Instagram Engagement Rate Calculation: Three Formulas and When Each One Lies

The formula, and the single decision hiding inside it

Every version of the sum has the same shape. Add up the interactions on a post, divide by an audience number, then multiply by 100. Interactions normally means likes plus comments. Plenty of people add saves and shares too, and if you do, say so when you quote the figure, because the reader will assume you did not.

The only real decision is what goes on the bottom. Followers is the stable choice: it barely moves from one day to the next, so a run of rates measured against it can be compared across months. Reach is the honest choice: it counts the accounts that actually saw the post, so it answers what share of the people who saw it did something. Impressions is the strict choice, because a single account can produce several impressions, so the same interactions end up divided by a larger number. None of the three is wrong; they answer different questions. Trouble only starts when two people compare rates without saying which denominator they used.

Three formulas, and the question each one answers

Here are the sums written out, with the job each is suited to. Pick one, write it down somewhere you will see it again, and use that same one every time you check.

• Engagement rate by followers. Likes plus comments, divided by follower count, times 100. Answers how loud an account is relative to its size. This is the figure a brand or an agency almost always means when they ask.
• Engagement rate by reach. Likes plus comments, divided by accounts reached, times 100. Answers what proportion of the people who saw the post reacted to it. This is the one to judge an individual post by.
• Engagement rate by impressions. Likes plus comments, divided by impressions, times 100. Answers how much interaction each view produced. Always the smallest of the three, which is exactly why it is rarely the one anybody quotes.
• Account engagement rate. Add the interactions across your last twelve feed posts, divide by twelve for an average, then divide that average by your follower count. Answers how the account performs in general, so one unusually good post cannot flatter the whole picture.

Where these numbers actually live inside Instagram

Likes and comments sit on the post itself, so anyone can count those. Reach and impressions do not. They live in Insights, which only appears on a professional account. Switching a personal account to creator or business costs nothing and can be switched back, and it is the only way to get the bottom half of two of the three formulas.

On a professional account, open a post and tap View Insights. Accounts reached is your reach number, and impressions sit in the same panel, sometimes grouped under a views heading. Saves and shares appear as small icons with counts beside them, and those two are the interactions people most often forget.

Your follower count is on your own profile. Take it on the same day as the interactions, because a rate built from this week's likes against a follower count noted three months ago is not measuring anything real.

The same post run through every formula

Take a post with 240 likes, 18 comments, 12 saves and 6 shares, on an account with 8,400 followers. Insights says the post reached 3,100 accounts and produced 4,600 impressions.

Counting only likes and comments, interactions come to 258. Against followers that is 258 divided by 8,400, which is about 3.1%. Against reach it is 258 divided by 3,100, about 8.3%. Against impressions it is 258 divided by 4,600, about 5.6%.

Now include the saves and shares, so interactions become 276. The three rates shift to roughly 3.3%, 8.9% and 6.0%. One post, one day, six defensible figures, all decided by two choices nobody can see when you quote the result. The sum matters far less than saying which sum you used.

Reels and stories break the standard sum

A reel can be shown to accounts that do not follow you, in numbers that make the follower denominator nonsense. A reel reaching 60,000 accounts on a profile with 8,000 followers gives a follower-based rate well above 100%: arithmetically correct, practically useless. For reels, divide by plays or by reach, and never fold reel rates into a feed average.

Stories have no likes and usually no public comments, so the standard interactions simply do not exist. The number used instead is the story view rate: viewers divided by followers, times 100. Replies and sticker taps are the interaction equivalent, and both sit in Insights under the individual story. Keeping feed, reels and stories apart is the biggest single improvement most people can make to their own reporting, because a blended average across all three tells you nothing you can act on.

What an engagement rate cannot tell you

Because it is a fraction, the rate moves for two different reasons and never says which. Interactions falling looks identical to followers rising. An account that gains 2,000 followers in a month while its likes stay flat shows a falling rate with nothing wrong at all.

It also says nothing about who engaged. A hundred likes from accounts that will never buy anything is worth less than ten comments from people in your market, and the ratio treats both the same. Brands understand this, which is why a follower rate is a screening filter rather than a decision.

So track it, but record interactions and followers as their own columns each month, not just the rate they produce. When the rate drops you can then see which half caused it, which is the only version of this number that leads to a decision.

Where buying likes fits into the arithmetic, and where it does not

Bought likes change the top of the fraction and nothing else. Add 500 likes to a post and the engagement rate by followers rises straight away, because the follower count has not moved. That is arithmetic rather than growth, and it is worth being clear with yourself about which one you are paying for.

There is a sensible use and a self-defeating one. The sensible use is social proof on a post you are actively sending people to, because a post sitting on a handful of likes reads as ignored. The self-defeating use is inflating a rate to show a brand, which tends to collapse the moment they ask for an Insights screenshot, because the reach behind those likes will not support them.

If you do buy, keep the size in proportion to the post and to the account. Likes here start at 100, are priced in pounds and paid through PayPal, and the account has to be public while they land. A post with 30 organic likes and 5,000 bought ones does not read as successful. It reads as bought.

Whichever formula you settle on, a single post's rate is noise. A trend is the useful thing, and it only exists if the same method is applied the same way each time.

• Pick one formula and one definition of interactions, and write both down next to your numbers.
• Measure on a fixed day each month, rather than whenever something has just done well.
• Use your last twelve feed posts, never your best three.
• Record interactions, followers and reach in separate columns so you can see which half moved.
• Keep reels and stories on their own rows and out of the feed average entirely.

Questions people ask before ordering