What is a good Instagram engagement rate?

Median Instagram engagement is 0.30% per post measured against follower count, according to the 2026 Rival IQ and Quid benchmark report. Socialinsider, working from 35 million posts across 447,613 pages, puts the Q2 2026 median at 0.45%, also by follower count. Above roughly 1% by follower is genuinely good. Above 2% puts you well ahead of the median account in almost any industry.

Those numbers will feel wrong if you have read almost any other page on this question, because the folklore figure is around 3% and the folklore says under 1% means something is broken. The folklore is not lying to you. It is quoting a different formula. That is the first thing this page fixes, and it is the thing nobody else on the results page bothers to explain.

Why one source says 0.30% and another says 3%

There is no agreed definition of Instagram engagement rate. There are two, they produce answers that differ by roughly a factor of ten, and both are used by serious analytics companies in reports they publish every year.

Rival IQ and Quid divide interactions by follower count. Socialinsider does the same. Hootsuite divides interactions by reach — the number of unique accounts that actually saw the post. A post can only be engaged with by someone who saw it, so dividing by reach measures how compelling the post was to the people it reached. Dividing by followers measures something else: what share of the audience you have assembled is still showing up.

Both are legitimate. They answer different questions. What is not legitimate is comparing your by-reach number against a by-follower benchmark and concluding you are ten times better than the median, or the reverse.

The same post, run through both formulas

Take one real-shaped example. An account has 10,000 followers. It publishes a Reel. The Reel is seen by 1,200 unique accounts and collects 120 interactions in total.

• By follower count: 120 divided by 10,000, times 100 = 1.20%. Against the 0.30% median, that is four times better than typical.
• By reach: 120 divided by 1,200, times 100 = 10.0%. Against a roughly 3% by-reach median, that is more than three times better than typical.

Same post. Same likes, same comments, same saves, same shares. Two correct numbers, 1.20% and 10.0%, and an eightfold gap between them that comes entirely from what went under the line. If you screenshot the 10.0% and put it in a media kit next to a benchmark that was calculated by follower, you have made a claim you cannot defend when the brand checks it.

So the rule is short: always ask which denominator a benchmark used. If the report does not say, the report is not usable. There is a second, smaller source of divergence worth knowing about too — sources also differ on what goes into the numerator. Some count likes and comments only. Others add saves and shares. That will move your number by a few tenths of a point rather than by a factor of ten, but it is another reason two honest tools will hand you two different answers for one post.

How to calculate your own rate, step by step

Both calculations take about thirty seconds. You need a professional account — Creator or Business — for Instagram Insights to give you the reach figure.

1. 01Open the post in Insights and add up the interactions: likes, comments, saves and shares. Call that number your interactions.
2. 02For the by-follower rate, divide interactions by your current follower count and multiply by 100.
3. 03For the by-reach rate, divide the same interactions by the reach figure — accounts reached, not views — and multiply by 100.
4. 04Repeat across your last ten to twenty posts and take the median rather than the mean. One post that outperformed will drag an average upward and tell you nothing about a typical week.
5. 05Write down which denominator you used, in the same place you wrote the number. Six months later you will not remember.

A worked example you can copy the shape of: a post picks up 850 interactions on an account with 25,000 followers. That is (850 divided by 25,000) times 100 = 3.4% by follower count. Against a 0.30% median that is an exceptional post, roughly eleven times the typical rate, and it is worth checking whether something unusual drove it before you treat it as your baseline.

One caution on reach. Instagram's Views metric counts repeats — one person watching a Reel three times registers three views. Reach counts unique accounts. If you divide interactions by views instead of by reach, you get a third number, smaller than both, and it is not comparable to anything anyone publishes.

What good looks like by format

Socialinsider's 2026 dataset is the best-documented methodology publicly available for this question: 35 million posts across 447,613 pages, all measured by follower count. The format split matters more than most account-level advice does.

Overall engagement in Q2 2026 was 0.45%, down about 24% year on year. Carousels have held up best and still lead. Reels barely moved. Single images have lost more than a quarter of their engagement rate since 2024 and are now clearly the weakest of the three.

Read the table as a target, not as a verdict. If your carousels run at 0.6% and your single images run at 0.3%, you are not doing worse on images than everyone else. You are doing exactly what the platform average does, and the useful conclusion is about format mix rather than about your captions.

The decline is the story

The single most important thing to understand about your 2026 engagement rate is that it fell, and so did nearly everybody's. Rival IQ and Quid recorded the median dropping from 0.36% to 0.30% by follower — a 17% year-on-year decline. Socialinsider recorded a steeper 24% drop on its own overall figure. Two independent datasets, two different methodologies, same direction.

Posting frequency fell as well. The median account now publishes 3.69 posts per week, a second consecutive annual decline. Brands and creators are posting less and getting less back per post.

If your own rate dropped this year, the first hypothesis to test is not that your content got worse. It is that you moved with the platform. Compare your 2026 rate against the 2026 median, and compare your own trend against the direction of the industry trend. That is a fair comparison. Comparing against a screenshot someone posted in 2022 is not.

The tier ladders everyone quotes are not sourced

Search this question and you will find a table on almost every page giving engagement rate by account size — something like nano 6.23%, micro 3.86%, mid 2.4%, macro 0.4%, mega 1.2%. Those tables copy each other. Trace them back and they do not terminate in a published report with a stated sample and a stated denominator. They terminate in other blog posts, and a good number of them appear to be machine-generated for search traffic.

We are not printing one, because we cannot show you where the numbers came from. The one tier figure that does trace to a named source is HypeAuditor's: nano influencers at 1,000 to 10,000 followers average 2.19% engagement, and 76% of all Instagram influencers are nano. That is a real figure from a company that audits accounts for a living, and it is the only one on this page in that shape.

The underlying pattern the fake ladders are gesturing at is real: smaller accounts do tend to post higher engagement rates by follower, because a small follower list is more likely to be people who actually chose you. But the precise decimals are invented. If a number carries two decimal places and no sample size, treat it the way you would treat a price with no currency on it.

Why a cross-industry median is a weak comparison for you

Even with the denominator settled, a single platform-wide median is a blunt instrument. The Rival IQ data shows the spread between industries is more than tenfold, top to bottom. Higher education runs around 2.10%. Nonprofits sit near 0.56%. Both are medians of real accounts, measured the same way, in the same report.

A university account at 1.8% is underperforming its category badly while sitting six times above the platform median. A nonprofit at 0.9% is doing well while looking mediocre against the same yardstick. The platform median is useful for deciding whether a number is plausible. It is close to useless for deciding whether your account is doing a good job.

What brands actually measure

If you care about engagement rate because you want brand deals, the priority order is worth knowing. The ANA surveyed 84 client-side marketers and asked which KPIs matter most for influencer work. Engagement rate came top at 90%. Impressions followed at 87%, reach at 86%. Follower count does not appear on the list at all.

That is worth pairing with the pay data, which points the same way: the rate you can charge is set by what your audience does, not by how many of them there are. We cover the numbers on that in the guide to how many Instagram followers you need to get paid.

What buying followers does to this number

We sell Instagram followers, likes and views, so take this section as coming from an interested party — and then check the arithmetic, because arithmetic does not have an interest.

Engagement rate by follower is interactions over followers. A follower purchase increases the denominator. It does not increase the numerator, because purchased followers did not find you through your content and will not reliably interact with it. So buying followers on their own mechanically lowers your engagement rate. That is not a platform penalty and we are not going to dress it up as one. Instagram has never described account-level engagement rate as a ranking input, and anyone telling you it does is inventing a mechanism. It is just division.

Row two is what a follower-only purchase does: a 37% fall in the visible rate, with the account's actual performance unchanged. Row three is the coherent version. Followers went up 60%, so interactions went up 60%, and the ratio held. That is what the profile would look like if the growth had been real.

Which is the honest case for pairing. If you are adding followers, adding engagement in proportion is what keeps the profile internally consistent — the 6,000-follower package alongside a likes package rather than followers alone. The 500-like package spread across several recent posts is the small version of that. If you are only buying one thing, our guide on which product to buy first works through the decision properly, including the cases where the answer is neither.

The limitation worth stating plainly: no package fixes an engagement rate. It can keep one from visibly collapsing while a follower count moves, which is a narrower and more honest claim.

What changed in 2025 and 2026

Four changes since the start of 2025 affect whether your number means what you think it means.

• Views replaced impressions and plays. Announced August 2024, with the API cutover on 21 April 2025. Views is now the primary metric across Reels, Live, photos, carousels and Stories, and it counts repeats. Any benchmark published before August 2024 was built on impressions or plays and is not comparable to a views-based number today.
• Engagement kept falling. Down about 17% year on year in the Rival IQ data and about 24% in Socialinsider's, with posting frequency down to 3.69 posts per week. Two consecutive years of decline in both.
• The May 2026 account sweep moved denominators. Overnight on 6-7 May 2026 Instagram removed a large volume of bot, spam and inactive accounts. Typical small and mid-sized creators lost 2% to 5% of their follower count. If you calculate by follower, a smaller denominator with the same interactions means your rate went up in May without you doing anything. Our guide to losing followers overnight covers what happened and what Meta said about it.
• The originality rules tightened on 30 April 2026. Accounts that repeatedly repost other people's work are no longer eligible for recommendations across the app, now including photos and carousels rather than Reels alone. Followers still see the posts; non-followers do not. That compresses reach, which raises a by-reach engagement rate while leaving a by-follower rate flat — another reason to know which one you are looking at. If you think this has hit you, the Account Status check is where Instagram actually tells you.

The practical upshot for 2026: recalculate your baseline from your own last twenty posts, in the format mix you actually publish, using one stated denominator. Compare it to the 0.30% and 0.45% by-follower medians and to your own industry if you know it. Then ignore every number you have ever seen quoted without a sample size.