Most dashboards display the least useful metric first: follower count. It's a reassuring number, easy to compare, and it says almost nothing about an account's real performance. Here's what to look at instead.

1. Engagement rate relative to reach

This is the central indicator. It's calculated by dividing total interactions by the number of people who actually saw the post.

The nuance matters: many tools calculate engagement against follower count, which distorts everything. An account with 10,000 followers whose posts only reach 800 people doesn't have a 2% engagement rate but a 25% one. The first reading suggests failure; the second reveals a highly responsive audience.

Look at this metric over time, never in isolation. Its trend is far more informative than its absolute value.

2. Share of reach from non-followers

What percentage of people reached weren't already following you? This metric directly measures your ability to escape your existing circle.

An account whose reach comes 95% from followers is a closed loop: it's talking to people who already know it. Growth runs through this number, not through post count.

On recommendation-heavy networks — TikTok, Instagram, Pinterest — a non-follower share below 30% signals content that isn't leaving its bubble.

3. Save and share rate

Likes are the cheapest interaction, and therefore the least meaningful. Saves and shares commit more: one signals content someone wants to find again, the other content they'll publicly recommend.

Algorithms know this and weight these signals heavily. A post with 40 likes and 25 shares travels much further than one with 200 likes and 2 shares.

It's also the most useful metric for steering production: saved content is content that delivers practical value.

4. Completion rate

On sequential formats — video, carousel, thread — the question isn't how many people started, but how many reached the end.

A low completion rate on a video almost always indicates a hook or pacing problem, rarely a subject problem. On a carousel, a sharp drop between the first and second slide signals an opening promise the rest doesn't keep.

It's the most directly actionable metric: it tells you precisely where your content loses people.

5. Traffic driven to your site

If your objective is commercial, this is the only metric connecting social activity to a measurable outcome. You read it in your web analytics, not in network statistics.

Two precautions. First, use UTM parameters to distinguish sources — otherwise part of your social traffic gets counted as direct. Second, don't judge a network on raw volume: a hundred qualified visitors beat a thousand who leave in three seconds.

6. Growth rate, not total

Follower count has a use, but a derived one: it's the change that informs. Gaining 200 followers in a month after gaining 50 the month before means something is working, and deserves investigation.

Watch unfollows too. Positive net growth can mask heavy churn — a symptom of a gap between what attracts people and what keeps them.

Metrics not to track

Absolute follower count. It says nothing about audience quality and pushes you toward counter-productive decisions.

Impressions alone. An impression is a display, not a read. Without the associated engagement rate, the number is meaningless.

Single-post metrics. Variability between posts is enormous. A four-week average is vastly more reliable than one isolated post, good or bad.

A simple reading method

Once a month, note these six figures and compare them to the previous month. Don't try to optimise everything: identify the metric that dropped most and address only that one.

This discipline avoids the most common trap in social performance analysis — looking at a lot of numbers and drawing no decisions from any of them.