Instagram engagement rate: how to calculate it and what counts as good

Engagement rate is the number brands, managers and creators ask about first. It is also one Instagram never defines. There is no “engagement rate” field in the app and no official target. What exists are the raw counts in Insights and a handful of formulas the industry agreed on. Pick the wrong one and you will compare a reel’s rate with a carousel’s and draw the wrong conclusion.

This guide gives you the formulas, shows where each input lives in Insights as of October 2026, and explains why “what counts as good” depends on which formula someone used.

Which formula should you use?

All four start the same way: add up interactions on a post, divide by an audience number, multiply by 100. They differ in the audience number.

The by-followers version is what most published benchmarks use. Rival IQ’s 2025 report, for example, defines engagement rate as likes, comments, shares and similar interactions divided by total follower count.

The by-reach version is the one to use when you are deciding what to post next. It answers a direct question: of the people Instagram showed this to, how many did something?

Likes are usually the largest part of the numerator, which is why the rate moves when likes move. That includes paid likes. If you ever buy Instagram likes for a launch post, mark that post in your tracking sheet and leave it out of your median. Otherwise your baseline drifts upward, and every normal post after it looks like a failure.

Where do the numbers come from in Insights?

Instagram’s Insights glossary, published in the Meta Business Help Center, defines the inputs you need.

• Interactions: actions people take on your posts, such as likes, comments, saves and shares.
• Accounts engaged: unique accounts that interacted with your posts.
• Views: the number of times a post was played or displayed. For reels, Instagram’s help page says it is the number of times a reel starts to play or replay.
• Viewers: unique accounts that saw the post on screen at least once. This is the number older guides call reach.

You can pick any preset or custom range within the past 90 days. Follower trends appear once you have at least 100 followers, and demographic breakdowns need more than 100 accounts reached or engaged.

If you pull numbers through Meta’s API instead, the total_interactions field is likes, saves, comments and shares, minus unlikes, unsaves and deleted comments. That is a cleaner figure than adding the public counts yourself.

The views metric is newer than many guides admit. In August 2024 Instagram said it was “introducing ‘Views’ as the primary metric across all formats including reels, live, photos, carousels, and stories.” Reels plays were relabeled as views, and repeat views now count on photos and stories too. Any by-views rate from before that change is not comparable with one from after it.

A worked example

Here is one carousel, with made-up round numbers to show the arithmetic.

• By followers: 235 ÷ 8,000 × 100 = 2.9%
• By reach: 235 ÷ 3,100 × 100 = 7.6%
• By views: 235 ÷ 4,400 × 100 = 5.3%

Same post, three different answers. None is wrong. The mistake is quoting one and comparing it with a benchmark built on another.

What counts as a good engagement rate?

The honest answer: one that beats your own recent median, measured the same way.

Published benchmarks are useful for direction and poor as targets. They sample different account types, count different actions, and divide by different numbers. Rival IQ’s 2025 study, for instance, looked at brands across 14 industries, divided by followers, and found Instagram engagement fell 16% compared with the year before. It also found carousels beat reels on engagement for brands in that sample. A solo creator with a close community, a news publisher and a retailer will sit far apart on any such chart, and that tells you little about your next post.

A more useful routine:

1. Pick one formula. Use by reach for your own decisions.
2. Calculate it for your last 20 to 30 posts, excluding posts that had paid promotion or paid engagement.
3. Take the median, not the average. One viral post skews an average for months.
4. Score each new post against that median after the same number of days, for example seven.
5. Split the median by format. Reels, carousels and single photos rarely land at the same rate.

If a format sits above your median three times in a row, make more of it. That is the decision engagement rate is good for.

Which interactions should count for more?

A like and a send are both “one interaction” in the formula. They do not mean the same thing to Instagram.

In January 2025, Adam Mosseri named watch time, likes and sends as the top three ranking signals. He told creators to track likes per reach and sends per reach, and said sends matter slightly more for reaching people who don’t follow you. Saves are a strong sign that a post is useful, and comments show that people cared enough to type.

So add two side ratios next to your engagement rate:

• Sends per reach: shares ÷ viewers. This tracks how far a post can travel beyond your followers. See Instagram shares and sends for why.
• Saves per reach: saves ÷ viewers. This tracks whether the post is worth coming back to.

A post with a modest engagement rate and a high sends per reach often grows the account more than a post with lots of likes from existing fans.

What makes an engagement rate look worse than it is?

Several things lower the number without saying anything about the post.

The hidden-likes row matters more than people think. Since May 2021, anyone on Instagram can hide like counts on their own posts. Public engagement rate tools can’t see those likes, so they underreport. Your own Insights still show the full number.

The last row is the honest side of buying followers. Accounts that deliver follows do not like or save your posts. A follower order raises the denominator and leaves the numerator alone, so a by-followers rate will drop. If you plan an order, note the date and switch to by-reach for the following weeks. The trade-offs are covered in Instagram followers or likes.

How often should you check it?

Weekly is enough for most accounts. Check each post at the same age, since Insights keep counting for days after posting, and a two-day-old post will always look weaker than a two-week-old one. Instagram’s reel insights stay available for two years after a reel is published, so you can rebuild a median from older posts without third-party tools.

If you manage several accounts, keep a simple sheet: date, format, viewers, interactions, sends, saves, and a column marking any post that had paid promotion or paid engagement. That one column keeps the rest of the sheet honest. More Instagram measurement guides are in the Instagram section of the blog.