Instagram profile analyzer

Instagram profile analyzer: engagement rate per post, posting frequency and what performs

Price
Free plan: 100 credits, no card. One analysis uses about 15–25 credits.
What you get
Followers, posting frequency (posts per week) and format mix (Reels, carousels, photos), median likes and comments, engagement rate per post by followers and, for Reels, by plays, the 5 best and 5 weakest of the 24 latest posts with their topics, and 3 takeaways for a brand or seller.
Coverage
Any public Instagram account, in any country. Private accounts, Stories and Instagram Insights (reach, profile visits, saves on older posts) are not available. Pinned posts, posts that hide their like count and collab posts owned by another account are left out of the rates.
Time to report
Usually 5–10 minutes.
Account
Sign-in with Google required to run it. You can read the example and fill in the form without an account.

Enter a public Instagram username. Nolyn reads the profile and its 24 most recent posts, calculates the engagement rate of every post by followers and, for Reels, by plays, measures how often the account posts and in which formats, and shows which posts carry the account, with the numbers behind each one. To work out a single post yourself, use the free Instagram engagement rate calculator.

Example: @owala
Analyze a profile

Public accounts only. You can paste the profile link.

Start from an example:

Sign in to run it. New accounts get 100 free credits; research uses credits for AI and data lookups.

How do you analyze an Instagram profile?

Read the account's latest posts and work out each one's engagement rate by followers and, for Reels, by plays; then take the median and look at which posts and formats sit above it. In Nolyn's October 18, 2026 sample, @owala posted about 5 times a week and its median post reached 0.47% of its 1.18M followers, while the Pokémon launch Reel reached 5.20%.

What you get

Engagement rate per post, two ways

By followers, (likes + comments) ÷ followers, as a median and range across the latest posts; and for Reels, (likes + comments) ÷ plays. Both formulas are shown, so you can compare accounts of different sizes fairly.

Posting frequency and format mix

How many posts a week, how many are Reels, carousels and photos, median likes and comments, and how far Reels reach beyond the followers.

What the best posts have in common

The five best and five weakest recent posts with their topics and formats, the themes that repeat among the best, giveaways flagged, and three takeaways for a brand or seller.

Example: @owala

A real run on October 18, 2026 for @owala, the verified account of the water-bottle brand (1.18M followers). Nolyn read the profile and its 24 most recent posts, published between September 13 and October 15, 2026: about 5 posts a week. None was pinned, hid its likes or belonged to another account.

Followers

1.18M

1,125 posts · verified · Product/service

Posting frequency

5.1 a week

24 posts, Sep 13–Oct 15, 2026: 18 Reels, 4 carousels, 2 photos

Engagement rate by followers

0.47%

Median (likes + comments) ÷ followers; range 0.035–5.20%

Reels engagement by plays

2.45%

Median of 18 Reels; median 286,147 plays per Reel

PostPostedFormatLikesCommentsPlaysBy followersBy plays
Owala | Pokémon launch at TargetSep 16Reel60,5468081,329,2835.20%4.62%
Fall lineup: “swap out your emotional support bottle”Sep 24Reel22,248115646,5071.90%3.46%
Fright and Delight Halloween bottle revealSep 13Reel21,210186460,1991.81%4.65%
Midnight Manor bottle giveaway (closed)Sep 14Reel14,1947,059320,9431.80%6.62%
Homestead cow-print bottle launch (collab post)Sep 25Reel13,238131404,1821.13%3.31%
Puffer Minis on campus (College Tour)Sep 21Reel1,1582879,8300.10%1.49%
College Tour clipSep 28Reel1,1283676,8340.099%1.51%
Mic’d up at a “human claw machine”Oct 3Reel7201860,2140.063%1.23%
FreeSip and Puffer outfit postOct 9Reel505950,5860.044%1.02%
All-Day Crossbody on a fall hikeOct 15Reel3991348,7770.035%0.84%

The five best and the five weakest of the 24 posts by engagement rate by followers = (likes + comments) ÷ 1,179,970 followers. “By plays” = (likes + comments) ÷ Instagram plays, for Reels only. Counts as Instagram returned them at collection time.

The Pokémon launch topped Instagram too

The September 16 Owala | Pokémon Reel drew 1,329,283 plays, 60,546 likes and 38,427 shares: 5.20% of followers, and 11 times the median post's 5,485 likes. The same launch is also the best video in our TikTok account and video analyzer examples.

Giveaways bring comments, not likes

Three closed giveaways (Midnight Manor, Amor y Vida and College Tour merch) hold 11,274 of the 13,648 comments on the 24 posts, 82.6%. The median post has 70 comments. When an account runs giveaways, compare it on likes or plays rather than comments.

Product reveals carry the account

All five best posts announce a product: the Pokémon launch, the fall lineup (646,507 plays), the Fright and Delight Halloween bottle (460,199), the Midnight Manor giveaway and the co-authored Homestead cow-print launch (404,182). All five are Reels.

Campus and outfit clips sit at the bottom

The five weakest posts (48,777 to 79,830 plays) are College Tour clips, a claw-machine skit and two accessory outfit posts, at 0.035–0.10% of followers. Even so, the median Reel reached 286,147 plays, about 24% of the follower count.

Every figure was recomputed from the saved Instagram responses: the profile and two pages of 12 posts, fetched through the same metered data layer the tool uses, with a Nolyn test account. The newest post (October 15) was three days old at collection, so its numbers were still growing. We show a brand account's public numbers only.

Run it for your own product or account

How it works

  1. Read the profile

    Followers, following, number of posts, bio, category and whether the account is verified.

  2. Read the recent posts

    The 24 latest posts with date, format, likes, comments and, for Reels, plays and shares. Pinned posts, hidden like counts and other accounts' collab posts are set aside.

  3. Calculate

    Posting frequency, format mix, median likes and comments, and engagement rate per post by followers and by plays, with the number of posts behind each figure.

  4. Explain what works

    Best and weakest posts, repeating themes and formats, giveaways flagged, and three takeaways you can use.

Who uses the Instagram profile analyzer

Brands benchmarking a competitor

See how often a competitor posts, which formats it leans on and which product reveals beat its own median post.

Agencies vetting a creator

Check a creator's median engagement by followers and by plays before you pay for a post, and spot giveaways that inflate comments.

Creators planning the next posts

Find which of your own posts and formats sit above your median and post more of what already works.

Limits

Only public accounts and the numbers Instagram shows publicly. Nolyn cannot see Instagram Insights such as reach, impressions, profile visits, follower demographics or sales, and saves are missing on many posts. Stories are not included. Figures describe the most recent posts at the time of the run, and the newest posts are still collecting likes. Engagement rates by followers fall as accounts grow, so compare accounts of a similar size with the same formula.

Questions

How do you calculate Instagram engagement rate?

Per post: (likes + comments) ÷ followers, then the median across the latest posts, which one viral post or giveaway distorts less than an average. For Reels, Nolyn also divides likes + comments by plays, which shows how a Reel did with the people who actually saw it. To work out one post without signing in, use the free Instagram engagement rate calculator.

What is a good engagement rate on Instagram?

It depends on the account size, the niche and the formula. Rates by followers drop as accounts grow: @owala, at 1.18M followers, had a median of 0.47% on October 18, 2026, while O Boticário, at 13.5M, had 0.013%. Compare an account with accounts of a similar size measured the same way.

Can I analyze a competitor's or an influencer's Instagram?

Yes, any public account. You see their posting frequency, formats, engagement per post and which topics work, but not their private Insights. For finding new creators for a product, use the influencer search tool.

Does it include Reels plays?

Yes. For every Reel Nolyn reads Instagram's play count and gives engagement by plays alongside engagement by followers, plus the median plays across the Reels in the sample.

Is this Instagram profile analyzer free?

You can start free: new accounts get 100 one-time credits, no card required. One analysis like this uses about 15–25 credits, so the free credits cover about 4–6 runs. Paid plans add monthly credits.

Do I need to sign up?

To run it, yes: sign in with Google. It is free and asks for no card. You can read the example and fill in the form before signing in.

Related tools

All tools

Ecommerce research guides & examples

Browse all resources

Research example

Stanley vs Owala vs HydroJug: what 447,586 Amazon ratings say about leaks and durability

Amazon review analysis of 5 best-selling insulated tumblers: 73% of Stanley Quencher leak mentions are negative; durability is Owala's and HydroJug's weak spot.

Research example

KKV and MINISO best sellers in China: what 31 popular Xiaohongshu posts show

What shoppers in China like at KKV and MINISO, from 31 popular Xiaohongshu (RedNote) posts: nail polish and must-buy lists at KKV, a guava lip balm at MINISO.

Research example

TikTok search trends, October 2026: 28 product searches with too few videos

What people in the US searched on TikTok in October 2026, and where videos are missing: 28 product searches with popularity, 7-day trend and video counts.

Research example

How many followers do you need to sell on TikTok Shop?

Sellers need no followers; affiliates need 1,000 (5,000 to lift limits). In our data, 27 of 74 product videos with 500K+ plays came from accounts under 10K.

Comparison

Kalodata alternatives: free and cheaper options, from EchoTik and FastMoss to AI research

Free and cheaper Kalodata alternatives compared on data, history, price and limits, checked on each official site: EchoTik's free plan, FastMoss, an AI research option, and where PiPiAds fits.

Comparison

FastMoss vs Kalodata: pricing, data and which to choose

FastMoss vs Kalodata compared on data scale, history, pricing, limits and extras, checked on both official sites, plus what sellers say on Reddit and where EchoTik and an AI research tool fit.

Comparison

FastMoss: what it is, what it costs and who it fits

What FastMoss is, Basic, Pro and Ultimate pricing (monthly and yearly), history and daily limits by plan, the 7-day free trial, and the 20 regions it covers. Checked on fastmoss.com on October 18, 2026.

Comparison

Kalodata pricing: what it costs, the free trial and cheaper options

How much is Kalodata? Monthly plans from its own App Store listing (US Basic $45.99, US Standard $99.99), how the 7-day free trial works, who it fits and cheaper or free options. Checked October 17, 2026.

Comparison

EchoTik: what it is, what the free plan covers and what it costs

What EchoTik is, what the free plan includes, Basic, Pro and Enterprise pricing ($13.9 to $57 a month), history and daily limits by plan, regions and refunds. Checked on echotik.live on October 18, 2026.

Comparison

Shoplus has shut down: what it was, refunds and where to go next

Shoplus, the TikTok analytics tool, stopped all services on September 18, 2026. What it offered, its last prices, the refund window, and alternatives with checked pricing. Checked on shoplus.net on October 18, 2026.