Short answer
X's analytics report engagement rate as engagements divided by impressions: the share of people who saw a post and then interacted with it. If you don't have impressions, engagements divided by followers is a rougher stand-in. There is no single 'good' rate; compare a post with your own recent posts.
Two ways to divide
By impressions. Engagements ÷ impressions × 100. This is the figure X's own analytics show, and the better one for judging a single post, because it accounts for how many people actually saw it.
By followers. Engagements ÷ followers × 100. Useful when you're looking at someone else's account (impressions are only visible to the author) or comparing accounts. It drifts with reach: a post that travels far beyond your followers can score over 100%.
When you enter totals for several posts, put the number of posts in the Posts box. The calculator then also gives engagements per post as a share of followers, which is the version you'll see in most creator media kits.
Worked example
A post shows 10,000 impressions, 150 likes, 20 reposts, 25 replies, 5 quotes, 10 bookmarks and 40 clicks. That is 250 engagements.
- By impressions: 250 ÷ 10,000 = 2.5%
- By followers, with 2,000 followers: 250 ÷ 2,000 = 12.5%
- Impressions as a share of followers: 10,000 ÷ 2,000 = 500%, so the post travelled well beyond the account's followers.
Using the rate without fooling yourself
- Compare like with like. Posts with images, threads and replies all behave differently. Line up similar posts.
- Look at the median, not the average. One post that takes off will drag an average up for weeks.
- Watch which parts move. A high rate built on profile clicks says something different from one built on replies.
- Don't chase the number with engagement bait. Asking for likes or reposts in exchange for something can fall under X's rules on fake engagement, and it inflates a rate without meaning anything.
Related: follower ratio calculator and why follow-for-follow tactics backfire.
Questions
Why is my rate by followers higher than by impressions?
Because a post usually reaches more people than your follower count, so you divide by a bigger number. The opposite happens when a post reaches only a slice of your followers. Neither is wrong; they answer different questions.
What counts as an engagement?
X's analytics count the times people interacted with a post: likes, reposts and replies, plus clicks anywhere on it, such as links, the profile, media, hashtags or expanding the post. Click types are easy to miss, so if you have the total from analytics, paste it into the total box.
What is a good engagement rate on X?
We don't give one, on purpose. Published 'benchmarks' vary wildly with the sample and method, and rates fall as accounts grow. The useful comparison is with yourself: work out the rate for your last 10 to 20 posts and see which ones beat your median.
Does the calculator read my analytics?
No. It does arithmetic on numbers you type. Nothing is fetched from X and nothing is stored.