Engagement Rate Calculator for Every Social Network

On October 28, 2026 we opened every study and help page this page cites and copied the figures from the pages themselves, not from roundups that quote them. The benchmarks come from four 2026 publications: Socialinsider's 2026 social media benchmarks and its separate Instagram, TikTok and LinkedIn studies, the 2026 Social Media Industry Benchmark Report that Rival IQ now publishes under Quid, and Buffer's State of Social Media Engagement 2026. Follower tier figures for Instagram come from HypeAuditor's benchmark guide, last updated in August 2026.

For the formulas we went to each network's own documentation: Meta's developer pages for Instagram media insights and Threads insights, LinkedIn's content analytics help, Pinterest Analytics help, the TikTok video object, the YouTube Data API, the Mastodon status entity, Telegram's statistics API and the Discord message resource. Those pages say which counts exist. They decide which formula is even possible on a network.

Two warnings from that reading shape the whole page. First, the studies do not agree because they do not measure the same thing: one divides by followers, another by reach, one reports the average, another the median. Second, two of the four label 2025 data as 2026 benchmarks while they wait for a full year of 2026 posts. We say which is which in every row, and the calculator only compares your number with a benchmark built the same way.

This page covers the 11 networks Spreadcast publishes to: Instagram, Facebook, Threads, LinkedIn, Pinterest, TikTok, YouTube, Bluesky, Mastodon, Telegram and Discord. We did not survey anyone and we did not run our own benchmark study. Where a number is ours, it is a made up worked example and labeled as one.

Engagement rate is the share of people who did something with a post: liked it, commented, shared it, saved it, or on some networks clicked it. It is written as a percentage. The top of the fraction is the interactions. The bottom is the group you measure against, which is where all the arguments start.

Raw counts mislead because they grow with audience size. A post with 300 likes on an account of 150,000 followers did worse than a post with 60 likes on an account of 1,500. Dividing by audience puts both on the same scale: 0.2% against 4%. That is the whole point of the metric. It lets a small account and a big one, or this month and last month, be judged by the same yardstick.

What counts as an interaction depends on the network, and the network decides that by what it shows you. Instagram reports likes, comments, shares and saves. Facebook reports reactions, comments and shares. LinkedIn adds clicks to its own figure. Pinterest counts saves, Pin clicks and outbound clicks. YouTube gives likes and comments. Bluesky has likes, reposts, replies and quotes. Mastodon has favourites, boosts and replies. Telegram has views, reactions and forwards. Discord has reactions and replies.

Engagement rate is not reach and it is not growth. A post can reach 20,000 people and convince almost none of them to react, and a post seen by 400 can start a long thread. Track the three side by side: followers for growth, reach or views for distribution, engagement rate for how well the content lands with the people who saw it. Our guide to social media KPIs lays out the full set and how they fit together.

One more trap. Engagement rate rewards content that asks for a reaction, and the networks know it. Meta has demoted posts that beg for likes since 2017. A rate earned by a giveaway or a "comment YES" prompt tells you little about whether your normal posts work. The useful number is the rate on your ordinary posts, tracked over months.

Start by writing down the interactions for one post. On Alex's post that is 52 likes, 7 comments, 4 shares and 9 saves. Add them: 72. If a network shows a combined figure, Instagram's API for example has a total interactions metric that already nets out unlikes and deleted comments, use that figure rather than adding the parts yourself.

Next, pick the bottom of the fraction and write down where you got it. For a follower based rate, that is 2,400. Divide: 72 ÷ 2,400 = 0.03. Multiply by 100 to turn it into a percentage: 3.00%. For a reach based rate, divide by 1,600 instead and you get 4.50%. Keep two decimals. On Instagram and Facebook most brand accounts land below 1% by followers, and the difference between 0.31% and 0.38% matters there.

To get the engagement rate across several posts, you have two choices. The first is to calculate each post's rate and take the average. The second is to add up all interactions and divide by the sum of the bottoms. Say Alex's next post earned 20 interactions from a reach of 400. Averaging the two rates gives (4.50% + 5.00%) ÷ 2 = 4.75%. Pooling gives 92 ÷ 2,000 = 4.60%. Pooling weights big posts more. Averaging treats every post as one vote, which is how the studies report a typical post.

Then decide between average and median. One viral post pulls an average up for months. The median, the middle post when you sort them, ignores it. Rival IQ and Buffer both report medians for that reason, Socialinsider reports averages. If your own rate is calculated as an average and you compare it with a median benchmark, you will look better than you are whenever you had one big hit.

Last, fix the time window. Interactions keep arriving long after posting on some networks, search driven ones like Pinterest and YouTube especially, so a post measured after one day looks worse than the same post after a week. Pick one window, such as seven days after posting, and stick to it. Spreadcast's Analytics uses the first week of interactions for its engagement rate for exactly this reason: a post from yesterday and a post from last month are then measured over the same stretch.

If you track this in a sheet, put one row per post with the date, network, format, each interaction type, followers on that day, reach and views. The template in our social media audit template has those columns ready, and the content calendar shows how to link each row back to what you planned.

Use followers when you compare accounts. It is the only bottom you can read from the outside, so it is how every public benchmark of brands works, and how you vet a creator before paying them. Its weakness is that most of your followers never see a given post. The bigger the account, the smaller the share that sees each post, so follower based rates fall as accounts grow, whatever the content quality.

Use reach when you judge your own content. It divides by the people who saw the post, followers and strangers alike, so a post that travelled far outside your audience is not punished for it. Buffer's 2026 report uses reach for exactly that reason and gets rates ten times higher than the follower based studies: about 5.5% on Instagram against 0.30% to 0.45%. Both are right. They answer different questions.

Use impressions or views when that is what the network gives you. LinkedIn's own engagement rate is interactions per impressions. Pinterest's is engagements divided by the times your Pins were seen. TikTok and YouTube show views on every video, and the video studies divide by views. On a network that shows views but not reach, a view based rate is the closest thing to a reach based rate you can get.

Some networks leave you only followers. Bluesky and Mastodon show no view counts on posts, and Discord messages carry reaction counts but no views. On those three, divide by followers or members, or simply track interactions per post, which is what Buffer does for Bluesky and Mastodon.

Whatever you pick, write it next to the number every time. "Engagement rate 4.2%" means nothing on a report. "4.2% by reach, median of 18 posts, first seven days" can be compared with next month and with a benchmark built the same way.

Most people who search for an engagement rate calculator want the Instagram one. The formula is (likes + comments + shares + saves) ÷ followers × 100 for a public comparison, or the same top divided by reach for your own posts. Meta's documentation lists likes, comments, shares and saved for feed posts and Reels, plus reach, which it describes as unique accounts that saw the media, and views. It also says impressions are deprecated for media created after July 2, 2024, so a calculator that still asks you for impressions on a new post is out of date.

So what is the average engagement rate on Instagram in 2026? By followers, Socialinsider's Instagram study puts the average brand post at 0.48% for 2025 and its third quarter 2026 update at 0.45%, counting likes and comments over followers across 67.5 million posts from accounts with 1,000 to 1 million followers. Rival IQ's report puts the median at 0.30%, down from 0.36% a year earlier. By reach, Buffer's median is about 5.5%. Those three numbers describe the same network.

Format matters more than most tips suggest. In Socialinsider's data, carousels averaged 0.55% by followers in 2025 and 0.50% in the third quarter of 2026, Reels 0.52% and 0.48%, single images 0.37% and 0.35%. Buffer, measuring by reach, found carousels at a 6.90% median, single images at 4.44% and Reels at 3.31%. Reels win on reach instead: Buffer found they reach about 36% more people than carousels and about 125% more than single images. A carousel is the engagement play; a Reel is the discovery play.

For creators the numbers run higher because their followers chose them personally. HypeAuditor's guide, updated in August 2026, gives rough ranges of 4% to 5% for accounts with 1K to 5K followers, down to 1% to 2% above 500K, and puts brand accounts at roughly 0.5% to 1%. HypeAuditor divides likes plus comments by views rather than followers, so read those as view based figures.

When you vet an influencer, compare their follower based rate with the tier, look at the comments for real sentences rather than emoji strings, and check whether the rate holds across their last 20 posts or rests on two. A tenfold gap between one post and the rest usually means a giveaway or a paid push. Our guide to promoting an app on Instagram covers the creator side in more depth, and the Instagram post generator shows the sizes and caption limits that carousel posts need.

TikTok is a video network, so the standard TikTok engagement rate divides by views: (likes + comments + shares + saves) ÷ views × 100. TikTok's developer video object carries like, comment, share and view counts for every video, and the app shows saves beside them. A follower based rate also works and is what you will see in brand benchmarks, but on TikTok most views come from people who do not follow you, so dividing by followers can produce rates above 100% on a video that travels.

By views, Socialinsider's TikTok study, published in April 2026 from 40.7 million organic videos, puts the average at 3.85% for the first half of 2026 and 3.90% in the third quarter. By followers, its main benchmarks page lists TikTok at 2.50% for 2026, down from 3.70% for 2025, and Rival IQ's median is 2.01%. Buffer's median by reach is about 4.6%. TikTok still leads every network on follower based engagement in both brand studies.

Account size barely moves TikTok's view based rate, which is unusual. Socialinsider's 2025 figures by page size run from 4.40% for accounts under 5K followers to 3.75% for 50K to 100K, and back up to 3.95% above 100K. A small account is not at a disadvantage there, because the For You feed hands each video to strangers first.

When you use the calculator for TikTok, pick views as the bottom unless you are comparing yourself with a brand benchmark by followers. Count saves: on TikTok a save often means "I want to try this later", the strongest signal you get from a how to video. And give the video three to five days before you judge it, so the slower views and saves are counted too.

If you make videos for an app, our guide to promoting an app on TikTok explains which formats earn saves and shares rather than passive views.

LinkedIn calculates engagement rate for you on Company Page posts, and it uses impressions. Its content analytics help page defines engagement rate as the ratio of interactions per impressions, and counts clicks on your content, company name or logo among them. It also warns that impressions are an estimate, and that metrics can take 48 hours to show, except reactions and comments, which update at once. A LinkedIn engagement rate calculator earns its keep when all you have are the counts, such as a competitor's post or an export from a tool that leaves the rate out.

The formula, then, is (reactions + comments + reposts + clicks) ÷ impressions × 100 if you want to match LinkedIn's figure, or the same without clicks if you want to match the studies. Clicks are often the biggest number on a LinkedIn post, so including them or not changes the result a lot. The calculator has a separate box for clicks so you can do either.

The average engagement rate on LinkedIn in 2026 is higher than on any other network on an impressions or reach basis. Socialinsider's LinkedIn study, published in March 2026 from 2.45 million organic posts by pages with 1K to 1M followers, puts the average at 5.20%, counting reactions, comments and shares over impressions. Buffer's median, which includes clicks, is about 6.2%, the top of its table.

Format is where LinkedIn gets interesting. Socialinsider's 2025 averages by impressions: native documents, the PDF carousels, 7.00%; multi image posts 6.45%; video 6.00%; single images 5.30%; plain text 4.50%; polls 4.20%; posts with a link 3.25%. A link post earns about half the rate of a document post. If you share an article, put the link in the first comment and keep the post itself as a document or an image.

For carousel specs and examples that work, see LinkedIn carousel size and our LinkedIn carousel examples. The LinkedIn post generator covers the caption side, including where the "see more" cut falls.

The usual YouTube engagement rate is (likes + comments) ÷ views × 100, with shares added if you read them from YouTube Studio. The public YouTube Data API gives view, like and comment counts for every video. Dislikes have been private since December 13, 2021, so they cannot be part of a public formula. YouTube's help page on engagement adds that like counts in Analytics can differ from the watch page and can drop as invalid likes are removed.

Views changed meaning recently, which matters for anyone comparing old and new rates. YouTube's help now says views are counted the moment a video starts to play, across Shorts, long videos and live streams, and the Data API notes that from August 24, 2026 the public view count follows the same rule. More plays count as views, so the same likes divided by a bigger view count give a lower rate. A drop in your YouTube engagement rate around that date may be an accounting change, not a content problem.

Benchmarks for YouTube are thin. Rival IQ added YouTube to its 2026 report for the first time and found a median engagement rate of 0.21% by subscribers across industries, with Food far ahead and Tech and Financial Services trailing. Buffer does not publish a YouTube rate at all. It reports the median video it publishes earned 433 views. No study we checked publishes a view based YouTube rate for 2026, so the calculator shows the subscriber benchmark only.

For Shorts, treat YouTube like TikTok: divide by views, wait a few days, and look at shares and comments over likes. For long videos, engagement rate is a weak signal next to watch time and average view duration, which YouTube's own analytics put first. A tutorial watched to the end with few likes is doing its job.

On Facebook the interactions are reactions, comments and shares, and the formula is (reactions + comments + shares) ÷ followers × 100 for a public comparison, or divided by reach for your own Page. Count every reaction type, not only likes: a love or a haha is the same act of engaging.

Facebook has the lowest follower based engagement of the big networks. Socialinsider's benchmarks put the average brand post at 0.15% for 2024 and 2025 and 0.13% in the third quarter of 2026. Rival IQ calls Facebook engagement flat and does not print a median on its page. By reach the picture flips: Buffer's median is about 5.6%, second only to LinkedIn. People who see a Facebook post react to it about as often as on Instagram. Far fewer of your followers get to see it.

Meta's measurement is also moving. As Social Media Today reported in August 2025, Meta is deprecating the impressions metric in its Page insights API and replacing it with views, which it wants as the single distribution metric across Facebook and Instagram. If an older report used impressions as the bottom, recalculate it with views before comparing.

Do not chase the rate with bait. Meta began demoting posts that ask people to like, comment or share in 2017, extended it to comments and to video audio later, and says Pages that do it repeatedly face stronger demotion. A higher rate that costs you reach is a bad trade.

If you read the last cards and thought "we should be on Threads and Bluesky, but nobody has time for four more feeds", that is the usual reason those accounts sit empty. Every extra network means another caption length, another image size, another app to open at nine in the morning.

Spreadcast publishes to all 11 networks on this page from one calendar: Instagram, Facebook, Threads, LinkedIn, Pinterest, TikTok, YouTube, Bluesky, Mastodon, Telegram and Discord. It designs the image in your brand and writes a caption for each network that fits its limits, so a post goes to Threads, Bluesky and a Telegram channel in the same pass as Instagram. Pro is $29 a month for 10 social accounts in any mix, and the pricing page lists what each plan includes.

Put the studies side by side and Instagram's average engagement rate is 0.30%, 0.45% or 5.5%, depending on whom you ask. None of them is wrong. Four choices explain almost all of the gap.

The bottom of the fraction. Socialinsider and Rival IQ divide by followers. Buffer divides by reach. Socialinsider's LinkedIn and TikTok studies divide by impressions and views. Reach is always smaller than followers on a brand account, so reach based rates are always higher, roughly ten times on Instagram and Facebook.

Average or median. Socialinsider reports averages, which a handful of viral posts can lift. Rival IQ and Buffer report medians, the typical post. Expect averages to sit above medians.

Who is in the sample. Socialinsider covers brand accounts with 1,000 to 1 million followers. Rival IQ picks 150 companies at random in each of 18 industries, from a database of over 200,000, and includes paid posts. Buffer's data comes from accounts that publish through Buffer and posted at least 10 times in the past year, across at least four different weeks. Different accounts, different results.

Which interactions count. Socialinsider counts only likes and comments on Instagram, while Buffer and Meta's own total add shares and saves. On LinkedIn, Buffer's figure includes clicks and Socialinsider's does not. More counted interactions, higher rate.

The practical rule: compare your number only with a benchmark built the same way, and trust your own trend over any study. If your median rate by reach went from 3% to 4% over a quarter, that is real progress, whatever a report says the network average is. For the bigger picture behind these numbers, our social media statistics page collects user and usage data for each network.

A good engagement rate is one above the typical post on the same network, measured the same way, for accounts of your size. That sounds like a dodge, but it is the only definition that survives contact with the studies above. 1% by followers is excellent on Facebook and below par on TikTok. 3% by reach is poor on LinkedIn and fine on Threads.

Here is a workable shorthand for brand accounts, built from the 2026 figures. By followers: above 0.5% on Instagram is good, above 0.2% on Facebook is good, above 2.5% on TikTok is good. By reach: above 5% on Instagram, Facebook or LinkedIn is good, above 4% on TikTok or Pinterest is good, above 3.5% on Threads is good. Creators can expect more, and a rule like "anything above 1%" for Instagram creators is too low to mean anything.

Is 2% a good engagement rate? On Instagram by followers, yes, about four times the brand average. On TikTok by followers, it is about the median. By reach on any network, it is below the typical post. Is 10% good? By followers on Instagram it is unusual enough to check for a giveaway or bought engagement. By reach on a small account it happens on posts that hit a nerve.

Then there is the better question: good compared with what you did last month? Benchmarks help you notice if something is badly off. Your own median, tracked over months on a fixed formula, tells you whether the work is improving. Most teams that track engagement well look at one number per network per month and one line of notes on what changed.

A good rate on the wrong goal is still the wrong goal. If you post to drive app installs, sign ups or sales, engagement rate is a leading signal, not the result. Put it next to clicks to your link and followers gained, and judge the posts by the business number when you have one.

Counting your own replies. A comment thread where you answered every person doubles the comment count. Either subtract your replies or count unique commenters. The calculator assumes you entered other people's interactions only.

Using today's follower count for old posts. An account that grew from 2,000 to 6,000 followers this year will see its January posts drop to a third of their real rate if you divide them by today's number. Note the follower count when you post, or use reach, which is fixed per post.

Mixing paid and organic. A boosted post reaches people outside your audience on a budget, and its reach and interactions mix into the post's totals. Report organic and paid separately, and compare organic posts only with organic benchmarks such as Socialinsider's. Rival IQ includes paid posts in its sample, which is one more reason its figures and Socialinsider's differ.

Comparing across formulas. A 4% rate by reach next to a 0.4% competitor rate by followers is not a win. Run the competitor's public numbers through the calculator by followers first.

Including giveaway posts in the average. A contest with "tag three friends" can show ten times your normal rate. Leave it out of the median or mark it, or it will hide what your ordinary posts are doing.

Ignoring counting changes. Instagram's API stopped reporting impressions for posts made after July 2, 2024, Meta is moving Facebook to views, and YouTube changed when a view counts in 2026. A sudden jump or fall in your rate around one of those dates is worth checking against the network's change log before you change your content plan.

Ignoring bots. An account that bought followers has a low rate by followers and a suspicious pattern of short comments. If you vet influencers, the rate is a quick filter, not proof. Look at who is commenting.

From the outside you only see what the network makes public, and that decides the formula. Reach, saves and impressions are private on every network, so a competitor's engagement rate is always by followers, and on Instagram it is likes plus comments over followers. That is also why Socialinsider's Instagram benchmark counts only likes and comments: it is the version anyone can reproduce.

Open the account and take its last 12 regular posts. Skip anything from the last two days, because those are still collecting interactions. Skip pinned posts, which sit at the top for months and keep gathering likes. Skip giveaways and posts that say paid partnership, or note them separately. Write down likes and comments for each, and the follower count once.

Run each post through the calculator by followers, then sort the 12 rates and take the middle two. That median is the account's typical post. Compare it with your own median, built the same way from your own public numbers, not with your reach based rate from insights. If the account hides like counts, which Instagram lets people do, you only have comments, and the honest answer is that you cannot compare.

On TikTok and YouTube the views are public, so you can work out a view based rate for any video: likes plus comments, plus shares where the app shows them, divided by views. On LinkedIn you can see reactions, comments and reposts but not impressions, so use followers. On Bluesky and Mastodon the counts are on every post, and followers on the profile.

Then read the comments. A real audience writes sentences, asks questions and argues. A rate propped up by bought engagement shows the same short comments from accounts with no posts of their own, often within minutes of posting. If you are about to pay a creator, ask for a screenshot of reach and saves from their own insights for three recent posts. Anyone serious will send it.

Here is what tracking engagement by hand looks like for a small team posting to four networks. Once a week someone opens Instagram insights, LinkedIn page analytics, the Threads app and Pinterest analytics, copies likes, comments, shares, saves, reach and followers for each post into a sheet, fixes the rows where a number was missed, and works out the rates. That is an hour a week for a handful of posts, and it is the first task skipped when things get busy. The posting itself takes longer: an image designed per post, a caption per network, each one uploaded by hand.

Here is what it looks like in Spreadcast. Posts are designed in your brand and published to Instagram, Facebook, Threads, LinkedIn, Pinterest, TikTok, YouTube, Bluesky, Mastodon, Telegram and Discord from one calendar. The Analytics tab then shows followers and interactions for each network and likes, comments, reach and views for each post where the network shares them, with an engagement rate by followers worked out from each post's first week. Nobody copies a number into a sheet.

For an app, the posts can start themselves. Spreadcast checks your App Store listing every day, and when a new version goes live or the app hits a ratings milestone, it designs the post, writes a caption per network and publishes it on full autopilot or after one tap of approval. For anything else you type one line in Create, or pick one of twelve post types for apps, such as a testimonial, a how to or a feature post, and it designs a single image or a carousel. You can also upload your own pictures and video and schedule them.

If you work in Claude or ChatGPT, the Spreadcast MCP server lets the assistant read your analytics, design posts and carousels, write captions, and create, schedule and approve posts from the chat. Ask it which posts beat your median rate last month and to draft three more like them.

Pricing is plain. Free makes 3 posts to download for one app, with a small Spreadcast mark. Pro is $29 a month or $290 a year for clean posts with no cap, publishing to 10 social accounts in any mix, autopilot or one tap approval, and posting from Claude or ChatGPT. Studio is $79 a month for 5 apps and 50 accounts, each with its own brand. If you only need a burst, a Post Pack is $19 once for 10 posts designed and published, and a 30 Day Pass is $39 once for everything in Pro for 30 days. Paid plans have a 7 day full refund. Paste your link at spreadcast.app/start to see the first posts before you pay.

Engagement rate earns its place as the quality check between reach and results. Reach and views say whether the network showed the post. Engagement rate says whether the people who saw it cared. Follower growth, link clicks and the business number at the end say whether it paid off. A monthly report with those four lines per network, plus a sentence on what you changed, beats a dashboard with forty metrics nobody reads.

Keep the formula fixed for the whole year. If you switch from followers to reach in June, the line jumps and the comparison with the first half of the year is gone. If the network forces a change, as Meta did with views, note it on the report the month it happens.

Use the benchmarks twice a year, not every week. Check whether you are roughly in line with your network and size, then go back to beating your own median. And when a post does far better than usual, look at why before you celebrate the number: the format, the topic, the hour, the first line. That note is worth more than the rate. If you want ideas for the next batch, our list of 52 post ideas for an app and the guide to batching a month of content in one afternoon are good places to start.