousand impressions, known as CPM, is spend divided by impressions, multiplied by 1,000.

Remember where people can click. Some places, such as Instagram feed captions, do not carry clickable links, so plan how people reach you.

Match the metric to the goal

Use this as a starting point.

• Goal: be known. Watch reach, impressions and follower growth.
• Goal: build a community. Watch comments, replies and share rate.
• Goal: teach or inspire. Watch saves and average watch time.
• Goal: sell. Watch clicks, sign-ups and cost per result.

How to track them without a spreadsheet headache

Use each network’s own analytics for detail, and a single tool for the overview. Content Postiz shows reach, engagement and follower trends across your connected channels, so you can read them in one place.

Set a weekly five-minute check and a monthly review. In the monthly review, list your best and worst posts by network, ask what they had in common, and decide one thing to do more of and one thing to stop.

Common mistakes

• Chasing vanity metrics. Follower counts and likes are easy to see and often unrelated to your goals.
• Comparing across networks. A good number on one network can be a poor one on another.
• Looking at single posts. Look for patterns across several.
• Changing too fast. Give a change a few weeks before you decide it failed.

What to watch on each network

The same word can mean different things on different networks, so learn what each one gives you. Use this as a map, and check each network’s own definitions.

• LinkedIn: impressions, reactions, comments and click-throughs. Comments are the signal that a post started a conversation.
• Instagram: reach, saves, shares and profile visits. Saves and shares tend to say more than likes.
• X: impressions, replies, reposts and link clicks. Replies show real interaction.
• TikTok: views, average watch time, completion rate and shares. Watch time says whether the video held attention.
• YouTube: click-through rate on the thumbnail, average view duration and returning viewers.
• Facebook: reach, reactions, comments and shares on your Page posts.

Build a simple scorecard

A scorecard puts the few numbers you care about on one page. It should take minutes to update.

• One row per network.
• One goal metric per row, such as clicks, saves or replies.
• Two support metrics, such as reach and follower change.
• A column for last month and a column for this month.
• A line for what you will do about it.

Use your own history as the benchmark

It is tempting to look for industry averages, but they depend on how each study measured things and on the kinds of accounts it included. Your own last three months are a better yardstick, because they come from your audience and your content.

Ask whether this month is better than last month, and why. That question leads to decisions. “Are we above average?” rarely does.

Track clicks with UTM links

If you want to know which posts bring visitors to your site, tag your links. UTM parameters are small additions to a URL that tell your web analytics where a visit came from, such as the network, the campaign and the post.

Use a consistent naming pattern, such as the network as the source and the campaign name as the campaign. Write the pattern down so everyone on the team uses it, or your reports turn into a mess.

Vanity metrics versus useful ones

Vanity metrics look good and tell you little. Useful metrics change what you do next.

• Follower count alone is vanity. Follower growth among your target audience is useful.
• Total likes are vanity. Saves, shares and replies show deeper interest.
• Total impressions are vanity. Reach plus a click-through to your site is useful.
• Number of posts is vanity. Results per post show what is worth repeating.

When a number drops

A fall in a metric is a prompt to look, not to panic. Work through these questions in order.

• Did you change how often or when you post?
• Did you change the format or topic?
• Did the network change something, such as a feature or ranking?
• Was there a holiday, an event or a news story that drew attention elsewhere?
• Is it one bad post or a pattern across several?

Leading and lagging metrics

Some metrics predict results and others confirm them. Saves, replies and watch time are leading: they move first and hint at what is coming. Sign-ups and sales are lagging: they confirm that the work paid off, but often weeks later.

Track both, and when a lagging number looks bad, check the leading ones to find out where the chain broke.

Compare posts fairly

A post sent to a bigger audience will have bigger numbers. To compare fairly, use rates instead of totals, such as engagement rate or click-through rate.

Also compare like with like. A video should be compared with other videos, a carousel with other carousels, and a post published on a Tuesday morning with other weekday-morning posts.

How many posts before you draw a conclusion

One post tells you very little. Social media results are noisy, and a single post can win or lose for reasons that have nothing to do with the idea.

As a rule of thumb, wait for several posts of the same kind before you decide something works or fails, and keep testing it for a few weeks before you change your plan.

Read the comments, not only the counts

Numbers tell you what happened. Comments often tell you why. Read them each week. Note the questions people ask, the words they use and what they push back on. These are free research and a good source of the next post.

What to automate

Automate the collecting and keep the thinking for yourself. A tool can gather the numbers across networks each week. You decide what they mean. In Content Postiz you can ask your AI to summarise last week’s performance in plain words and propose two experiments, and then schedule the posts that test them.

A worked example: reading one month

Say you posted twelve times on LinkedIn this month. Your goal was more visits to the pricing page, and your tracked links show visits from social.

You list the twelve posts with their reach, comments and link clicks. Three of them brought most of the clicks, and all three were short how-tos. Two posts with the most likes brought almost no clicks, and both were opinion posts.

The lesson is clear and specific: for the goal of visits, short how-tos work better than opinions, even though opinions collect more likes. Next month you write more how-tos, keep a couple of opinion posts for conversation and check whether the clicks follow. That is what metrics are for: not a report, but a decision.

Frequently asked questions