Outlier Detection — Free

Instagram Viral Post Analyzer

Type a public handle. We compute the account's own median engagement, then surface every post that landed at more than double the baseline — with the exact multiplier shown next to each one.

Try:
Engagement signal·30-post sample·Free forever·0 signup
2x
Viral threshold
Median
Baseline math
0
Login needed
instagram-viral-post-analyzer
Features

What the analyzer surfaces on every run

Median baseline, viral threshold, every outlier with its multiplier, and the format pattern hiding inside the hits — computed live, no spreadsheet.

Per-account median baseline

The middle post in the recent window sets the ruler — not an industry average, not a follower-count guess, the account's own steady-state.

2x viral threshold

Any post whose likes-plus-comments crossed twice the median gets flagged with its exact multiplier — 2.3x, 4.7x, 9.1x — right on the card.

Outlier card stack

Each flagged post shows its thumbnail, caption snippet, raw likes, raw comments, and the ratio against baseline in one tidy row.

Format pattern read-out

Compare hashtag counts, caption length, and post-type mix between the viral set and the steady-state set — the recipe hides in the gap.

Empty-window safety

When no post crosses the 2x line, the page says so clearly and points to a different account — no hallucinated “winners” out of thin air.

Zero setup, zero signup

Type the handle. The math runs in the browser the moment the public feed lands. No account of yours, no install, no waiting room.

How It Works

Handle in, outliers out — in three motions

Outlier detection — posts that scored more than 2x the median engagement, with a pattern read.

1

Type a handle

Any public handle. The analyzer fetches recent posts and computes the engagement baseline.

Public handleNo login
2

We flag outliers

Median engagement = baseline. Threshold = 2x baseline. Anything above the threshold is flagged as a viral outlier.

engagement > 2 × median
3

Read the pattern

Each outlier appears with its multiplier, plus a pattern read on what the viral posts have in common.

>2xPatternTagsLength
The Four Reads

Four signals the analyzer pulls out of every feed

Baseline, threshold, outlier list, format pattern — the four numbers a creator actually needs to copy what worked.

Median baseline

The account becomes its own ruler.

Most analytics tools measure your posts against an industry slab — a number that means nothing to a niche creator. This analyzer takes the median of the recent feed instead. The middle post defines steady-state, and every other post is judged against that single, honest number. A small account and a celebrity account each get a fair read.

  • Median resists single huge posts that warp the average
  • Account-relative, never industry-relative
  • Recomputed on every fresh pull, never cached stale
The account becomes its own ruler
2x threshold

The viral line. Finally pinned down.

Posts that doubled the median are unmistakably out of pattern — they did something the rest of the feed did not. The 2x line is the threshold marketing teams have been eyeballing for years, made explicit. A post at 1.9x is steady-state with a good day. A post at 2.0x crossed into outlier territory and deserves a closer look at format, caption, and timing.

  • Explicit threshold — no “feels viral” guesses
  • Multiplier shown for every post that crossed the line
  • Resists noise — a single big post can't move the line
The viral line that everyone agrees on, finally pinned down
Outlier list

Viral hits. One stack. With multipliers.

Every flagged post lands in a single ranked stack: thumbnail on the left, caption snippet in the middle, multiplier badge on the right. Sorted descending so the biggest outlier sits at the top. Click any card to open the original post on Instagram in a new tab. No carousel of fluff metrics — just the posts that did the actual work and the number that says by how much.

  • Sorted from highest multiplier downward
  • Caption snippet and raw counts on every card
  • One click to open the original post
The viral hits, in one stack, with their multipliers
Pattern read-out

Format signal. Hidden inside the winners.

The analyzer compares the viral subset against the steady-state subset on three axes: hashtag count, caption length, and post-type mix. When all the outliers carry three hashtags and the baseline carries fifteen, that's a recipe signal. When every outlier is a reel and the baseline is mostly images, that's another. The page surfaces the gap so you can copy what worked instead of guessing.

  • Hashtag count: viral vs steady-state
  • Caption length: viral vs steady-state
  • Post type distribution: reels vs carousels vs images
The format signal hiding inside the winners
Plain Guide

A plain guide to reading viral outliers

A viral post is not the post with the most likes — it's the post that out-performed what the same account normally posts. That single shift in definition is what the analyzer is built on, and it changes which posts get attention.

Why the median, not the average

Averages get yanked around by a single huge post. The median is the middle value — sort the feed by engagement, pick the post in the middle, and you have a number that ignores the outlier you're hunting for.

Why 2x is the line, not 1.5x or 3x

The 2x line is the convention working creators and growth teams settled on years ago — small enough to flag real momentum, large enough to ignore a lucky comment thread. Below 2x is a good day. At 2x or above, something landed.

What the pattern panel is actually saying

If the viral posts all share a trait the baseline lacks — three hashtags instead of fifteen, a 60-character caption instead of 600, the reels format instead of carousels — that trait is the candidate variable. Worth a test.

When no outliers means no outliers

Some feeds are calm by design. If nothing crossed the 2x line, the page says so plainly and suggests trying a different account. There's no fake winner inserted to fill the panel.

Feature Detail

Seven jobs the outlier engine handles

One handle in — seven outputs every working creator needs to copy what landed.

Median computation

The recent feed is sorted by engagement and the middle post fixes the baseline. Stable, single-number, account-relative.

2x threshold flag

Anything at or above twice the median lands in the viral set with its precise multiplier on the card.

Multiplier badge

2.4x, 4.7x, 9.1x — the exact ratio against baseline, never rounded into a vague “hit” label.

Hashtag count compare

Average hashtag count on viral posts versus steady-state posts — the recipe signal nobody else surfaces.

Caption length compare

Mean caption characters in the viral set versus the baseline set — long-form versus snap.

Format distribution

Reels versus carousels versus single images inside the outlier subset — the format that broke pattern.

Empty-window honesty

If nothing crosses 2x, the page says so and suggests a different handle — no padded winners list.

Side by Side

Why outlier math beats raw “most likes” sorting

What you actually want to learn Sorting by raw likes This analyzer
Catch posts that out-performed the account itselfNoYes
Show the exact multiplier above baselineRaw count onlye.g. 4.7x
Resist warping by a single huge postSkewedMedian holds
Fair to a small creatorTiny absolute countsAccount-relative
Surface the format pattern behind the winsNot shownSide panel
CostPaid analytics suiteFree, no signup
Who Uses It

Who keeps the analyzer in their morning loop

Growth teams

Weekly competitor sweeps — which competitor posts broke the 2x line, what format did they share, what should we test next sprint.

Solo creators

Reading their own feed for the same outlier signal — not which post got the most likes ever, but which post out-performed the recent norm.

Influencer scouts

Triaging a long candidate list — creators with consistent outliers are worth a brief; creators with flat feeds rarely move a campaign.

Editorial planners

Spotting a post that punched above weight in a niche, then commissioning a follow-up before the story cools off.

Brand strategists

Reading what a sponsorship target's outliers have in common — tone, format, hashtag count — before drafting the brief.

Inside The Engine

The five stages between handle and outlier panel

A clean pipeline — pull, sort, threshold, classify, render. Every step explicit, every number traceable.

1

Pull the recent feed

Backend hits the public profile endpoint and returns the most recent batch — usually the last twelve posts with likes and comments attached.

2

Sort and pick the middle

Each post gets a combined engagement number — likes plus comments. The list is sorted, the middle value is taken, and that becomes the median baseline.

3

Set the viral threshold

Baseline times two becomes the viral line. Every post is tested against that single number — above or below. No fuzziness, no judgement call.

4

Compute multipliers and patterns

For each viral post, engagement divided by baseline gives the multiplier. Hashtag counts, caption lengths, and post types are pooled across viral versus baseline subsets.

5

Render and close

Baseline box, threshold box, sorted outlier stack, pattern panel. Nothing stored after the page paints — recompute is one search away.

Active Users

Three short notes from working analysts

“The median baseline is the part that sold me. I was sorting feeds by raw likes for years and missing the actual signal — a 3.4x outlier on a small account is a way better lead than the celebrity's best week.”

— Wren Kalanick, Growth Lead

“I drop ten competitor handles into this every Monday. Three minutes later I have a list of every post that out-performed each account's own norm, with a multiplier I can put in a deck.”

— Otis Donnelly, Brand Analyst

“The format pattern read-out caught something I'd been missing on my own feed — every one of my outliers had a sub-80-character caption while my baseline averaged 320. That was the variable.”

— Sable Crowther, Independent Creator

FAQ

Direct answers to common questions

A post is flagged viral when its likes plus comments come in at twice the account's own median for the recent window. The baseline is the account, not an industry average, so a smaller creator can still register viral hits without being measured against a celebrity feed.
Average engagement is dragged around by one or two huge posts — exactly the posts we're trying to flag. The median is the middle value, so it stays steady even when an outlier exists. That makes the 2x multiplier honest instead of self-cancelling.
The window is whatever the public feed endpoint returns on the first pull — typically the most recent 12 posts. That's a wide enough window to spot trends and a small enough window to keep the multiplier responsive to fresh changes in the creator's output.
No. The feed is gated at the platform's own server for private accounts, so no third-party reader can compute a baseline on a profile it cannot see. The page will say the profile is private and stop there — clean failure, no fake numbers.
Every search triggers a fresh pull through the same public feed endpoint, then the median and the multipliers are recomputed on the spot. There's no stored snapshot layer to go stale — the number you see is the number that exists at request time.
Some feeds are remarkably consistent — each post lands in a tight band around the median, and nothing crosses the 2x line. That's not a bug. It usually means the creator has a loyal, predictable audience rather than format experiments that occasionally pop.
No, because the baseline is the account's own median, not a global threshold. A profile averaging 200 engagements will register a 480-engagement post as viral; a profile averaging 50,000 needs a 100,000+ hit to register the same multiplier.
Reels and images are pooled into a single engagement number — likes plus comments per post — because that's how a creator's audience reacts on the surface they see. The pattern panel does break out post-type distribution separately so you can read format intent.
Yes. Each viral card shows the post's raw likes and comments, the multiplier ratio against the baseline, and the engagement total used in the comparison. There's nothing happening behind a curtain — the arithmetic is visible inline.
Outlier detection on the public feed is free with no signup, no card, no daily cap. The save-to-dashboard layer for tracking an account across sessions is the paid path, but the live analysis runs without an account at all.

Drop a handle. Read the outliers. Copy what worked.

The median is the line. Twice the median is the threshold. Above that, the analyzer shows you every post that broke pattern — and the format trait they shared.

Run an Analysis See What the Engine Surfaces

What users are saying

4.8 · 12 reviews
Vikram J.
★★★★★

Used the engagement-rate calculator before pitching to a brand for sponsorship. Came in with hard numbers instead of vague growth claims. They signed.

Hannah C.
★★★★★

The top-posts analyzer is gold for content strategy. Pulled my own account and saw which post types actually drive saves. Re-balanced my calendar that week.

Sarah K.
★★★★★

Compare-accounts feature won me a client. Walked into a sales call with a side-by-side of their account vs three competitors. Closed in one meeting.

Brandon L.
★★★★½

Solid metrics across the board. Would love an export-to-CSV button for client decks, but the on-screen visuals are already excellent.

Priya S.
★★★★★

The viral-post analyzer told me exactly what hook structure was working for a creator I was studying. Reverse-engineered it for my own niche.

Owen P.
★★★★★

Follower-to-following ratio check helped me spot a likely fake-engagement influencer before a partnership. Saved budget and reputation.

Aanya M.
★★★★★

Engagement-rate calculator gave me a real benchmark instead of guessing whether my numbers were good. I am at 3.2% which apparently is excellent. Glad I know now.

Mike R.
★★★★½

Great metric breakdown. Would be nice to track changes month over month in one view, but the per-snapshot data is already very useful.

Tom B.
★★★★★

Most-liked posts analyzer is a quick win for understanding any account in 60 seconds. I use it for prospect research weekly.

Mariana C.
★★★★★

Competitor analysis showed me my main rival was actually losing engagement quarter over quarter even though follower count was growing. Confidence boost when I needed it.

Lucas O.
★★★★½

Numbers feel accurate vs what I see in Creator Studio. Would love Portuguese-language label support, but the data itself is excellent.

Reggie M.
★★★★★

I use the top-posts tool every Monday morning on six accounts I track. Five minutes total. Replaces a whole research session.