Social media metrics that actually matter (and those that just flatter)
Dashboards overflow with numbers, and that is precisely the problem: when everything is measured, nothing is a priority. Here is a simple reading grid for what to look at, how often, and above all what to do with it.
Vanity metrics vs. action metrics
A vanity metric goes up or down without telling you what to change: total cumulative likes, for instance. An action metric answers a precise question: “does this format work better than the other?”, “is my audience growing faster this month?”. Before looking at a number, ask which decision it could change. None? Move on.
Reach: a thermometer, not a goal
Reach measures how many people saw your content. It depends heavily on algorithms, varies wildly from post to post, and compares poorly across formats. Use it as a trend thermometer over several weeks, never as a verdict on a single post.
Engagement: prefer the rate over the volume
100 reactions from 1,000 people reached beat 300 from 20,000. Engagement rate (interactions divided by reach) is the most honest metric for comparing your own content: it neutralises the algorithm’s distribution whims.
Look at the nature of engagement too: a thoughtful comment or a share weighs more than a reflex like. Those are the signals that tell you what your audience wants more of.
Growth is read over time
Today’s follower count says nothing; its curve over eight weeks says everything. That is why Perroly records a daily snapshot of your accounts, and of the ones you watch: trends emerge, one-off accidents fall into perspective.
Compare yourself, intelligently
Competitor watching is not for self-flagellation: it provides context. If your whole sector sees reach drop the same month, the problem is not your content. Pick three to five accounts comparable to yours (not the stars of the field) and observe their pace, their formats, what sparks reactions.
The 15-minute weekly ritual
One appointment a week is enough:
- Open the dashboard of each connected account.
- Note the trend for reach and engagement rate: up, flat, or down.
- Identify the best and worst post of the week, and form a hypothesis about each.
- Decide on a single change for next week. Just one.