How to calculate engagement rate, and why benchmarks disagree
Quid puts Instagram's median engagement rate at 0.30%. Buffer puts it near 5.5%. Both are correct — they divide by different things. Here are the formulas, and how to pick one.
7 minute read
Engagement rate is not one number
Ask three tools for your engagement rate this month and you will get three answers, none of them wrong. Engagement rate is a family of formulas that agree on the numerator and argue about the denominator. Divide engagements by followers and you have measured how much of your audience responded. Divide by reach and you have measured how well a post did with the people it actually got in front of. Divide by views or impressions and you are closer to an ad-efficiency number. Those are three different questions, so they return three different values, and on a single post the gap can be an order of magnitude. Any benchmark you find is only usable if you know which one it used, and most pages quoting benchmarks never say.
The four formulas, written out
Write them down and the problem stops being abstract. Engagement rate by followers is total engagements on a post, divided by your follower count, times 100. Engagement rate by reach is total engagements divided by reach — the number of unique accounts that saw it — times 100. Engagement rate by impressions or views swaps the denominator for every delivery or play, repeats from the same person included. Engagement rate by post means computing one of those for each post and then taking the average or median across them, rather than summing every engagement and every denominator once. Hootsuite lists six variants, including a daily rate that divides by followers and days rather than by posts. For most accounts reach is smaller than the follower count and views are larger than reach, so the same post produces a low by-follower figure and a flattering by-reach one.
What counts as an engagement is also unsettled
The numerator is less standardised than it looks. Quid, which publishes the benchmark report that used to carry the Rival IQ name, counts likes, comments, favourites, reposts, shares, reactions and video interactions. Socialinsider's Instagram figure is likes plus comments only. Buffer counts likes, comments, shares and reposts but explicitly excludes saves. Saves matter a great deal here: on Instagram they are among the strongest ranking signals and one of the most common actions on a carousel, so including or dropping them moves the answer materially. Before comparing your rate to anything, check what the other side is adding up. Two accounts behaving identically can differ by half simply because their tools disagree about whether a save is engagement.
Why published benchmarks disagree by more than 10x
Take Instagram and three named studies from the same stretch of time. Quid's 2026 Social Media Industry Benchmark Report, published in March 2026 from a random sample of 150 companies in each of 18 industries, puts the all-industry median at 0.30% by follower, down from 0.36% the year before. Socialinsider, working from 35 million Instagram posts across 447,613 profiles between January and December 2025, puts it at 0.48% — also by follower, but counting only likes and comments. Buffer, from more than 52 million posts through 3 December 2025, reports a median around 5.5%, because Buffer divides by reach and notes openly that the data covers Buffer users only. Same platform, same period, roughly an eighteen-fold spread. Nobody made a mistake.
If you want a benchmark, take it all from one study
The only safe way to use published benchmarks is to draw every figure from a single study running a single formula on a single sample. Quid's report is the usual choice for that, because it covers several platforms on one definition — total interactions divided by follower count — across the same pool of brand accounts. On that basis TikTok led every platform at 2.01%, Instagram sat at 0.30%, YouTube made its debut at 0.21%, and X came in at 0.03%, doubled from a 0.015% floor in 2024. Facebook is discussed but never given a single all-industry median, which tells you something in itself. These are brand accounts sampled by industry rather than creators, so read the ranking more than the absolute level.
The denominator moved under everyone in 2025
There is a second reason your own historic numbers refuse to line up: Meta changed the metric. On 21 April 2025, with API version 22.0, Instagram deprecated impressions and plays in favour of views. Reach still counts unique accounts; views count every play and replay, repeats from the same person included, so views run higher than impressions did. If your engagement rate was impression-based it either broke or quietly became views-based, and in the second case your rate fell without anything about your content changing. Stories are the awkward exception and still report impressions. Any year-on-year engagement comparison that crosses April 2025 on Instagram needs a footnote, or you will conclude your work got worse when your denominator simply got bigger.
Which formula to actually use
For deciding what to make next, use engagement rate by reach. It answers the question that matters at the post level — of the people who saw this, how many did something — and it does not punish you for a follower count you no longer reach. Use by-follower only when you are comparing against a published benchmark that uses it, and quote that benchmark by name alongside your number. Use by-impressions for paid, because impressions are what you bought. Calculate per post and report the median rather than the mean, since one outlier will otherwise redraw your whole month. Then write your formula down somewhere your future self will find it, because a number you cannot reproduce in six months is worth nothing.
Treat it as a diagnostic, not a target
The last thing worth saying is that engagement rate is a symptom reading, not a goal. It falls when a post reaches far beyond your regular audience, which is usually exactly what you wanted. It rises on a small, loyal, stagnant account. Read it next to reach and watch time rather than alone, and track its direction over a quarter instead of agonising over a single week. The bigger lever is nearly always volume: more posts means more denominators, more variance to learn from, and a median you can trust. Getting there without losing your week is a workflow problem, which is what FrameOS handles — one recording becomes a batch of captioned vertical clips in a single pass. 300 credits for 3 days · no card.
FAQ
What is a good engagement rate?
It depends entirely on the formula. On a by-follower basis, Quid's 2026 benchmark puts the all-industry median at 2.01% for TikTok, 0.30% for Instagram, 0.21% for YouTube and 0.03% for X. On a by-reach basis, Buffer's data from over 52 million posts puts typical medians between roughly 2.5% and 6.2%. Compare like with like or the answer is meaningless.
Should I calculate engagement rate by followers or by reach?
By reach for anything you are using to make content decisions, because it measures how the post landed with the people who actually saw it. Use by-follower only when you are lining your account up against a published benchmark that uses that denominator, and name the benchmark when you report the number.
Why does every tool give me a different engagement rate?
Because they use different denominators and different numerators. One divides by followers, another by reach, another by views. One counts saves as engagement, another does not. Neither is broken. Pick one tool's definition, use it consistently, and stop comparing its output to numbers computed a different way.
How do I calculate engagement rate for Instagram Reels?
Add likes, comments, shares and saves, divide by the Reel's reach, and multiply by 100. Reach is the count of unique accounts, so this tells you what share of the people served the Reel responded. Avoid using views as the denominator for Reels, since views include replays and inflate quickly on anything that loops.
Sources
- 2026 Social Media Industry Benchmark Report · Quid (formerly Rival IQ)
- 2026 Instagram Benchmarks · Socialinsider
- 2026 TikTok Benchmarks · Socialinsider
- The State of Social Media Engagement in 2026: 52M+ Posts Analyzed · Buffer
- Instagram replaces "Impressions" with "Views": What you need to know · Metricool
Related pages
Keep reading
More Creator Growth →The best time to post on Instagram, according to four studies
Buffer says Thursday 9am. Later says 5am. Sprout says early afternoon. They analysed billions of posts and still disagree — here is why, and what to do about it.
8 minute readThe best time to post on TikTok, according to five studies
Buffer analysed 7.1 million posts and rates Saturday the best day. Sprout Social analysed 2 billion engagements and says avoid the weekend entirely. Both are right, for different accounts.
8 minute readThe best time to post on LinkedIn, according to five studies
Buffer's 4.8 million posts point to Wednesday at 4pm and rank Tuesday among the worst days. Sprout Social's 2 billion engagements call Tuesday the best day of the week. Here is why.
8 minute readTurn one long video into a week of posts
300 credits for 3 days · no card.
Start free