Back to the hub

Engagement rate calculator

Interactions over reach, which is the only version of the sum that survives being shown to a client.

By Redação Swonkie

Equipa editorial

2026-07-21

Interactions = likes + comments + shares + saves.

Engagement rate

3,00%

Normal

There are two ways to work out an engagement rate, and the difference between them is not a detail.

Over followers. Interactions divided by follower count. It is the more widely quoted and the easier to game, because the denominator is a historical number that no longer describes who sees the posts. On an account with a few years behind it, a good share of those followers are inactive.

Over reach. Interactions divided by the people who actually saw it. That is the one above. It answers the useful question - of those who saw this, how many reacted - and it treats large and small accounts fairly.

What counts as an interaction

Likes, comments, shares and saves. Saves and shares matter more than likes for distribution on most networks, but for the rate they all go in the same bucket. Link clicks do not: that is a different metric with a different job.

Reading the result

  • Below 1%: the content is not speaking to the people seeing it. Before changing everything, check the reach is not coming from paid distribution or from a post that travelled outside the audience.
  • 1% to 3%: where most brand accounts live.
  • 3% to 5%: good.
  • Above 5%: check the reach. High rates usually come from small samples, and a post seen by 40 people proves nothing.

The most common reporting mistake

Comparing one month's rate with another without looking at volume. Publishing less tends to raise the average rate, because the weak posts drop out first. A rate going up while total reach falls is bad news wearing a good face. Which is why the rate never travels alone in a report: it goes next to reach and to the number of posts.

See it in the product

Keep going

Prompts
Analytics

2026-10-06

Prompt: what to measure for this goal, and what to ignore

The list of what you will ignore is more useful than the list of what you will measure. It is the one that stops the report growing.

Pick three metrics before you start, and write down which ones do not matter. That is the part that stops the thirty-page report.

I will tell you this brand's goal for the next three months.

Answer with:
1. The three metrics that answer that goal. No more than three.
2. For each one, what makes it rise misleadingly, and what to put beside it so you do not fall for
   that.
3. The metrics we will deliberately ignore, with the reason in one line.
4. The baseline I need to record today to be able to compare in three months.

If the goal is badly framed for measurement, say so and propose a measurable version before
answering the rest.

Redação Swonkie

Equipa editorial

Read the full piece
Prompts
Analytics

2026-09-03

Prompt: turn a monthly report into something a client reads

If a number went up because of a single post, say so. That is the difference between a report and a sales deck.

Numbers do not convince on their own. The prompt that turns a month of metrics into three paragraphs a person actually reads.

You have last month's data for a profile. Write the executive summary of a report for a client who
does not work in marketing.

Required structure, three paragraphs:
1. The most relevant thing that happened, with the number that proves it.
2. Why it happened, tied to concrete decisions we made.
3. What we will do next because of it.

Rules: no technical term without an explanation in the same sentence. No percentage without the
absolute value beside it. If a number went up because of a single post, say so. If any indicator
got worse, it belongs in the summary, not buried in the tables.

200 words maximum. Do not invent causes: if the data does not explain a change, write that it does
not.

Redação Swonkie

Equipa editorial

Read the full piece