How to calculate Instagram engagement rate from a brand's last 24 posts.

For each post, add likes and comments and divide by the follower count. Then take the median across recent posts. The median stops one viral post or one giveaway from setting the number. Followers are the usual base on Instagram because, in this data, posts come back with likes and comments but no view counts.

Leave out posts with hidden likes. Two Glossier posts came back with exactly 3 likes and 35 and 14 comments. That pattern fits a hidden like count, not a post nobody liked. Counting them as 3 drags the median down. Whatever you choose, write the formula next to the rate.

There is no single good number. Published benchmarks use different formulas and account sizes, so a rate only means something against similar accounts counted the same way.

This run shows why size matters. Glossier, with 3.1M followers, had a median of 0.12%. Its partner on the giveaway, an account with 156,880 followers, had 1.68%, 14 times higher. Glossier drew more likes per post in absolute terms, 3,736 against 2,345. Compare a creator with accounts of its own size and niche, never with a brand ten times larger.

One post asked people to follow two accounts, like the post and comment to enter a draw for a Chanel bag. It got 5,978 comments, 75% of the 8,010 comments on all 24 posts. The next highest had 327.

The 15 comments the run read show what that looks like. 13 came from one account, the same sentence posted 13 times with a different emoji, all within 26 seconds. A giveaway buys entries, not interest. Leave giveaway posts out of the rate you quote to a client. Here that moved the median from 0.12% to 0.11%. The fewer posts in the window, the more one giveaway weighs.

• Median likes and comments over followers, hidden likes left out.
• Which posts carry most of the comments, and why.
• A sample of comments from the top post: repeats, tags, one-word replies.
• The same check on two or three accounts of the same size.

Not directly. No public data says which followers are real. What you can see is whether followers turn into likes and comments. An account with a large following and a rate far below accounts of its size is worth questioning. Inactive followers look the same as bought ones, so treat a low rate as a reason to ask, not as proof.

Comments are the second signal. Runs of identical or one-word comments on one post point to giveaways, comment pods or bots. This run read one page of 15 comments out of 5,978, enough to spot the pattern on one post but not to score the whole account.

A free calculator returns one rate and does not show which posts it used. This run showed every post with its likes, comments and date, and you choose what to leave out. Profile and posts for one account cost $0.0044. The comments page cost $0.0047. The whole run, two accounts and one comments page, cost $0.0135.

The tools read public profiles, posts and comments. They do not see reach, impressions, saves, shares or story views, which only the account owner sees. Hidden likes show as 3 in this data. The run did not read reel views or score followers. Vet brand accounts, or creators who agreed to it, and not private individuals.