How to Calculate Engagement Rate
One formula, five denominators, and no way to compare the answers across apps.
How to calculate engagement rate is the easy half: add the interactions on a post, divide by the size of the audience that saw it, multiply by a hundred. Interactions means likes, comments, saves and shares, and that part is the same wherever you go.
The hard half is the denominator, because each platform gives you a different one in its own analytics, and they are not the same kind of number. One gives you people. One gives you plays, counting repeats. Two give you impressions. One does not offer the figure at all. So the formula is portable and the result is not: a rate calculated on Instagram and a rate calculated on TikTok are answers to different questions, and comparing them tells you nothing about either.
Here is what each app actually hands you, which is the part that decides your answer.
Instagram Insights reports accounts reached — people, deduplicated. One person who saw your post four times counts once. TikTok reports video views, which are plays, and a replay counts again, so the same viewer can inflate your denominator by themselves. X and LinkedIn both report impressions, which are appearances rather than people, so the same person scrolling past twice is two. YouTube does not publish an engagement rate field at all; it gives you views and average view duration and expects you to build your own measure out of them.
Run the same post through those and the rate falls as the denominator gets looser. A deduplicated people count is the smallest bottom number, so it produces the highest rate. A plays count that includes replays is the largest, so it produces the lowest. Nothing about the post changed. That is why a benchmark like a certain percentage being good, copied from an article about one network, is not a target on another — it is a number from a different measuring system wearing the same name.
The numerator: what counts as an interaction
Likes, comments, saves and shares. The first two are obvious and the last two are the ones people leave out, which quietly lowers every rate they calculate.
Saves and shares are also the interactions worth most as signal, because they cost the viewer something. A like is a tap. A save is a decision that the post is worth returning to; a share is a decision to put your content in front of someone else using their own name as the recommendation. Platforms have said as much about how they weigh these, and it matches intuition.
Decide what goes in your numerator once and keep it fixed. The comparison that matters is your post against your other posts, and that comparison breaks the moment you change the formula halfway through.
Reach, impressions, and views are not synonyms
Reach counts people. Impressions count appearances. Views count plays. They are three different quantities and the gaps between them are large.
For a single post, impressions are always at least reach, usually more, because people encounter things twice. For video, views can exceed both, because a replay is another play but not another person and sometimes not another appearance.
The practical rule is that you cannot mix them in one comparison. If you calculate on reach for one post and on impressions for another, the first will look better by an amount that has nothing to do with how the posts performed. This is the most common error in engagement reporting and it survives because both numbers sit on the same analytics screen, one above the other.
Doing it per platform
On Instagram, open Insights on a professional account and use accounts reached. That gives you a per-person rate, and the same screen splits reach into followers and non-followers, which is more useful than the rate itself.
On TikTok, the analytics tab gives you video views. Your rate will look lower than Instagram's for identical performance, and that is the denominator, not your content.
On X and LinkedIn, use impressions from the post's own analytics. Both are appearance-based, so those two are the only pair here that can be compared to each other with any honesty.
On YouTube, there is no engagement rate field. Build one if you need it — likes and comments over views — and label it as yours, because nobody else's YouTube engagement figure will have been built the same way.
Per post or per follower
There are two common bases and they answer different questions. Dividing by reach or impressions asks how well a post did with the people who saw it. Dividing by follower count asks how much of your audience you are activating.
The follower version is the one brands often quote, because it is the only one they can calculate about you from the outside — your follower count is public and your reach is not. It is also the more fragile of the two: it falls as an account grows, since follower counts accumulate people who stopped watching while a given post still only reaches the active ones.
So a large account with a low follower-based rate and a small account with a high one may be performing identically with the people actually present. Know which base someone is using before you accept a comparison built on it.
What a good rate is, and why nobody can tell you
Nobody can give you a cross-platform number, because a cross-platform number would have to assume a shared denominator and there is not one. Any single figure offered as the benchmark for social media is quietly assuming one platform's measurement.
What you can do is build your own baseline. Calculate the same way, on the same platform, for thirty days of your own posts, and take the middle of that range as your normal. After that the metric starts being useful, because it tells you when something departed from your normal and is worth understanding.
This is less satisfying than a target and considerably more useful. A benchmark tells you how you compare to an average built from accounts that are not yours; a baseline tells you which of your posts did something different.
What the rate cannot tell you
It cannot tell you whether an audience is worth having, because it is a ratio and ratios hide size. A post seen by very few people that gets a handful of interactions produces a spectacular rate and no outcome.
It also cannot tell you what to change. A low rate is a symptom with several possible causes: the wrong people saw it, the right people saw it and were unmoved, or the format gave nobody anything to do. Those need different responses and the rate does not distinguish between them. The split between followers and non-followers usually does.
And if the goal is how a post reads to a stranger rather than how your existing audience behaves, the rate is not the metric at all. The visible interaction count is, and that is a different thing to work on.
Questions people ask about engagement rate
What is engagement rate?
The share of the audience that interacted with a post: likes, comments, saves and shares divided by the audience figure, as a percentage. The interaction half is consistent across platforms; the audience half is not.
What is a good engagement rate?
There is no cross-platform answer, because each app supplies a different denominator. A rate calculated on deduplicated reach is structurally higher than one calculated on plays. Build a baseline from your own posts instead.
Should I divide by followers or by reach?
Reach or impressions answers how the post did with people who saw it. Followers answers how much of your audience you activated, and it is the version outsiders can calculate about you. They are not comparable.
Why is my TikTok rate lower than my Instagram rate?
Because TikTok's denominator is video views, which counts replays, while Instagram's is accounts reached, which counts people once. A looser denominator produces a lower rate for identical performance.
Where is engagement rate in YouTube analytics?
It is not there. YouTube reports views and average view duration and does not publish an engagement rate field, so any YouTube figure you see was built by whoever quoted it.
Do saves and shares count?
Yes, and leaving them out lowers every rate you calculate. They are also the interactions that cost the viewer the most, which makes them the more meaningful part of the numerator.
Keep reading
- add likes across networks — If the number you need to move is the visible interaction count rather than the ratio, you can add likes across networks and keep the rate as your own diagnostic.
- what reach means on Instagram — The denominator is the part worth understanding first, and what reach means on Instagram covers the one that splits into followers and non-followers.
- why the Facebook like count is six reactions — The numerator has its own trap on one network, because why the Facebook like count is six reactions means you are adding up something other than likes.