How to Find Viral Posts on Instagram (Without Sorting by Likes)

To find viral posts on Instagram, compare each post with what that account usually gets, not with other accounts. A post that triples its creator’s usual tells you the idea worked. A post with a big like count often only tells you the account is big.

One week of our data shows how far apart those two things can be, and the method below works whether you do it by hand or with a library.

Why are likes a bad way to find viral posts?

Take the 12 biggest outliers in our library posted in the week to 9 October 2026. Sorted by likes and comments, the list starts with @victorianpoetry’s carousel at 121,155. Sorted by how far each post beat its own creator’s usual, that same carousel drops to eighth, at 2.7×.

At the bottom of the likes list sits “Canva Apps you should know PART 1” by @contentwithsprouthive: 536 likes and comments. It scored 7.7×, second only to the week’s top outlier.

Here is what the same week looks like when you work out each creator’s usual (likes + comments divided by the score):

• @victorianpoetry: 121,155 likes + comments, 2.7×, so a usual of about 44,900
• @case.kenny: 86,501, 2.6×, a usual of about 33,300
• @paintingsanalyzed: three carousels at 55,048 to 58,103 each, all 2.5–2.6×
• @jasonrewired_: 9,766, 5.2×, a usual of about 1,880
• @contentwithsprouthive: 536, 7.7×, a usual of about 70

A like count can’t tell these apart. A 2.7× post from a big account and a 7.7× post from a small one look like a hit and a flop, when the small one beat its baseline by far more.

The same account can produce both kinds. At the end of September, @victorianpoetry posted “Don’t get too drunk. Don’t drink from an open cup.”, which reached 2,385,045 likes and comments: 54× the usual. Same account, three days apart. One post beat the usual by 2.7×, the other by 54×. Only the comparison with the account’s own track record shows which idea broke out. If the maths behind the score is new to you, how the outlier score works goes through it step by step.

Where people usually look, and what it misses

• Hashtag pages, “top posts” lists and anything ranked by raw likes or views. Big accounts win these by default, so you learn who is big, not which idea worked.
• A competitor’s grid. You can see the numbers (when likes aren’t hidden), but there’s no baseline until you work one out.
• Engagement rate (likes ÷ followers). Better than raw likes, but it compares a post with a follower count, not with what that account’s posts normally get.

The fix for all three is the same: one baseline per account.

How to find viral posts in your niche: a six-step method

1. Pick 15 to 30 accounts, and mix the sizes

Include small accounts on purpose. @contentwithsprouthive’s 7.7× came from an account whose usual is around 70 likes and comments. A list made only of big names would never surface it.

Add one or two neighbouring niches as well. A format that’s ordinary in one niche can be new in yours.

2. Work out each account’s usual

Take the median likes + comments of the account’s recent carousels. We use up to 20 carousels from the same creator, the ones posted closest in time, and need at least 7 before we score anything. Compare carousels with carousels. Mixing in Reels gives you a baseline for a different kind of post.

3. Leave out the posts that distort the baseline

Our score skips paid partnerships, giveaways and posts with comments turned off when it builds a creator’s usual, and it ignores posts under a week old, which are still collecting engagement. Do the same by hand. A giveaway with 10,000 comments tells you nothing about the next normal post.

4. Decide on a bar before you start

Across the 9,923 carousels we’ve scored:

• 49.3% beat their creator’s usual at all (1× or more)
• 23.8% reached 2×
• 14.9% reached 3×
• 8.5% reached 5×
• 3.6% reached 10×

So 1× is the floor, not a result: about half of anyone’s posts land there by definition. A 3× carousel is roughly one in seven, which makes it a sensible bar for saving a post. One caveat: these rates come from a mix of creators we track, not a random slice of Instagram.

5. Check the baseline behind any huge multiple

A score is only as steady as the usual underneath it. This month’s biggest multiple in the library is a food carousel by @heatherjanecollum at 2,894.7×, with 68,025 likes and comments. That puts the creator’s usual at about 23.

It’s a real outlier. But a multiple that size says as much about the small baseline as about the post. Look at the absolute numbers next to the score, and give more weight to patterns you see across several accounts than to one extreme result.

6. Write down why it worked, then repeat weekly

For every post that clears your bar, note three things: the format (list, how-to, story and so on), what the first slide promised and how many slides it used. After a few weeks those notes become your list of proven structures.

New carousels keep arriving, and a post’s numbers keep moving in its first days. That’s why the library re-checks carousels at about 4 days and again at 8–12 days old. By hand, a weekly pass over your list is enough.

Should you do it by hand or use a library?

By hand it costs nothing and you see every post yourself. It’s also slow: 20 accounts at 25 carousels each is 500 posts to log, and you only ever see accounts you already know.

With a library, the scoring is already done. Top Carousels holds 3,924 outlier carousels from 192 creators across 13 niches, each scored against its own creator’s usual. A free account browses outliers by niche, five at a time. Pro opens the whole library, with filters for score, likes and comments, slide count, date, niche and format, plus saved searches. Either way you can keep the best references in collections with private notes.

For a single account, the account analyzer reads any public account’s last 40 carousels and shows which ones beat that account’s usual, scored the same way. It needs at least 7 carousels that have been up for a week, and it can’t read private accounts or Reels. A free account can analyse 1 account; Pro analyses 10 a month.

What to do once you’ve found one

Don’t copy it. Study what would survive if you removed the topic: the first-slide pattern, the format, the slide count, the ending. The library holds 854 outliers posted in the last 30 days alone, from 139 creators, with a median of 8 slides and a median score of 1.9×, so there’s rarely a shortage of references.

Then write your own version about your subject. The hook patterns behind outlier carousels help with the first slide, and I wrote up a full remix from reference to finished post, which got 3K+ likes in 3 days.

Frequently asked questions

How do I find viral posts in my niche on Instagram?

List 15–30 accounts in and around your niche, including small ones. For each, work out the median likes + comments of its recent carousels, then divide every post by that number. Anything at 3× or more is worth saving. Only 14.9% of the 9,923 carousels we scored got there.

Why shouldn’t I sort Instagram posts by likes?

Because likes mostly measure audience size. In the week to 9 October 2026, a carousel with 121,155 likes and comments scored 2.7× its creator’s usual, while one with 536 scored 7.7×. Sorted by likes, the stronger result came last.

What counts as a viral post on Instagram?

There’s no fixed number. A useful definition is a post that beats its own account’s usual by a wide margin. In our data, 23.8% of carousels reached 2× their creator’s usual, 14.9% reached 3× and 3.6% reached 10×.

Can I see which of a competitor’s posts went viral?

Yes. By hand, work out the median likes + comments of their recent carousels and compare each post with it. Top Carousels’ account analyzer does this for any public account’s last 40 carousels; a free account can analyse one account.