Trusted by 2,000+ brands and agencies across 25+ countries
Engagement rate has no single definition. Instagram in 2026 reads 0.30% as a median from Quid, published 5 March 2026 and divided by followers, and 3.54% from Metricool, published 18 June 2026 with no denominator stated at all. A rate quoted without its denominator cannot be checked. Enter your numbers once and this tool computes the rate on each published formula it can, names the publisher behind every one and shows how your result sits against that publisher's own figure. No signup, no card.
| Formula | What it divides | Your rate | Against the published figure |
|---|
Followers, reach, impressions and views are four different questions about the same post. Dividing by followers asks how much of your audience responded. Dividing by reach asks how many of the people who actually saw it responded. The tool fills in every version it has the inputs for, so you can see the spread rather than picking one number and hoping it travels.
Publishers count different interactions. Socialinsider's Instagram rate counts likes and comments. Its TikTok rate adds shares and saves. Quid counts likes, comments, favorites, reposts, shares, reactions and video interactions together. The tool runs your numbers through each of those formulas separately instead of inventing an average of them.
A rate only means something next to a benchmark built the same way. Each result is shown against the published figure from the publisher whose formula produced it, with the multiple between them. Where a publisher has no figure for that combination the tool says so and leaves the comparison empty.
The calculation runs in your browser. Nothing you type is sent anywhere, so you can paste in client numbers without thinking about it. It is one of the free tools on Flinque, and the full platform is there when a single rate stops being enough.
Engagement rate looks like a settled metric. It is not. There is no agreed formula, no agreed set of interactions to count and no agreed thing to divide them by. Two analysts can take the same post and the same week of data and publish numbers that differ by a factor of ten, with neither of them doing anything wrong. The difference is almost always the denominator.
Here is the clearest demonstration available. Three large studies measured Instagram engagement in 2026. Quid put the all-industry median at 0.30%. Socialinsider put the 2025 average at 0.48%. Metricool, working from more than 24 million posts across more than 375,000 accounts, reported 3.54%. Those three numbers are not in conflict about Instagram. They are answers to different questions.
Quid divides total interactions by follower count and reports the median across its sample. Socialinsider divides likes and comments by followers and reports the average. A median and an average of the same data differ whenever the distribution is skewed, and engagement is heavily skewed, because a small number of posts take a large share of the interactions. That explains the gap between 0.30% and 0.48%.
Metricool's 3.54% is a different case. Its sample is the largest of the three and its figure is roughly seven times higher, but the publisher does not state which interactions it counts or what it divides them by. Without the denominator the number cannot be placed beside the other two. It is reported here because it exists and is widely quoted, and it is labelled as not comparable because that is the honest reading.
A rate with no denominator attached is unusable. If a creator sends you a 6% engagement rate and a media kit with no formula, you cannot tell whether that is remarkable or ordinary. Divided by followers, 6% on Instagram is twenty times the published median. Divided by reach it is unexceptional, because reach is usually a fraction of the follower count and a smaller denominator produces a bigger percentage.
This is why the tool above computes every version at once. Seeing a modest rate against followers and a much larger one against reach for the same post is the point, not a problem to be resolved. The worked examples further down this page show that spread on real arithmetic. Both figures are true. They answer different questions, and knowing which one you are being shown is the whole skill.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026 · Metricool 2026 Instagram study, published 18 June 2026
Before comparing your rate to anybody's benchmark, check that the formula matches. These are the formulas behind the figures on this page, taken from each publisher's own methodology note rather than from a summary of it.
| Formula | Interactions counted | Divided by | Published by |
|---|---|---|---|
| Instagram, Socialinsider formula | likes + comments | followers | Socialinsider, 16 January 2026 |
| TikTok, Socialinsider formula | likes + comments + shares + saves | followers | Socialinsider, 16 January 2026 |
| Facebook, Socialinsider formula | likes + comments + shares | followers | Socialinsider, 16 January 2026 |
| X, Socialinsider formula | likes + shares | followers | Socialinsider, 16 January 2026 |
| Any platform, Quid formula | likes + comments + shares + saves | followers | Quid, 5 March 2026 |
| LinkedIn, Socialinsider formula | likes + comments + shares | impressions | Socialinsider LinkedIn, 16 March 2026 |
| Divided by reach, no publisher | likes + comments + shares + saves | reach | no publisher uses this denominator |
| Divided by impressions, no publisher | likes + comments + shares + saves | impressions | no publisher uses this denominator |
| Divided by views, no publisher | likes + comments + shares + saves | views | no publisher uses this denominator |
Socialinsider counts likes and comments on Instagram but adds shares and saves on TikTok. That is not an oversight. Saves are a meaningful signal on TikTok and were not exposed consistently on Instagram, so the formulas diverge for a reason. It does mean a TikTok rate and an Instagram rate from the same publisher are not strictly comparable either, because the numerator differs as well as the platform.
Quid takes the opposite approach and counts everything it can see, including favorites, reposts, reactions and video interactions, then divides by followers. A wider numerator produces a higher rate for the same post, which is worth remembering when a Quid median looks generous next to a Socialinsider average.
Socialinsider's LinkedIn figure divides by impressions rather than followers, so it belongs in a different column entirely. LinkedIn's 5.20% average is not evidence that LinkedIn outperforms Instagram by a factor of ten. Impressions are a much smaller denominator than a follower count for most pages, and a smaller denominator makes the percentage larger. Comparing the two directly is the single most common mistake made with these numbers.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026 · Socialinsider LinkedIn Benchmarks, published 16 March 2026
Most engagement tools pick one denominator and never mention it. That single choice moves the result by a factor of ten, so this one shows all of them.
Publishers count different interactions. Blending their formulas produces a number that matches none of their benchmarks, so each is run on its own.
Every figure here carries its publisher and publication date next to it, so you can open the source and check it.
Where no published figure exists for a platform and denominator, the tool reports nothing rather than filling the space with an estimate.
The arithmetic happens in your browser. Client follower counts and reach figures never leave the page.
Engagement rate is one signal. The full platform adds audience data and authenticity scoring across more than 10 million verified creators.
Every figure the calculator compares your rate to is listed here, with the statistic it represents and the date it was published. There are nine of them. That is a small number for a metric this widely used, and the shortage is the reason so many engagement benchmarks circulating online cannot be traced to a source.
| Platform | Divided by | Published figure | Statistic | Source and date |
|---|---|---|---|---|
| followers | 0.30% | median | Quid, 5 March 2026 | |
| followers | 0.48% | average | Socialinsider, 16 January 2026 | |
| TikTok | followers | 2.01% | median | Quid, 5 March 2026 |
| TikTok | followers | 3.70% | average | Socialinsider, 16 January 2026 |
| YouTube | followers | 0.21% | median | Quid, 5 March 2026 |
| X | followers | 0.03% | median | Quid, 5 March 2026 |
| X | followers | 0.12% | average | Socialinsider, 16 January 2026 |
| followers | 0.15% | average | Socialinsider, 16 January 2026 | |
| impressions | 5.20% | average | Socialinsider LinkedIn, 16 March 2026 |
Two of those rows deserve a note. Socialinsider's TikTok figure appears as 3.70% in its own table and 3.73% in its body text on the same page. The table value is used here and the discrepancy is reported rather than quietly resolved. Quid's X figure of 0.03% is its 2025 median, which doubled from 0.015% the year before, and it remains the lowest published engagement rate of any platform in this set by a wide margin.
Quid describes Facebook as flat in 2026 but publishes no number for it, so the only Facebook figure here is Socialinsider's 0.15% average. Where you see one source for a platform rather than two, that is because only one published a usable figure.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026 · Socialinsider LinkedIn Benchmarks, published 16 March 2026
Follower counts are public. Reach and impressions are not, unless the creator shares a screenshot of their analytics. If you are weighing up ten creators you have never worked with, followers is the only denominator you can apply consistently to all ten, and it is the one almost every published benchmark uses. Quid and Socialinsider both divide by followers for Instagram, TikTok, Facebook and X, which means a follower-based rate is the only kind you can check against an outside figure.
It has a real weakness. A follower count includes people who no longer open the app, and it can be bought. A follower-based rate therefore punishes accounts with long histories and flatters accounts with small, recently acquired audiences. That weakness is also what makes it useful for spotting bought followers, because a padded audience inflates the denominator and drags the rate down.
If you have access to the analytics, reach answers a better question. Of the people who saw this, how many did something about it. That isolates the content from the distribution, which is what you want when deciding whether a post worked rather than whether an account is worth hiring.
The catch is that nobody publishes a reach-denominated benchmark for Instagram, TikTok, YouTube, X or Facebook. Not one of the sources behind this page does. You can compute the rate, and the tool above will, but you will be comparing it against your own history rather than an industry figure. That is often fine. It is only a problem if you report it as though it were benchmarked.
On video-first surfaces a view-based rate is the common currency, and HypeAuditor now divides likes and comments by views as its main Instagram formula. The difficulty is that a view is a platform-defined number and platforms redefine it. When a platform changes what counts as a view, every historical view-based rate shifts without anybody touching the content.
There is a further problem specific to YouTube. No credible source in this set publishes a view-denominated YouTube engagement benchmark. Quid's YouTube median of 0.21% divides by subscribers, not views. If you compute a view-based YouTube rate you will get a number several times larger, and there is nothing published to compare it with.
The worst outcome is a deck where Instagram is follower-based, LinkedIn is impression-based and YouTube is view-based, with all three in the same bar chart. The chart will show LinkedIn winning by a distance and the finding will be an artefact of arithmetic. If you must put platforms side by side, use followers for all of them and accept that the absolute numbers will look small.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026 · Socialinsider LinkedIn Benchmarks, published 16 March 2026
Save 50% on annual billing across paid plans.
Every example here can be reproduced in the calculator above and checked on paper. The arithmetic is deliberately simple so that any disagreement is about method rather than rounding.
An account with 25,000 followers posts something that takes 1,200 likes and 48 comments. Reach on the post was 9,000.
On Socialinsider's Instagram formula the numerator is likes plus comments, so 1,200 plus 48 is 1,248. Divided by 25,000 followers that is 0.04992, or 4.99%. Against Socialinsider's published 2025 average of 0.48% that is roughly ten times higher. Against Quid's median of 0.30% it is roughly seventeen times higher.
Divided by reach instead, 1,248 over 9,000 is 13.87%. That is a much larger number and it has no published benchmark to sit against. Both figures describe the same post. The first says the audience is unusually responsive relative to its size. The second says the people who saw it reacted strongly, which you would expect given the first.
Suppose that post also took 15 shares and 60 saves. Quid counts interactions broadly, so the numerator becomes 1,200 plus 48 plus 15 plus 60, which is 1,323. Divided by 25,000 followers that is 5.29%.
The rate moved from 4.99% to 5.29% without the post changing at all. Only the counting rule changed. A 0.3 percentage point difference sounds trivial until you are comparing two creators whose media kits used different rules, at which point it is the whole gap between them.
A company page post takes 140 likes, 22 comments and 9 shares, with 11,000 impressions. Socialinsider's LinkedIn formula divides engagement by impressions, so 140 plus 22 plus 9 is 171, and 171 over 11,000 is 1.55%.
Socialinsider's published overall LinkedIn average is 5.20%, so this post sits well below it, at roughly a third of the published figure. Note what you cannot do with this result. You cannot compare 1.55% to the Instagram example's 4.99%, because one is divided by impressions and the other by followers.
HypeAuditor publishes a worked example of its own. An account with 50,000 followers, 1,500 likes and 100 comments gives 3.2%. That arithmetic is 1,600 divided by 50,000, which is indeed 3.2%, so the example is follower-based. It is worth noting that the same page names views as its primary denominator, so its worked example and its stated formula do not match. Checking a publisher's own example against its own formula is a quick way to find out how carefully a benchmark was assembled.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026
This section exists because the gaps in published engagement data are larger than the data, and most pages on this subject fill them silently. These are the things the calculator above will not tell you, and why.
Tier-based engagement bands are the most requested numbers in influencer marketing and the least well supported. Because demand is high and supply is nil, the gap gets filled by figures that have no traceable origin. They are quoted, then quoted again, until after a few rounds they look like established fact.
The test is simple. Take any engagement benchmark you are about to rely on and try to find the publisher, the date, the sample size and the denominator. If you cannot find all four, the number is not a benchmark. It is a rumour with a decimal point.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026 · HypeAuditor Instagram engagement rate benchmarks, created 30 April 2025, last updated 7 August 2026
"We tested three platforms before landing on Flinque. The creator data is more reliable. The filters actually work. Our team spends half the time finding the right influencers now."
"Flinque replaced two tools we were paying for separately. Creator discovery and performance tracking in one place. And the data is actually accurate."
"Our campaign turnaround time dropped significantly after switching to Flinque. What used to take a full day of research now takes under an hour."
A rate is only as useful as the label attached to it. Four things make a reported engagement rate checkable, and leaving any of them out is what causes the arguments.
Write it as a sentence rather than a number in isolation. For example, 4.99% engagement, likes and comments divided by followers, averaged across the last 12 posts, measured in October 2026. That is longer than 4.99% and it is the only version anybody can verify or argue with properly.
If you are receiving rather than sending, ask for those four things before you accept a figure. A creator or agency that cannot say what their own number divides by is not withholding information. They usually do not know, because the tool that produced it did not say.
When you run a post through the calculator above you will usually get three or four different rates. Do not average them. An average of a follower-based rate and a reach-based rate is a number with no meaning and no benchmark. Pick the denominator that matches the question you are answering, report that one and keep the others for context.
If you are vetting creators, that means followers, because it is the only denominator you can apply to everybody and the only one with published figures to check against. If you are reviewing your own content, use reach or impressions and compare against your own history instead of an industry number.
Source: Quid 2026 Social Media Industry Benchmark Report, published 5 March 2026 · Socialinsider Social Media Benchmarks, published 16 January 2026
The calculator settles what a rate actually is. The full platform adds audience quality, authenticity scoring and discovery across more than 10 million verified creators.
It depends entirely on what the rate is divided by, and on which published figure you compare it to. On Instagram, Quid put the all-industry median at 0.30% in its report published 5 March 2026 and Socialinsider put the 2025 average at 0.48% in its report published 16 January 2026, both dividing by followers. A follower-based Instagram rate above those figures is above the published middle. A reach-based rate cannot be judged this way, because no source in this set publishes a reach-denominated benchmark.
Because engagement rate has no single formula. Dividing by followers, reach, impressions or views answers four different questions, and publishers also count different interactions. The tool shows each version separately and names the publisher whose formula produced it, instead of picking one and presenting it as the rate.
None of them is correct in the abstract. Use followers when comparing creators, because it is the only denominator that is public for every account and the only one with published benchmarks. Use reach or impressions when judging whether a piece of content worked. The mistake is not choosing one, it is failing to say which you chose.
No, not as they are usually published. Socialinsider's LinkedIn average of 5.20%, published 16 March 2026, divides engagement by impressions. Its Instagram average of 0.48% divides by followers. Impressions are a far smaller denominator than a follower count, so the LinkedIn figure is larger for arithmetic reasons rather than because LinkedIn audiences are more responsive.
Quid reports a median of 0.30% dividing total interactions by follower count, published 5 March 2026. Metricool reports 3.54% from more than 24 million posts, published 18 June 2026, but does not state which interactions it counts or what it divides them by. Without a stated denominator the two figures cannot be compared, which is why this page carries the Metricool number with that warning attached rather than alongside the others.
No, and deliberately so. No source read for this page publishes engagement rate by follower tier with a stated sample size and one consistent denominator. HypeAuditor publishes tier bands, describes them as rough, states no sample size and mixes a view denominator with a follower denominator inside the same table. Rather than present invented thresholds, this tool applies no tier adjustment at all.
10M+ verified creators across 4 platforms, 12 filters and a fake-follower score on every profile. No card.