There is no fixed number. A post with 40 likes on Instagram can be excellent and a post with 4,000 can be limp, because a like only carries meaning next to the number of people who actually saw the post. The measure that works is engagement over reach. Add up the likes, comments, saves and shares on a post, divide that total by the accounts reached, then multiply by 100. Anything sitting at or above your own median for your last dozen posts is a healthy result. Measured this way, a 900 follower account regularly beats a 90,000 follower account, and the raw like counts you have been comparing yourself against tell you almost nothing on their own.
Table of Contents
Three denominators produce three different answers for one post
When somebody quotes an engagement rate at you, the first question is what they divided by. Take a single post with 800 likes and 60 comments, published by an account with 20,000 followers, which reached 12,000 accounts and picked up 30,000 views.
| Divided by | Sum | Result | Who uses it |
|---|---|---|---|
| Followers (20,000) | 860 / 20,000 | 4.3% | Brands, media kits, most published benchmarks |
| Reach (12,000) | 860 / 12,000 | 7.2% | Creators judging whether a post worked |
| Views (30,000) | 860 / 30,000 | 2.9% | Video and Reels reporting |
Same post. Same likes. Three numbers, and the highest is two and a half times the lowest. This is why two people can argue about whether a post did well and both be right. Instagram’s own Insights leads with reach and views these days, where reach counts the unique accounts that saw your content and views counts the times it was displayed. A media kit that quotes 4.3% and a creator who says the post hit 7.2% are describing the same afternoon.
The tidy benchmark you were about to be handed
This is normally the paragraph where an article gives you a comfortable figure. Somewhere between 1 and 3 percent is healthy, it says, and under 1 percent means you are in trouble. You were waiting for me to repeat it, weren’t you? I am going to leave it alone, because a benchmark without its denominator is worthless, and because those published averages are moving downward every year anyway. Rival IQ’s 2025 benchmark report calculates engagement as total interactions divided by follower count, and found Instagram engagement dipped across every industry it tracks, by roughly 16 percent year on year. A figure you memorised from an article two years old is describing a version of Instagram that no longer exists.
Build your own benchmark in about ten minutes
Your own history is a better yardstick than any industry average, because it holds your niche, your audience and your posting habits steady. Open Insights, switch the window to the last 90 days, and work through your recent posts one format at a time.
- List your last 12 feed posts. Keep Reels in a separate list, since they reach non followers at a completely different rate.
- For each one, note the accounts reached and the total interactions.
- Divide interactions by reach for each post, so you end up with 12 percentages.
- Take the median rather than the average. One viral post drags an average somewhere useless.
- Write that median down. That is your number now. A post above it worked, a post below it did not, and the argument ends there.
Do it again every couple of months. A median that drifts down while your follower count climbs is telling you something useful about who has been joining.
Why the rate falls as an account grows
New creators find this deeply unfair, and it is worth understanding rather than resenting. A 500 follower account is mostly followed by people who chose it deliberately and open the app expecting to see it. At 100,000, your audience includes a large layer who followed once during a viral moment, never engaged again, and now scroll past without registering you. They still count in the denominator. They contribute nothing to the numerator.
So a 3 percent engagement rate at 100,000 followers is a far harder achievement than 8 percent at 400. Judging yourself against an account ten times your size, in either direction, produces nonsense. Compare with accounts roughly your own size in roughly your own subject.
Two accounts, identical followers, wildly different like counts
Four things explain nearly every case of this.
- Reach percentage. If one account reaches 30 percent of its followers per post and the other reaches 8 percent, the like gap follows automatically. Check yours by dividing reach by follower count across your last ten posts.
- Niche behaviour. Humour, fan accounts and pets collect likes almost reflexively. Accounts about accounting software, industrial equipment or legal advice collect saves and quiet profile visits from people who would never tap a heart. The second kind can be far more valuable per follower.
- Format. Reels are pushed to people who do not follow you, which inflates reach and deflates the rate calculated over reach. Carousels earn saves. Static posts land mostly on your existing audience.
- Audience quality. Follower packages, follow for follow rounds and giveaway spikes all leave behind accounts that will never engage. They sit in your denominator forever.
That last one is measurable. If your reach per follower has been sliding while your follower count rises, the new arrivals are dead weight, and no amount of better captions will fix an audience that was never interested.
Likes stopped being the signal they used to be
Here is the part that should change what you do on Monday. Instagram’s ranking systems have moved on from treating likes as the headline measure of a good post. Adam Mosseri, who runs Instagram, has said publicly that sends per reach is among the things he watches most closely, and saves and watch through time now carry serious weight in what gets recommended. A like is cheap. Sending a post to a friend in DMs costs the viewer something, so it says far more about whether the content landed.
The practical consequence is uncomfortable if you have been optimising for hearts. Content built to collect likes tends to be agreeable and forgettable. Content built to be sent to somebody tends to be useful, funny in a specific way, or genuinely surprising. If you set yourself a like target and hit it while your sends and saves stay flat, you have optimised the wrong dial and your reach will tell you so within a fortnight.
Bought likes break the comparison for everyone
Two things happen when likes get topped up. The obvious one is that the account you have been measuring yourself against may be inflated, which means the target in your head was never real. Scroll a suspiciously liked post and look at the comments underneath. A post with thousands of likes and four comments is telling you a story about where those likes came from.
The less obvious one matters more if it is your own account. Once you add likes to a post, your median is contaminated, and every future comparison you make against it is meaningless. Brands who audit properly are looking at your like to comment ratio and your reach figures, both of which sit outside anything a service touches. If you do use a service for Instagram likes, keep a clean baseline of untouched posts to measure against, and never let a topped up like count stand in for evidence that the content actually connected. The reach number is the one that tells you the truth.
What to send a brand when they ask for your rate
Give them both figures with the denominators labelled, over a stated window. Something like this reads as competent and stops the conversation going in circles: over the last 30 days, an engagement rate of 4.3 percent on followers and 7.2 percent on reach, across 12 feed posts and 8 Reels, with a median of 340 saves per carousel. Add your typical reach per post. Brands who know what they are doing care about that far more than the number sitting under your photos.
